design, synthesis and evaluation of novel dna alkylating...

TitleDesign, Synthesis and Evaluation of Novel DNA AlkylatingAgents Based on the Chemistry of Antibiotic KapurimycinA[3]( Dissertation_全文 )

Author(s) Okamoto, Akimitsu

Citation 京都大学

Issue Date 1998-03-23

URL https://doi.org/10.11501/3135524

Right

Type Thesis or Dissertation

Textversion author

Kyoto University

Design, Synthesis and Evaluation

of Novel DNJ\. Alkylating Agents Based on

the Chemistry of Antibiotic Kapurimycin A3

Altimitsu Okamoto

1998

Preface

The study presented in this thesis has been carried out under the

direction of Professor Isao Saito at the Department of Synthetic

Chemistry and Biological Chemistry of Kyoto University during April,

1993 to March, 1998. The study is concerned with design, synthesis and

evaluation of novel DNA alkylating agents based on the chemistry of

antibiotic kapurimycin A3•

The author wishes to express his sincere gratitude to Professor Isao

Saito for his kind guidance, valuable suggestions, and encouragement

throughout this work. The author is deeply grateful to Associate

Professor Kazuhiko Nakatani for his constant advice, valuable

discussions, and encouragement during the course of this study. The

author is also indebted to Professor Hiroshi Sugiyama for their helpful

suggestions. The author is also indebted to Dr. Yoshikatsu Ito and Dr.

Kenzo Fujimoto for their helpful suggestions.

The author wishes to thank Mr. Yoichi Uosaki and Mr. Mitsunobu

Hara, Kyowa Hakko Kogyo Co., Ltd. for providing natural kapurimycin

A3 and its structural infonmation for comparative analysis. The author

is grateful to Mr. Haruo Fujita and Mr. Tadao Kobatake for the

measurements of NMR spectra and mass spectra, respectively. The

author is thankful to Professor Jun-ichi Yoshida, Professor Yoshihiko

Ito and Professor Junzo Sunamoto for the measurements of cyclic

voltammetry, optical rotation and circular dichromism, respectively.

The author wishes to N[essrs. Mikito Yamanuki, Kazuhito Tanabe,

Shinsuke Sando, Satoshi Okuda, Takahiro Matsuno and Shinya Hagihara

for their collaboration. The author is also indebted to Mr. Junya Shirai,

Mr. Takashi Nakamura and Dr. Masami Takayama

collaboration.

for their

The author is also grateful to Messrs. Satoshi Maekawa, Tomonori

Sakurai, Hideki Okita, Yasuki Komeda, Kiyohiko Kawai, Nobuhiro

Higashida, Kazuhiko Fujisawa, Taisuke Iwanami, Hiroshi Miyazaki,

Y ohei Ozeki, Chikara Dohno, Yusuke Nomura, Kaoru Adachi and

Mitsuhiro Iwasaki for their helpful suggestions and hearty

encouragement. The author also thanks to other members of Prof.

Saito's research group.

The author thanks Japan Society for the Promotion of Science for

financial support (Fellowship for Japanese Junior Scientists).

Finally, the author expresses his deep appreciation to his parents,

Mr. Katsumi Okamoto and Mrs. Etsuko Okamoto for their constant

assistance and affectionate encouragement.

Akimitsu Okamoto

January, 1998

Contents

General Introduction ----------------------------------------------------1

Chapter 1 Novel 4-Pyranone Ring Formation for the Synthesis

of Kapurimycin Analogs----------------------------------- 11

Chapter 2 Essential Structure for Efficient DNA Alkylation

by Kapurimycin A3----------------------------------------- 49

Chapter 3 Guanine-Guanine Sequence Selectivity for DNA

Alkylation by Kapurimycin Analogs --------------------- 79

Chapter 4 Effect of Absolute Configuration of Epoxy Subunit

on Guanine-Guanine Sequence Selective Alkylation----103

Chapter 5 Sequence Selective Alkylation of Continuous Guanine

Sequences by DNA Intercalators Possessing Epoxy

Side Chain--------------------------------------------------121

List of Publications ---------------------------------------------------140

General Introduction

DNA sequence selectivity is achieved by both proteins 1 and small

molecules.2 DNA-binding proteins have evolved to recognize the

sequence dependent features of DNA in order to participate in a variety

of genetic events such as control of gene expression and replication of

DNA. 3 While overall sequence selectivity is generally lower, small

molecules such as DNA reactive drugs and carcinogens may also exhibit

DNA sequence recognition. These drugs and carcinogens may share

common DNA recognition motifs with proteins and because of their

relatively small molecular weight and corresponding reduced structural

complexity, they may serve as useful models for more complex protein

DNA recognition mechanisms.

DNA basically has the nature to undergo electrophilic modification

by alkylating agents. Nitrogen atom of guanine N7 position in DNA is

one of the most nucleophilic position in DNA. A large number of

guanine N7 alkylating agents are known at present such as aflatoxin B 1

oxide,4 pluramycin A,5 hedamycin6 as naturally occurred molecules, as

well as dimethyl sulfate,7 nitrogen mustard8 and bromoacetate

derivatives. 9 These DNA guanine N7 alkylating agents separates to

mainly two groups, the molecules which alkylate all guanine equally

(dimethyl sulfate and bromoacetate derivative are typical) and the

molecules which recognize sequence of some base including guanine

base and alkylate guanine N7 such as aflatoxin and pluramycin. The

latter type of guanine N7 alkylating agents were shown in Figure 1.

Each molecule in Figure 1 has a unique sequence selectivity for their

alkylation.

Me2

A cO

pluramycin A 5'-CG*-3'

0

NMe2

OMe odl:: 0

HOJ;:J

Me2

HO

altromycin B 5'-AG*-3'

0

hedamycin 5'-CG*-3'

0 0

NMe2

OMe odl::0

HOJ;:J 0

altromycin H 5'-AG*-3', 5'-TG*-3'

aflatoxin B 1 oxide 5'-GG*-3'

Figure 1. Structure and DNA sequence preferences for natural alkylating agents. G* denotes the alkylated guanine.

Aflatoxin B1 is a DNA-reactive natural carcinogen. Humayun et al. 4a

and Loechler et al.4c described extensive and carefully executed

experiments revealing the sequence selectivity of activated aflatoxin B 1

reaction with DNA. The statistical treatment of the data gave

trinucleotide priority to 5' -GG*G, followed closely by 5' -GG*T. They

invoked a sequence-dependent noncovalent binding site for aflatoxin B 1

oxide in DNA major groove. The importance of binding interactions

was also supported by other lines of evidence which they listed.

The sequence selectivity of DNA alkylation by recently isolated

pluramycin antibiotics has been also investigated by Hurley et al.6e.Io and

Wickham et al.6d They have divided these pluramycins into three groups

of sequence selectivity. The first group represented by pluramycin A

and hedamycin showed 5' -CG* > 5'-TG* >> 5' -AG* = 5' -GG* sequence

selectivity. Altromycin B preferred 5' -AG* sequence and altromycin H

showed the greatest reactivity with 5' -AG* and 5'-TG* sequences. 10a·

11

These agents which become alkylated to DNA in a sequence-selective

fashion may recognize the sequence by either noncovalent binding or by

alkylating mechanisms, or perhaps more likely by a combination of both

mechanisms. The sequence selective alkylating agents in the Figure 1

possess the common structural feature involving aromatic rings for

DNA binding and epoxide ring for DNA alkylation. However, it has

been unknown how their structures work as a factor for their sequence

selective alkylation. In order to get considerable insights into the

molecular basis of its sequence selective reaction with DNA, it is very

important to do experiments designed to maximize sequence selectivity

information.

In 1990, the novel antitumor antibiotic, kapurimycin A3, has been

found in the culture of Streptomyces sp. D0-115 by Kyowa Hakko

Kogyo group. 12'13 It exhibited strong antibacterial and cytotoxic

2 3

activities and showed a potent activity against murine leukemia P388 in

vivo. It consists of the tetrahydroanthra-y-pyranone skeleton and the

{3, y-unsaturated 8-keto carboxylic acid structure. In addition, at C2

position of ring skeleton it has the side chain with propenyl substituted

epoxide in Z-configuration.

Kyowa Hakko Kogyo group investigated covalent modification and

strand scission of DNA by kapurimycin A3 in 1990.14

They described

that it caused single strand cleavage of supercoiled pBR322 DNA and

binded to calf thymus DNA and the thermal treatment of this adduct

resulted in a release of a guanine attached to C 16 of kapurimycin A3

through one of its nitrogen atoms. Previously, our group also

investigated DNA cleavage reaction of kapurimycin A3 with

deoxytetranucleotide d(CGCG)2 •15 This study provided evidence that

kapurimycin A3

alkylated DNA at N7 of guanine to produce a

thermolabile adduct which could undergo depurination to produce a

more stable kapurimycin-guanine adduct, together with the formation

of its abasic site-containing oligomer. However, the relationship

between the structural characteristics of kapurimycin A3 and the guanine

selectivity in its DNA alkylation remains unsolved.

OAc

kapurimycin A3 (1)

In spite of a large number of the studies on the reactivity and

sequence selectivity of natural compounds that alkylate guanine N7 in

4

DNA, it is not enough to discuss the general structural requirements for

both their reactivities to guanine and their sequence selectivities for

DNA alkylation. The synthetic analogs of reduced structural complexity

for the natural products would serve as useful models for the more

complex natural products-DNA recognition and alkylation mechanisms.

For exemplary purposes, we designed the kapurimycin analogs with a

very simple structure which just consists of aromatic rings for DNA

binding and epoxide side chain for DNA alkylation. Subsequently, these

analogs have been synthesized and compared in DNA alkylation in order

to understand the molecular basis for the DNA sequence selectivity of

DNA alkylation.

This thesis consists of five chapters on the design, synthesis and

evaluation of novel DNA alkylating agents based on the chemistry of

antibiotic kapurimycin A3.

In chapter 1, novel 4-pyranone ring formation for the synthesis of

kapurimycin analogs is reported. The 6-endo-digonal selective

cyclization method of o-silyloxyphenyl ethynyl ketones under mild basic

conditions was developed.

In chapter 2, essential structure for efficient DNA alkylation by

kapurimycin A3 is discussed. It was shown that ABC-ring analog of

kapurimycin A3 effectively alkylated DNA guanine bases and the

sequence selectivity of DNA alkylation was closely similar to that of

natural kapurimycin A3•

In chapter 3, guanine-guanine sequence selectivity for DNA

alkylation by kapurimycin analogs is described. By comparison of DNA

alkylation by kapurimycin analogs involving different number of

aromatic rings, the mechanism of guanine-guanine sequence selective

DNA alkylation by kapurimycin analogs was elucidated.

5

In chapter 4, the effect of absolute configuration of epoxy subunit on

guanine-guanine sequence selective alkylation is discussed. By

comparison of DNA alkylation by both enantiomers of kapurimycin

analogs, it was shown that the stereochemical orientation of epoxide in

the complex with DNA is considerably important for sequence selective

alkylation by kapurimycin analogs.

In chapter 5, the sequence selective alkylation of continuous guanine

sequences by DNA intercalators possessing epoxy side chain is discussed.

It was shown that the interaction of DNA HOMO with LUMO of DNA

alkylating agents played a critical role in the selectivity toward

continuous guanine sequences such as GGG.

6

References

(1) (a) Ollis, D. L.; White, S. W. Chern. Rev. 1982, 87, 981-995. (b)

von Rippel, P. H.; Berg, 0. G. Proc. Natl. Acad. Sci. USA. 1986, 83,

1608-1612.

(2) (a) Dabrowiak, J. C. Life Sci. 1983, 32, 2915-2931. (b) Hertley, J.

A.; Gibson, N. W.; Kohn, K. W.; Mattes, W. B. Cancer Res. 1986, 46,

1943-1947. (c) Warpehoski, M. A.; Hurley, L. H. Chern. Res. Toxicol.

1988, 1' 315-333.

(3) Gniazdowski, M.; Cera, C. Chern. Rev. 1996, 96, 619-634.

(4) (a) Muench, K. F.; Misra, R. P.; Humayun, M. Z. Proc. Natl. Acad.

Sci. USA 1983, 80, 6-10. (b) Stone, M. P.; Gopalakrishnan, S.; Harris,

T. M.; Graves, D. E. J. Biomol. Struct. Dyn. 1988, 5, 1025-1041. (c)

Benasutti, M.; Ejadi, S.; Whitlow, M. D.; Loechler, E. L. Biochemistry

1988, 27, 472-481. (d) Baertschi, S. W.; Raney, K. D.; Stone, M.P.;

Harris, T. M. J. Am. Chern. Soc. 1988, 110, 7929-7931. (e)

Gopalakrishnan, S.; Byrd, S.; Stone, M. P.; Harris, T. M. Biochemistry

1989, 28, 726-734. (f) Stone, M. P.; Gopalakrishnan, S.; Raney, K.

D.; Raney, V. M.; Byrd, S.; Harris, T. M. In Molecular Basis of

Specificity in Nucleic Acid-Drug Interactions; Pullman, B., Jortner, J.,

Eds.; Kluwer Academic Publishers: 1990, p 451-480. (g) Raney, K. D.;

Gopalakrishnan, S.; Byrd, S.; Stone, M. P.; Harris, T. M. Chern. Res.

Toxicol. 1990, 3, 254-261. (h) Gopalakrishnan, S.; Harris, T. M.;

Stone, M. P. Biochemistry 1990, 29, 10438-10448. (i) Baertschi, S.

W.; Raney, K. D.; Stone, M. P.; Harris, T. M. J. Am. Chern. Soc. 1991 '

113, 4092-4096. (j) Raney, K. D.; S.; Harris, T. M.; Stone, M. P.

Chern. Res. Toxicol. 1993, 6, 64-68.

(5) (a) Maeda, K.; Takeuchi, T.; Nitta, K.; Yagishita, K.; Utahara, R.;

Osato, T.; Ueda, M.; Kondo, S.; Okami, Y.; Umesawa, H. J. Antibiot.

7

Ser. A 1956, 9, 75-81. (b) Kondo, S.; Miyamoto, M.; Naganawa, H.;

Takeuchi, T.; Umezawa, H. J. Antibiot. 1977, 30, 1143-1145. (c) For

a review of the antibiotics of the pluramycin group, see: Sequin, U.

Fortschr. Chern. Naturst. 1986, 50, 57-122.

(6) (a) Sequin, U.; Bedford, C. T.; Chung, S. K.; Scott, A. I. Helv.

Chim. Acta 1977, 60, 896-906. (b) Zehnder, M.; Sequin, U.; Nadig,

H. Helv. Chim. Acta 1979, 62, 2525-2533. (c) Bennett, G. N. Nucleic

Acids Res. 1982, 10, 4581-4594. (d) Prakash, A. S.; Moore, A. G.;

Murray, V.; Matias, C.; McFadyen, W. D.; Wickham, G. Chem.-Biol.

Interact. 1995, 17-28. (e) Hansen, M.; Yun, S.; Hurley, L. Chern. Biol.

1995, 2, 229-240.

(7) Maxam, A.; Gilbert, W. Methods Enzymol. 1980, 65, 499-560.

(8) Hartley, J. A. In Molecular Basis of Specificity in Nucleic Acid

Drug Interactions; Pullman, B., Jortner, J., Eds.; Kluwer Academic

Publishers: 1990, p 513-530.

(9) (a) Povsic, T. J.; Darvan, P. B. J. Am. Chern. Soc. 1990, 112,

9428-9430. (b) Povsic, T. J.; Strobel, S. A.; Darvan, P. B. J. Am.

Chern. Soc. 1992, 114, 5934-5941. (c) Taylor, M. J.; Darvan, P. B.

Bioconjugate Chern. 1997, 8, 354-364.

(10) (a) Sun, D.; Hansen, M.; Clement, J. J.; Hurley, L. Biochemistry

1993, 32, 8068-8074. (b) Sun, D.; Hansen, M.; Hurley, L. J. Am.

Chern. Soc. 1995, 117, 2430-2440. (c) Hansen, M.; Hurley, L. Ace.

Chern. Res. 1996, 29, 249-258.

(11) (a) Jackson, M.; Karwowski, J. P.; Theriault, R. J.; Hardy, D, J.;

Swanson, S, J.; Barlow, G, J.; Tillis, P.M.; McAlpine, J. B. J. Antibiot.

1990, 43, 223-228. (b) Brill, G. M.; McAlpine, J. B.; Whisttem, D.

N.; Buko, A. M J. Antibiot. 1990, 43, 229-237. (c) Hansen, M;

Hurley, H. J. Am. Chern. Soc. 1995, 117, 2421-2429. (d) Sun, D.;

Hansen, M.; Hurley, H. J. Am. Chern. Soc. 1995, 117, 2430-2440.

8

(12) (a) Hara, M.; Mokudai, T.; Kobayashi, E.; Gomi, K.; Nakano, H. J.

Antibiot. 1990,43, 1513-1518. (b) Yoshida, M.; Hara, M.; Saitoh, Y.;

Sano, H. J. Antibiot. 1990,43, 1519-1523.

(13) The absolute configuration of 1 has been determined to be 8S, 14S

and 16S. Uosaki, Y.; Saito, H. Abstract paper p 1013, 69th annual

meeting of the Chemical Society of Japan, Kyoto (1995).

(14) Hara, M.; Yoshida, M.; Nakano, H. Biochemistry 1990, 29,

10449-10455.

(15) (a) Chan, K. L.; Sugiyama, H.; Saito, I. Tetrahedron Lett. 1991,

52, 7719-7722. (b) Chan, K. L.; Sugiyama, H.; Saito, I.; Hara, M.

Phytochemistry 1995, 40, 1373-1374.

9

CHAPTER 1

Novel 4-Pyranone Ring Formation for the Synthesis of Kapurimycin Analogs

Abstract: The cyclization of o-hydroxyphenyl ethynyl ketones was

examined from theoretical and experimental standpoints in order to

develop efficient synthetic methods for the construction of 2-substituted

pyranones of significant biological activities. Ab initio studies at HF/6-

31 G* level on the cyclization indicated that both 6-endo-digonal and 5-

exo-digonal cyclizations giving benzopyranones and benzofuranones,

respectively, were endothermic and reversible in aprotic media, and the

irreversible protonation of the resulting anions would be critical for the

product formation. We generated phenoxide ion under aprotic

conditions in situ by desilylation of o-silyloxyphenyl ethynyl ketones

with potassium fluoride and 18-crown-6 in anhydrous DMF. Under

these conditions the cyclization of variety o-hydroxyphenyl ethynyl

ketones proceeded smoothly to produce benzopyranone derivatives with

exceedingly high selectivity. Theoretical and experimental results

strongly suggested that the presence of a small amount of proton donor

effecting the protonation of the resulting benzopyranone an1on was

essential for the high 6-endo-digonal selectivity.

11

Introduction

Kapurimycin A 1 (1) is an antitumor antibiotic possessing a novel

anthra-y-pyrone ring system with a vinyl epoxide side chain at the C2

position. 1 The structure of 1 is closely resembled to those of

pluramycin family antibiotics / which have a common 4H-anthra[ 1,2-

b ]pyrane ring system and characteristic functionalities attached to the C2

position as well as deoxyamino sugars at C8 and C 10 positions.

Pluramycin A (2) ,3 hedamycin (3)4 and 1 all having epoxide

functionalities on the side chain attached to the C2 position are known to

covalently bound to DNA by a nucleophilic ring opening of the epoxide

with guanine N7 in DNA.5-

8

AcO C02H

kapurimycin A3 (1)

Me2 A cO

0

pluramycin A (2) hedamycin (3)

In spite of the significant biological features and unique reactivity

toward DNA, the structure-activity relationships on these antibiotics

was not examined primarily due to the difficulty in the synthesis of 2-

substituted benzopyranone ring system.9·1° For example, this problem

has been addressed by the synthesis of 0-methylkidamycinone (5) (eq

1 ), 9

showing that conventional synthetic scheme for benzopyranone

systems using acid-catalyzed cyclization of 1-(o-hydroxyphenyl)-1 ,3-

diketones11 '12 (e.g., 4) is not applicable to the synthesis of these

12

antibiotics because of the competitive formation of undesired

dihydropyranone (e.g., 6).

0

4

R

HO II 0

7

o-hydroxyphenyl ethynyl ketone

0

0

0-methylkidamycinone (5)

0

6

R R

?} & + 0 0

#

8 9

benzopyranone benzofuranone via 6-endo-digonal via 5-exo-digonal

cyclization cyclization

(eq 1)

(eq 2)

To investigate structure-activity relationship of these antibiotics we

focused our attention on developing efficient synthetic methods for 2-

substituted benzopyranone ring systems from readily available

precursors. As a candidate for such a process, the 6-endo-digonal

cyclization of o-hydroxyphenyl ethynyl ketones was first examined (eq

2), because the starting phenyl ethynyl ketones could be readily

synthesized from salicylic aldehyde and acetylenic compounds.

13

According to the Baldwin's rule 13 the 6-endo-digonal cyclization is a

favorable process, although the cyclization of o-hydroxyphenyl ethynyl

ketones 7 under basic conditions are reported to produce not only

benzopyranone 8 via the 6-endo-digonal cyclization but also

benzofuranone 9 by a simultaneous 5-exo-digonal cyclization, with the

product ratio being highly dependent on the reaction conditions. 14-16 To

get insight into the factors governing the 6-endo-digonal and the 5-exo

digonal cyclization of o-hydroxyphenyl ethynyl ketones, we have

investigated this reaction from theoretical and experimental viewpoints.

We herein describe experimental results in combination with theoretical

calculations indicating that both 6-endo-digonal and 5-exo-digonal

cyclizations of o-hydroxyphenyl ethynyl ketones are reversible in

aprotic media, and that the irreversible protonation of the resulting

vinyl anion gives rise to the benzopyranone formation with exceedingly

high selectivity .17

R

2)11

I~ o ~

10

(For calculation studies R denotes Me)

Results and Discussion

Theoretical Calculations for 6-Endo-digonal and 5-Exo-digonal

Cyclizations.

To discuss the cyclization of o-hydroxyphenyl ethynyl ketone in

detail, ab initio molecular orbital calculations of phenoxide ion 10,

14

vinyl anions 11 and 12 (where R denotes Me), and two transition states

TS-6 and TS-5 for the 6-endo-digonal and 5-exo-digonal

cyclizations, respectively, were carried out. 18·19 While in our

preliminary communication 17 we reported the theoretical calculations at

the HF/3-21G(*) level, more accurate calculations at higher HF/6-31G*

level was performed at this time for precise discussions. For these

calculations the initial structures for 10, 11, and 12 were surveyed at

the semiempirical PM3 level. Two stable s-trans and s-cis conformers

were found for 10, with the former being more stable than the latter by

1.80 kcal/mol. Therefore, the s-trans conformer of 10 shown in Figure

1 was used for further calculations. 20 While the reaction of nucleophiles

to the carbon-carbon triple bond may proceed via either syn or anti

addition, 19ct·21 the £-configuration for the exocyclic alkene in 12

supported by previous theoretical studies 19b,c was used for the

calculation. We could not develop a reasonable transition state model

for the syn addition of the phenoxide ion in 10 to the carbon-carbon

triple bond via the 5-exo-digonal cyclization. These structures obtained

by the PM3 calculations were optimized at the HF/3-21G(*) and then at

the HF/6-31G* level. Two transition states TS-6 and TS-5 were

initially generated empirically using the transition structure module

incorporated in Spartan 18 and finally calculated at the HF/6-31 G* level.

Frequency analyses for TS-6 and TS-5 showed the only one imaginary

vibrational frequency at -524.62 and -525.25 cm-1, respectively (Figure

1). 22 In both structures the approaching angle of the phenoxide ion to

the carbon-carbon triple bond was 115.4° and 118.9°, respectively. The

potential energy diagram for the reaction of 10 to 11 and 12 was

shown in Figure 2. The potential energies of two transition states TS-6

and TS-5 were very close in each other, therefore there was no

obvious difference in the activation energies for the two cyclization

15

TS-5 (- 532.399409)

0 - C8 1.840 A LO-Cg- C9 = 118.9°

TS-6 (- 532.400397)

0- C9 1.898 A LO-C9-Cg = 115.4°

12 (- 532.409586) 10 ( - 532.438009) 11 (- 532.421144)

Figure 1. Optimized structures for 10, 11, 12, TS-6, and TS-5 at the HF/6-31G* level. Numbers in parenthesis indicated the total energy in hartree.

TS-5 ( +24.22) TS-6 ( +23.60)

12 ( + 17.84)

11 (+10.58)

10 (0.0)

Figure 2. Potential energy diagram for the reaction of 1 0 to 11 and 12. Numbers in parenthesis indicated the relative potential energy from 10 in kcal/mol.

16

processes (23.60 kcal/mol for .10+10---711 and 24.22 kcal/mol for

.10+10---712). On the other hand, the produced vinyl anion 11 was

more stable than anion 12 by 7.26 kcal!mol. As a result the activation

energy for the ring-opening reaction of 12 to 10 (6.38 kcal/mol) was

substantially smaller than that for the conversion of 11 to 10 (13 .02

kcal!mol). Since both cyclization reactions were shown to be

endothermic processes, the irreversible protonation of the resulting

anions 11 and 12 would be critical for the product formation. These

theoretical results led to the following speculations for the cyclization of

10. 1) Under kinetically controlled conditions the selectivity for the 6-

endo-digonal cyclization would not be so high. 2) Under

thermodynamic conditions the product formation is favorable for the 6-

endo-digonal process, if all three anions were equilibrated and the

selective protonation of 11 proceeded irreversibly.

Synthesis of o-Silyloxyphenyl Ethynyl Ketones.

To achieve thermodynamically controlled reaction conditions for the

selective 6-endo-digonal cyclization, we examined in situ generation of

the phenoxide in an aprotic medium by desilylation of o-silyloxyphenyl

ethynyl ketone with fluoride. The o-silyloxyphenyl ethynyl ketone 16

used for the cyclization studies was synthesized from the silyl-protected

salicylic aldehyde 13 and readily available 14 (Scheme 1). Addition of

bromomagnesium salt 14 to 13 gave benzyl alcohol 15. Oxidation of

15 with manganese dioxide (Mn02) cleanly produced ethynyl ketone

16. Desilylation of 15 followed by oxidation with Mn02 produced

phenol 18.

17

Scheme 1. 10THP TBSO

1

1

1

-OCHO _M_g_Br_1.,...4

OTHP OTHP

OH

13 15 16

~TBAF OTHP OTHP

OH

17 18

Effects of Reaction Conditions for The 6-Endo-digonal Cyclization.

Various reaction conditions for the 6-endo-digonal cyclization were

tested by using 16. Considering that the protonation of the vinyl anion

is essential for the product formation, we first examined the

commercially available THF solution of tetra-n-butylammonium

fluoride (TBAF) containing approximately 5% (v/v) of water as a

fluoride ion source. The reaction of 16 with TBAF in THF at 0 OC for

1.5 h produced both 19 and 20 in 90% yield with very low selectivity

(19:20 = 47:53) (eq 3). In the early stage of the reaction the phenol 18

was detected on TLC indicating the existence of phenoxide ion under the

conditions, which slowly underwent cyclization to 19 and 20. The

structure of 19 was unambiguously confirmed by the HMBC spectrum

indicating the hydrogen-carbon connectivities as shown in Figure 3.

The stereochemistry of the exocyclic alkene in 20 was not confirmed by

spectroscopic methods, but transition state for the 5-exo-digonal

cyclization (TS-5) may suggest the preferential formation of Z-isomer.

To reduce the concentration of proton donor (e.g., water) in the

18

reaction system, spray-dried potassium fluoride (KF) in the presence of

18-crown-6 was used for the fluoride source. The reaction of 16 with

spray dried KF-18-crown-6 in anhydrous DMF proceeded smoothly

giving 19 as a sole product in a quantitative yield (97o/o ). The

formation of 20 was not detected by 1H NMR analysis of the crude

mixture.

~

TBSO II 0

16

OTHP OTHP

F _____.,.... + 0

19 20

Figure 3. The selected hydrogen-carbon conectivities observed in the HMBC spectrum of 19.

OTHP

0 (eq 3)

To confirm the speculation that irreversible protonation of the vinyl

anion is essential for the product formation, the reaction of 16 under

KF-18-crown-6-DMF conditions was quenched with deuterated acetic

acid (CH3C02D) after the standard reaction period (2 h) at ambient

temperature. As expected there was no sign of the incorporation of the

deuterium into 19, presumably due to in situ quenching of the resulting

19

anion by moisture already contaminated in the reaction system. The

observation that the cyclization of 16 became exceedingly slow when

the reaction was carried out in the presence of activated molecular

sieves 4A, may support in situ protonation of the resulting anion. On

the other hand, the addition of a large excess of methanol (20% v/v) to

the reaction mixture dramatically changed the reaction course giving 2 0

in 66% yield accompanied by the minor formation of 19 (13%). The

formation of the substantial amount of 20 (9o/o) along with 19 (83%) by

the cyclization of phenol 18 under KF-18-crown-6-DMF conditions

revealed that even the phenolic hydrogen could be effective as a proton

donor in the cyclization to result in a decrease of the selectivity for the

formation of 19. These results indicated that the presence of only a

small amount of proton donor like moisture in the reaction system plays

a critical role not only in governing the selectivity but also in the

smooth product formation. Evidence for the vinyl anion formation in

the cyclization of 16 was obtained when the cyclization was carried out

in DMF-CH30D (99 atom% D) (4:1) solution (eq 4). Both d1-19 and

d 1-20 formed in a ratio of 1:5 contained deuterium at the exocyclic

olefinic position with the deuterium incorporation efficiency being

more than 97o/o in both cases.

OTHP OTHP

0..._97% KF, 18-crown-6

(eq 4) DMF-CH30D (4: 1) 0

16

20

H Me~OTHP

e

d

c

b

a I I I I I J I I I I I j j I I} I I I I I j j I I J~ I I I I

Figure 4. Selected 1H NMR spectra (5.5- 8.5 ppm) of the crude mixture for the reaction of 16 under the KF-18-crown-6-DMF conditions after aqueous work-up (aq. NH4Cl) at the indicated reaction time. The compounds were identified by the triplet-like signals of the olefinic hydrogen observed at 6.11 ppm for 16, 6.21 ppm for 18, 6.06 ppm for 19, and 5.90 ppm for 20. line a; 0 min, 16, line b; 10 min at -20 °C, line c; 30 min at -20 °C, line d; 1 h at -20 °C, line e; warming up to 0 OC after 1 h at -20 °c. * DMF

The time-course of the cyclization of 16 was monitored by 1H NMR

spectroscopy. The reaction mixtures of 16 under KF-18-crown-

6-DMF conditions at -20 OC were subjected to aqueous work-up at an

indicated time interval and the 1H NMR of each of the crude mixture

was recorded (Figure 4 ). The starting material 16 was no more

detected after 10 min reaction, with phenol 18, benzopyranone 19, and

benzofuranone 20 being observed in a ratio of 29:52: 19 (line b). While

after 30 min the ratio of three compounds reached to 9:54:37 (line c),

upon prolonged reaction (1 h) phenol 18 was completely consumed,

21

with the ratio of 19 and 20 being 81: 19 (line d). Warming the reaction

mixture to 0 OC followed by aqueous work-up resulted in an almost

exclusive formation of 19 (line e) . The fact that the amount of the

initially formed benzofuranone 20 decreased on a prolonged reaction

with increase of benzopyranone 19 suggested that there was an

equilibrium among either 18, 19, and 20 or their anions.

To identify the stage for the equilibration, we examined the

interconversion between 19 and 20 under KF-18-crown-6-DMF

conditions and found that both compounds were absolutely inert at

ambient temperature under the conditions.23 It was also confirmed that

pentacoordinate silicate having strong Lewis acidity formed in situ in

the reaction did not induce the ring opening of 20. Thus, the reaction

of a mixture of 16 and 20 (approximately 1:1) under KF-18-crown-6-

DMF conditions afforded 19 with a complete recovery of 20 (eq 5).

These experiments clearly indicated that protonation of benzopyranone

and benzofuranone anions (e.g., 11 and 12, respectively) was an

irreversible process under the conditions, and the equilibration should,

therefore, exist at anion states. When the reaction of 16 under the KF-

18-crown-6-DMF conditions was quenched with D20 before the

equilibration is completed (e.g., after 15 min at -20 OC), it was

confinned that deuterium is efficiently incorporated into the exocyclic

alkenic position of benzofuranone, while it was not the case for the

benzopyranone (Figure 5). Thus, under these conditions protonation of

the benzofuranone anion by a small amount of proton donor existing in

the reaction system was much less efficient than that of benzopyranone

an1on.

22

OTHP

+

16 20

20..H1

OTHP ~ OTHP

KF, 18-crown-6 §:"=:: + 0

0 DMF I "=:: 0

.&

OTHP

0

19 20 (unchanged)

19-H1

t * 20-H2

(eq 5)

Figure 5. Selected 1H NMR spectrum (5.8 - 6.9 ppm) of the crude mixture

obtained by quenching the reaction of 16 under the KF-18-crown-6- DMF

conditions with D20-ND4Cl after 15 min at -20 °C. The newly formed olefinic

hydrogens for 19 and 20 were indicated as 19-H1 and 20-Hl, respectively.

Hydrogens attached to the trisubstituted alkene were labled as 19-H2 and 20-

H2. A small singal for 20-Hl in comparison with that for 20-H2 clearly

indicated the incorporation of deutrium at this position. The signal marked

with asterisk was the olefinic hydrogen of 18.

Considering theoretical calculations and the experimental results

obtained above, we can rationalize the cyclization reaction of o

hydroxyphenyl ethynyl ketones under basic conditions as illustrated in

Scheme 2. Under the conditions where there is a sufficient amount of

proton donor, the reaction produces varying amounts of both

benzopyranone and benzofuranone anions 11 and 12, which are

protonated to give 8 and 9, respectively. On the other hand, under the

conditions where only a limited amount of the proton donor was

available, all three anions 10, 11 and 12 are equilibrated. The most

stable phenoxide anion 10 would be expected to be preferentially

23

protonated to give phenol 7. However, under the basic reaction

conditions, 7, if formed, would be equilibrated with 10 immediately.

While the protonation of the benzofuranone anion 12 was relatively

slow under the conditions, the benzopyranone anion 11 was

immediately and irreversibly protonated. As a result of the equilibrium

among these three anions and of the irreversible protonation of 11,

highly selective formation of benzopyranone 8 was attained. However,

at this moment we do not know the reason why the protonation of 12 is

relatively slow compared with that for 11.

Scheme 2.

9

TS-5 TS-6 - -I t 1 \

I ~ 2!: 11 o ~ I : o

-H+H +H+ R

2}0 7

Synthesis of Various 2-Substituted Benzopyranones.

With an efficient synthetic method for the selective formation of

benzopyranones in hand, we examined the reaction of phenyl ethyny1

ketones 21, 22, 23 , and 24 under KF-18-crown-6-DMF conditions.

The epoxy-substituted ketones 22 and 23 were synthesized using epoxy

alkynes prepared from commercially available (Z)- and (£)-3-methyl-2-

penten-4-yn-1-ol, respectively, as detailed in experimental section. As

24

expected benzopyranones 25, 26, 27, and 28 were selectively obtained

in good to excellent yields as indicated in the parenthesis. The

hydrolysis of THP group of 28 to the known 2-hydroxymethyl-4H

chromen-4-one gave further evidence for the structure.24 Under these

conditions the stereochemical integrity for the carbon-carbon double

bond and the epoxide moiety was completely retained.

OTHP OEE

OEE OTHP

0

21 22 23 24

OTHP OEE

0

§HP ~

0 0

~

25 (83%) 26 (73%) 27 (72%) 28 (73%)

To examine the feasibility of this synthetic method for kapurimycin

A3 synthesis, we investigated the construction of tricyclic ring system as

a simple kapurimycin model. Ethynyl ketone 29 was prepared from 1-

silyloxy-2-naphthaldehyde according to the procedure described for the

synthesis of 16. The cyclization of 29 under KF-18-crown-6-DMF

conditions at ambient temperature proceeded smoothly giving the

tricyclic compound 30 in 81% yield (Scheme 3). These results clearly

indicated that our method has a high potential for the synthesis of 1 and

25

its congeners to study the structure-reactivity relationship of

kapurimycin A3 .

Scheme 3.

KF, 18-crown-6

DMF

81%

29 30

26

Experimental Section

General Procedures. Theoretical calculations were performed on

SGI INDY (R4000SC personal workstation) with Spartan molecular

modeling software (version 3.1) and Gaussian 92 program. 1H NMR

spectra were measured with Varian GEMINI 200 (200 MHz), JEOL

JNM a-400 (400 MHz) and JEOL JNM a-500 (500 MHz)

spectrometers. Coupling constants (J values) are reported in Hz. 13C

NMR spectra were measured with Varian GEMINI 200 (50 MHz),

JEOL JNM a-400 (100 MHz) and JEOL JNM a-500 (125 MHz)

spectrometers. The chemical shifts are expressed in ppm downfield

from tetramethylsilane, using residual chloroform ( 8 = 7.24 in 1H

NMR, 8 = 77.0 in 13C NMR) and dimethylsulfoxide ( 8 = 2.49 in 1H

NMR, 8 = 39.5 in 13C NMR) as an internal standard. The following

abbreviations were used for the description of the signal multiplicity: s,

singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad. IR

spectra were recorded on a JASCO FT!IR-5M spectrophotometer.

Melting points were obtained on a Yanagimoto Seisakusho micro

melting point apparatus and are uncorrected. Electron impact mass

spectra (MS) and high-resolution mass spectra (HRMS) were recorded

on JEOL JMS-DX 300 or JEOL JMS-SX 102A. Precoated TLC plates

Merck silica gel 60 F254 was used for monitoring the reactions and also

for preparative TLC. Wako gel (C-200, particle size 75-150 pm,

Wako) was used for silica gel flash chromatography. Anhydrous

reactions were performed under N 2 atmosphere. Ether and

tetrahydrofuran (THF) were distilled under N2 from

sodiurn!benzophenone ketyl pnor to use. Yields refer to

chromatographically and spectroscopically CH NMR) homogeneous

materials, unless otherwise stated.

27

2-(t-Butyldimethylsilyloxy)benzaldehyde (13). To a solution of

salicylaldehyde (1.21 g, 10.8 mmol) and 2,6-lutidine (1.75 mL, 15.0

mmol) in dichloromethane (30 mL) was added t-butyldimethylsilyl

trifluoromethanesulfonate (3.50 mL, 15.2 mmol) at -78 OC, and the

mixture was stirred at -78 OC for 2 h. After diluted with sat. aq.

NaHC03 the reaction mixture was warmed to ambient temperature and

extracted with ethyl acetate. The organic phase was washed with brine,

dried over anhydrous N~S04, filtered, and concentrated in vacuo. The

crude product was purified by flash chromatography (Si02

, 2% ethyl

acetate/hexane) to give 13 (2.26 g, 89%) as a yellow oil: 1H NMR (400

MHz, CDC13) 8 10.45 (d, 1H, J = 1.0 Hz), 7.79 (dd, 1H, J = 1.9, 7.7

Hz), 7.44 (ddd, 1H, J = 2.0, 7.3, 8.4 Hz), 7.02 (m, 1H), 6.87 (dd, 1H, J

= 1.0, 8.3 Hz), 1.00 (s, 9H), 0.26 (s, 6H); IR (CHC13

) 3016, 2957, 2932,

2860, 1684, 1600, 1478, 1256, 1217 cm-1; MS m/e (%) 179 [(M-13ut]

(56), 57 (100); HRMS calcd for C9H 11 0 2Si [(M-13utJ, 179.0528; found,

179.0506.

(Z)-1-[2-(t- B utyldimethylsily loxy )phenyl] -4-methyl-6-(2-

tetrahydropyranyloxy)-4-hexen-2-yn-1-ol (15). To a solution

of (Z)-3-methyl-2-penten-4-yn-1-ol (5.30 g, 55.1 mmol) and 3,4-

dihydro-2H-pyran (15.0 mL, 164 mmol) in dichloromethane (60 mL)

was added a catalytic amount of pyridinium p-toluenesulfonate (PPTS)

at 0 °C, and the mixture was stirred for 3 h. The reaction mixture was

diluted with sat. aq. NaHC03 and extracted with ethyl acetate. The

organic phase was washed with brine, dried over anhydrous N~S04,

filtered, and concentrated in vacuo. The crude product was purified by

flash chromatography (Si02, 10% ethyl acetate/hexane) to give (Z)-3-

methyl-1-(2-tetrahydropyranyloxy)-2-penten-4-yne (9. 78 g, 98%) as a

28

yellow oil: 1H NMR (500 MHz, CDC13) 8 5.90 (m, 1H), 4.63 (dd, 1H, J

= 3.1, 4.1 Hz), 4.38 (ddq, 1H, J = 1.4, 6.2, 12.6 Hz), 4.22 (ddq, 1H, J = 1.0, 7.2, 12.5 Hz), 3.87 (m, 1H), 3.50 (m, 1H), 3.13 (s, 1H), 1.88 (q,

3H, J = 1.3 Hz), 1.81 (m, 1H), 1.70 (m, 1H), 1.60-1.49 (4H); IR

(CHC13) 3305, 3010, 2948, 2855, 1442, 1202, 1118, 1023 em-'; MS m/e

(%) 180 (M+) (0.8), 149 ( 4 ), 85 (1 00), 79 (57); Anal. Calcd for

C 1 1H 160 2: C, 73.30; H, 8.95. Found: C, 73.02; H, 8.95. To a solution of

ethylmagnesium bromide (0.47 mL, 3 Min ethyl ether, 1.41 mmol) in

THF (5 mL) was added a solution of the above acetylene compound

(0.25 g, 1.40 mmol) in THF (2 mL) at 0 OC and the mixture was stirred

at 50 OC for 1.5 h to give bromomagnesium salt 14. A solution of 13

(0.34 g, 1.44 mmol) in THF (2 mL) was added to the solution of 14 at

ambient temperature and the whole mixture was stirred for 1 h. The

reaction mixture was diluted with sat. aq. NH4Cl and extracted with

ethyl acetate. The organic phase was washed with brine, dried over

anhydrous N~S04, filtered, and concentrated in vacuo. The crude

product was purified by flash chromatography (Si02 , 15% ethyl

acetate/hexane) to give 15 (0.37 g, 64 %) as a colorless oil: 1H NMR

(500 MHz, CDC13) 8 7.58 (ddd, 1H, J = 2.0, 2.6, 7.6 Hz), 7.18 (m, lH),

6.96 (dt, 1H, J = 1.1, 7.5 Hz), 6.80 (d, 1H, J = 8.0 Hz), 5.84 (m, 2H),

4.65, 4.63 (tx2, total 1H, J = 3.6 Hz), 4.35-4.22 (2H), 3.84 (m, 1H),

3.47 (m, 1H), 3.03, 2.97 (dx2, total lH, J = 5.3 Hz), 1.88 (d, 3H, J = 0.8 Hz), 1.80 (m, 1H), 1.69 (m, lH), 1.59-1.46 (4H), 1.02 (s, 9H), 0.26

(m, 6H); IR (CHC13) 3430, 3011, 2953, 2860, 1480, 1454, 1258, 1021,

916, 840, 762 cm-1; MS m/e (%) 415 [(M-HtJ (2), 398 [(M-H20) +] (3),

314 (46), 257 (98), 85 (66), 75 (100); Anal. Calcd for C24H360 4Si: C,

69.19; H, 8.71. Found: C, 69.05; H, 8.64.

29

(Z)-1- [2-(t- B u tyldimethy lsily loxy) pheny I] -4- me thy 1-6- (2-

tetrahydropyranyloxy)-4-hexen-2-yn-1-one ( 16). To a solution

of 15 (0.92 g, 2.20 mmol) in dichloromethane (30 mL) was added

manganese dioxide(2.0 g) and the mixture was stirred for 3 h at ambient

temperature. The reaction mixture was diluted with ethyl ether,


flash chromatography (Si02, 15% ethyl acetate/hexane) to give 11 (0.89

g, 97%) as a yellow oil: 1H NMR (500 MHz, CDC13) 87.93 (dd, 1H, J =

1.7, 7.8 Hz), 7.38 (ddd, 1H, 1= 1.8, 7.3, 8.3 Hz), 7.00 (m, 1H), 6.87

(dd, 1H, J = 0.8, 8.3 Hz), 6.11 (m, 1H), 4.62 (dd, 1H, J = 3.1, 4.2 Hz),

4.43 (ddq, 1H, J = 1.3, 6.2, 12.9 Hz), 4.28 (ddd, 1H, J = 1.0, 7.3, 12.8

Hz), 3.83 (m, 1H), 3.46 (m, 1H), 1.97 (d, 3H, J = 1.2 Hz), 1.79 (m, 1H),

1.69 (m, 1H), 1.57-1.50 (4H), 0.99 (s, 9H), 0.21 (s, 6H); IR (CHC13)

3012, 2952, 2860, 2189, 1643, 1478, 1447, 1257, 1233, 1022, 914, 841,

772, 759, 748 cm-1; MS m!e (%) 357 [(M-13urJ (26), 273 (100), 235

(98), 85 (7 4 ).

(Z)-1-(2-Hydroxypheny 1)-4-methy 1-6- (2- tetrahydro

pyranyloxy)-4-hexen-2-yn-1-ol (17). To a solution of 15 (33.5

mg, 80.4 pmol) in THF (1 mL) was added tetrabutylammonium

fluoride (80 pL, 1.0 M in THF, 80.0 pmol) at 0 OC and the mixture was

stirred at ambient temperature for 20 min. The reaction mixture was

diluted with sat. aq. NH4Cl and extracted with ethyl acetate. The




mg, 95%) as a yellow oil: 1H NMR (400 MHz, CDC13) 8 7.31 (m, 1H),

7.21 (m, 1H), 6.90-6.85 (2H), 5.89 (m, 1H), 5.75 (d, 1H, J = 8.0 Hz),

4.72 (tx2, total lH, J = 3.2 Hz), 4.35-4.21 (2H), 3.86 (m, 1H), 3.51 (m,

30

1H), 1.90 (s, 3H), 1.84-1.46 (8H); IR (CHClJ 3356, 3015, 2949, 2927,

1487, 1234, 1021 cm- 1; MS m/e (%) 284 [(M-H20t] (7), 200 (97), 171

(38), 84 (91), 55 (100).

(Z)-1- (2-Hydroxypheny 1)-4-methyl-6- (2- tetrahydropyranyl

oxy)-4-hexen-2-yn-1-one (18). To a solution of 17 (11.9 mg, 39.4

pmol) in dichloromethane (1 mL) was added manganese dioxide (50.0

mg) and the mixture was stirred for 4 h at ambient temperature. The

reaction mixture was diluted with ethyl ether, filtered and concentrated

in vacuo. The crude product was purified by flash chromatography

(Si02, 25% ethyl acetate/hexane) to give 18 (10.2 mg, 86%) as a yellow

oil: 1H NMR (500 MHz, CDC13) 8 11.67 (s, lH), 8.00 (dd, 1H, J = 1.7,

8.0 Hz), 7.49 (ddd, 1H, J = 1.7, 7.1, 8.5 Hz), 6.97 (m, 1H), 6.93 (ddd,

1H, J = 1.1, 7.3, 8.0 Hz), 6.21 (m, 1H), 4.66 (dd, 1H, J = 3.1, 4.2 Hz),

4.48 (ddq, 1H, J = 1.3, 6.4, 13.0 Hz), 4.33 (ddd, 1H, J = 1.1, 7.3, 13.0

Hz), 3.87 (ddd, 1H, J = 3.1, 8.3, 11.4 Hz), 3.51 (m, 1H), 2.03 (q, 3H, J

= 1.3 Hz), 1.81 (m, 1H), 1.72 (m, 1H), 1.62-1.49 (4H); IR (CHC13)

2949, 2191, 1624, 1597, 1243, 1022 cm-1; MS m/e (%) 300 (M+) (0.2),

273 (3), 216 (13), 173 (31), 121 (44), 85 (100).

(Z)-2- [ 1-(2-Tetrahydropyrany loxy) bu ten-3-y l] -4H -ch rom en-

4-one (19). To a solution of 16 (20.3 mg, 49.0 pmol) and 18-crown-

6 (26.2 mg, 99.1 pmol) in N,N-dimethylformamide (1 mL) was added

spray dried potassium fluoride (5.7 mg, 98.1 pmol) at 0 OC and the

mixture was stirred at ambient temperature for 2 h. The reaction

mixture was diluted with sat. aq. NH4Cl and extracted with ethyl acetate.

The organic phase was washed with brine, dried over anhydrous

N~S04, filtered, and concentrated in vacuo. The crude product was

purified by flash chromatography (Si02, 15% ethyl acetate/hexane) to

31

give 19 (14.2 mg, 97 %) as a yellow oil: 'H NMR (500 MHz, CDC13) 8

8.17 (ddd, 1H, 1= 0.5, 1.7, 7.9 Hz), 7.65 (ddd, 1H, J = 1.7, 7.2, 8.4

Hz), 7.43 (ddd, 1H, J = 0.5, 1.1, 8.4 Hz), 7.38 (ddd, 1H, J = 1.0, 7.0,

8.0 Hz), 6.28 (s, 1H), 6.06 (m, 1H), 4.65 (m, 1H), 4.62 (ddq, 1H, J =

1.7, 5.7, 14.5 Hz), 4.41 (ddq, 1H, J = 1.4, 6.2, 14.3 Hz), 3.85 (m, 1H),

3.49 (m, 1H), 2.09 (q, 3H, J = 1.5 Hz), 1.81 (m, 1H), 1.72 (m, 1H),

1.60-1.51 (4H); 13C NMR (50 MHz, CDC13) 8 178.7, 164.1, 156.3,

135.4, 133.9, 129.2, 125.8, 125.3, 123.9, 118.1, 110.7, 98.7, 65.0,

62.4, 30.5, 25.2, 21.0, 19.3; IR (CHC13) 3013, 2949, 2873, 2855, 1649,

1642, 1567, 1444, 1383, 1212, 1132, 1024 em-'; MS m/e (%) 300 (M+)

(1), 216 (100) [(M-THP+Ht], 200 (77), 199 (75), 187 (63), 121 (83),

85 (79); HRMS calcd for C13H120 3 [(M-THP+Ht], 216.0787; found,

216.0806.

2-[2-Methyl-4-(2-tetrahydropyranyloxy)-2-butenylidenyl]

benzofuran-3-one (20). To a solution of 16 (14.4 mg, 34.7 pmol)

in tetrahydrofuran (1 mL) was added a tetrahydrofuran solution of

tetra-n-butylammonium fluoride (35 pL, 1 M in tetrahydrofuran, 35

pmol) at 0 OC and the mixture was stirred for 90 min at that

temperature. The reaction mixture was diluted with sat. aq. NH4Cl and



crude product was purified by flash chromatography (Si02, 25% ethyl

acetate/hexane) to give 20 (5.0 mg, 48%) as a colorless oil accompanied

with 19 (4.4 mg, 42%). 20: 'H NMR (400 MHz, CDC13) 8 7.75 (ddd,

1H, J = 0.6, 1.4, 7.7 Hz), 7.61 (ddd, 1H, J = 1.4, 7.3, 8.5 Hz), 7.21 (dt,

1H, J = 0.7, 8.3 Hz), 7.17 (dt, 1H, J = 0.7, 8.4 Hz), 6.83 (d, 1H, J = 0.9

Hz), 5.90 (m, lH), 4.66 (dd, 1H, J = 3.1, 4.3 Hz), 4.46 (ddq, 1H, J = 1.4, 6.4, 13.3 Hz), 4.26 (ddq, 1H, J = 1.1, 7.3, 13.3 Hz), 3.89 (m, 1H),

32

3.54 (m, 1H), 2.24 (q, 3H, J = 1.3 Hz), 1.82 (m, 1H), 1.72 (m, 1H),

1.64-1.48 (4H); 13C NMR (50 MHz, CDClJ 8 185.14, 166.28, 147.29,

136.99, 135.70, 132.46, 124.71, 123.36, 121.66, 112.96, 109.62, 98.38,

63.44, 62.31, 30.42, 25.23, 22.51, 19.29; IR (CHC13) 3020, 2947, 1717,

1606, 1462, 1300, 1129, 1031 em-'; MS m/e (%) 216 [(M-THP+Ht]

(23), 185 (37), 134 (100).

Internal quenching with DMF-CH30D. The reaction of 16 (33.3

mg, 0.08 mmol) with potassium fluoride (9 .3 mg, 0.15 mmol), 18-

crown-6 ( 44.6 mg, 0.17 mmol) in N,N-dimethylformamide (2 mL) and

methanol-d (0.5 mL, 99 atm% D) was carried out for 10 min at ambient

temperature and worked-up as usual. Integration of the signal at 6.83

ppm for 20 in 'H NMR showed the deuterium content of the produced

20 was 97%. More than 97o/o deuterium content for 19 was determined

by the disappearance of the olefine hydrogen in 1 H NMR.

Chromatographic separation afforded d1-19 (2.7 mg, 11 %) and d 1-20

(11.6 mg, 48o/o). d1-19: MS m/e (%) 301 (1) (M+), 217 (100) [(M

THP+Ht], 202 (79), 201 (80), 188 (56); HRMS calcd for C13H, 10 3D1

[(M-THP+Ht], 217.0849; found, 217.0769. d1-20: MS m/e (o/o) 301 (1)

(M+), 217 (28) [(M-THP+Ht], 186 (41); HRMS calcd for C13H11 0 3D1

[(M-TI-IP+Ht], 217.0849; found, 217.0779.

Time-course of the cyclization of 16. Four reactions of 16 under

the standard KF-18-crown-6-DMF conditions (see the procedure for

preparation of 19) were carried out at -20 °C, and three of four

reactions were quenched by adding sat. aq. NH4Cl after 10, 30, and 60

min at that temperature. The remaining reaction was warmed up to 0

OC after 60 min at -20 OC and quenched as previous. Each reaction

mixture was extracted as for the preparation of 19 to give a crude

33

mixture, which was analyzed by 1H NMR in CDC13• The result was

shown in Figure 4.

Cyclization of 16 in the presence of 21. In a NMR tube a

solution of DMF-d7 (1 mL) containing 16 (5.0 mg, 0.012 mmol), 21

(3.6 mg, 0.012 mmol), and 18-crown-6 (12.8 mg, 0.048 mmol) was

prepared and 'H NMR of the starting mixture was recorded. To the

solution was added potassium fluoride (2.8 mg, 0.048 mmol) at room

temperature and the mixture was sonicated for 10 min. 'H NMR

spectrum of the resulting dark brown solution was then recorded.

D 20 quenching before the equilibration is completed. The

reaction of 16 described for the time-course experiment was quenched

by adding D20-ND4Cl solution after 15 min at -20 °C. The resulting

mixture was worked-up as usual and 1H NMR spectrum of the crude

product was recorded in CDC13. The result was shown in Figure 5.

(E)-1- [2- (t-B u ty ldimethy lsily I oxy )pheny I] -4-me thy 1-6- (2-

tetrahydropyranyloxy)-4-hexen-2-yn-1-one (21). According to

the method described in the synthesis of 15 the reaction of (E)-3-

methyl-2-penten-4-yn-1-ol (5.29 g, 55.0 mmol) gave (E)-3-methyl-1-

(2-tetrahydropyranyloxy)-4-pentyn-2-ene (7 .87 g, 79%) as a colorless

oil: 'H NMR (400 MHz, CDC13) 8 6.04 (m, 1H), 4.61 (m, 1H), 4.26 (dd,

1H, 1 = 6.2, 13.2 Hz), 4.09 (dd, 1H, J = 7.2, 13.2 Hz), 3.84 (m, 1H),

3.50 (m, 1H), 2.80 (s, IH), 1.82 (s, 3H), 1.80-1.48 (m, 6H); IR (CHC13

)

3306,3012,2948,2855, 1442, 1200, 1119, 1024 em-'; MS m/e (%) 180

(M+), (4), 149 (11), 85 (100)~ Anal. Calcd for C11

H16

02

: C, 73.30~ H,

8.95. Found C, 73.45~ H, 9.01. This compound (1.24 g, 6.87 mmol)

34

was treated with ethylmagnesium bromide as described in the synthesis

of 15 to give the corresponding bromomagnesium salt, which was

reacted with 13 (1.36 g, 5.74 mmol) to afford (E)-1-[2-(t

butyldimethylsiloxy)phenyl]-4-methyl-6-(2-tetrahydropyranyloxy)-2-

hexyn-4-en-1-ol (1.34 g, 55% based on 13) as a colorless oil: 'H NMR

(400 MHz, CDC13) 87.56 (dd, 1H, 1 = 1.8, 7.7 Hz), 7.18 (ddd, lH, 1 = 1.8, 7.4, 8.0 Hz), 6.96 (dt, 1H, 1 = 1.1, 7.5 Hz), 6.81 (dd, 1H, 1 = 1.1,

8.1 Hz), 5.98 (m, 1H), 5.80 (d, 1H, 1 = 5.5 Hz), 4.60 (m, lH), 4.24 (dd,

1H, 1 = 6.2, 12.5 Hz), 4.10 (dd, 1H, 1 = 7.1, 13.2 Hz), 3.84 (m, 1H),

3.50 (m, 1H), 2.71 (d, 1H, 1 = 5.6 Hz), 1.83 (d, 3H, 1 = 1.5 Hz), 1.81-

1.49 (m, 6H), 1.02 (s, 9H), 0.29 and 0.26 (sx2, total 6H)~ IR (CHC13)

3592, 3015, 2954, 2860, 1488, 1454, 1258, 1023, 912, 840, 745 em- '~

MS mle (%) 359 [(M-'BurJ (3), 315 (12), 275 (19), 257 (35), 179

(100). Oxidation of this alcohol (1.16 g, 2.79 mmol) with manganese

dioxide (ca. 2 g) produced 21 (1.05 g, 91 %) as a yellow oil: 'H NMR

(400 MHz, CDC13) 87.90 (dd, 1H, 1 = 1.9, 7.8 Hz), 7.38 (ddd, 1H, 1 =

1.8, 7.2, 8.1 Hz), 7.00 (ddd, 1H, 1 = 1.1, 7.3, 7.8 Hz), 6.86 (dd, 1H, 1 =

1.1, 8.2 Hz), 6.29 (m, 1H), 4.62 (t, 1H, 1 = 3.4 Hz), 4.32 (ddq, 1H, 1 = 1.1, 6.1, 14.0 Hz), 4.15 (ddd, 1H, 1 = 0.9, 7.0, 13.9 Hz), 3.84 (ddd, 1H,

1 = 3.2, 8.3, 11.6 Hz), 3.51 (m, 1H), 1.90 (d, 3H, 1 = 1.2 Hz), 1.80 (m,

1H), 1.70 (m, 1H), 1.62-1.49 ( 4H), 0.99 (s, 9H), 0.21 (s, 6H); IR

(CHC13) 3016, 2953, 2860, 2190, 1642, 1478, 1448, 1256, 910, 840,

761 cm- 1 ~ MS m!e (o/o) 357 [(M-'BurJ (12), 273 (8), 245 (46), 203 (24),

179 (17), 149 (18), 85 (100).

( 4R *, 5R *)-1-[2-(t-Bu tyldimethylsilyloxy)phenyl]-4,5-epoxy-

6-(1-ethoxyethyloxy)-4-methyl-2-hexyn-1-one (22). To a

solution of (Z)-3-methyl-2-penten-4-yn-1-ol (2.51 g, 26.2 mmol) and

disodium hydrogenphosphate (9.06 g, 63.8 mmol) in dichloromethane

35

(100 mL) was added m-chloroperbenzoic acid (9.08 g, 52.6 mmol) at 0

OC and the mixture was stirred at ambient temperature for 20 h. The

mixture was diluted with sat. aq. N~S203 and sat. aq. NaHC03, and



crude product was purified by flash chromatography (Si02 , 30o/o ethyl

acetate/hexane) to give (2S*, 3R *)-2,3-epoxy-3-methyl-4-pentyn-1-ol

(2.45 g, 84%) as a colorless needle: 1H NMR (500 MHz, CDC13) 8 3.91

(dd, 1H, J = 4.6, 12.3 Hz), 3.82 (dd, 1H, J = 6.2, 12.4 Hz), 3.08 (dd,

lH, J = 4.7, 6.1 Hz), 2.38 (s, 1H), 1.76 (br, 1H), 1.57 (s, 3H); IR

(CHC13) 3427, 3305, 3015, 1440, 1377, 1090 cm-1; Anal. Calcd for

C6H80 2: C, 64.27; H, 7.19. Found C, 64.15; H, 7.10. To a solution of

this epoxide (0.45 g, 4.00 mmol) and ethyl vinyl ether (0.77 mL, 8.05

mmol) in dichloromethane ( 4 mL) was added a catalytic amount of

pyridinium p-toluenesulfonate at 0 OC and the mixture was stirred at

ambient temperature for 4 h. The reaction mixture was diluted with

sat. NaHC03 and extracted with ethyl acetate. The organic phase was

washed with brine, dried over anhydrous N~S04, filtered, and

concentrated in vacuo. The crude product was purified by flash

chromatography (Si02 , 15% ethyl acetate/hexane) to give (2R*, 3R*)-

2,3-epoxy-1-(l-ethoxyethoxy)-3-methyl-4-pentyne (0.65 g, 88%) as a

colorless oil: 1H NMR (400 MHz, CDC13) 8 4.76 (q, 1H, J = 5.4 Hz),

3.85-3.62 (m, 3H), 3.48 (m, 1H), 3.06 (t, 1H, J = 5.4 Hz), 2.35 and

2.36 (sx2, total lH), 1.55(s, 3H), 1.32 (dd, 3H, J = 1.0, 5.4 Hz), 1.19 (t,

3H, J = 7.1 Hz); IR (CHC13) 3272, 3260, 2980, 2934, 2878, 1134, 1060

em 1; MS mle (%) 169 [(M-MerJ, (5), 95 (13), 73 (100); Anal. Calcd

for C 10H 160 3 : C, 65.19; H, 8.75. Found C, 65.02; H, 8.92. To a solution

of this compound (0.92 g, 5.01 mmol) in tetrahydrofuran (5 mL) was

added n-butyllithium (3.10 mL, 1.62 Min hexane, 5.02 mmol) at -78

36

OC and the mixture was stirred at -78 OC for 15 min. After addition of

a solution of 13 (1.18 g, 5.01 mmol) in tetrahydrofuran (5 mL) was

added at -78 OC the mixture was stirred for 2 h. The mixture was

diluted with sat. aq. NH4Cl and extracted with ethyl acetate. The



flash chromatography (Si02 , 15-50o/o ethyl acetate/hexane) to give

( 4R *, 5R *)-1-[2-(t-butyldimethylsilyloxy )-phenyl]-4,5-epoxy-6-( 1-

ethoxyethoxy)-4-methyl-2-hexyn-1-ol (77o/o as a mixture of four

diastereomeric isomers) as a colorless oil: 1H NMR (400 MHz, CDC13) 8

7.51 (m, 1H), 7.18 (m, 1H), 6.95 (m, 1H), 6.80 (m, 1H), 5.72 (m, IH),

4.73 (m, 1H), 3.85-3.58 (3H), 3.45(m, 1H), 3.08 (m, 1H), 2.86-2.66

(1 H), 1.56 (m, 3H), 1.28 (m, 3H), 1.16 (m, 3H), 1.01 (s, 9H), 0.27 (m,

6H); IR (CHC13) 2955, 1488, 1259 cm-1; MS mle (%) 374 (2), 331 (9),

307 (25), 273 (100), 243 (88), 179 (100); Anal. Calcd for C23H360 5Si:

C, 65.68; H, 8.63. Found C, 65.77; H, 8.73. Oxidation of this alcohol

(41.0 mg, 97.5 pmol) with manganese dioxide gave 22 (30.7 mg, 75%)

as a yellow oil: 1H NMR (400 MHz, CDC13) 87.90 (dd, 1H, J = 1.8, 7.9

Hz), 7.40 (ddd, 1H, J = 1.8, 7.3, 8.3 Hz), 7.00 (m, 1H), 6.87 (ddd, 1H, J

= 0.4, 1.1, 8.3 Hz), 4.74 (q, 1H, J = 5.3 Hz), 3.88-3.59 (3H), 3.46 (rn,

1H), 3.19 (m, 1H), 1.64 (s, 3H), 1.30 (d, 3H, J = 5.3 Hz), 1.15 and 1.14

(tx2, total 3H, J = 7.1 Hz), 1.00 (s, 9H), 0.22 (s, 6H); IR (CHC13) 2933,

1650, 1478, 1256 cm-1; MS m/e (%) 418 (M+) (0.3), 289 (41), 235 (48),

152 (72), 143 (87), 121 (100).

( 4S*, SR *)-1-[2-(t-Bu tyldimethylsilyloxy)phenyl]-4,5-epoxy-

6-(1-ethoxyethoxy)-4-methyl-2-hexyn-1-one (23). According

to the method described in the synthesis of 22 the reaction of (E)-3-

methyl-2-penten-4-yn-1-ol (2.51 g, 26.1 mmol) gave (2R *, 3R *)-2,3-

37

epoxy-3-methyl-4-pentyn-1-ol (2.16 g, 74%) as a colorless oil: 1H NMR

(400 MHz, CDC13) 8 3.83 (dd, 1H, J = 4.4, 12.4 Hz), 3.69 (dd, 1H, J =

6.2, 12.4 Hz), 3.36 (dd, 1H, J = 4.5, 6.2 Hz), 2.31 (s, 1H), 1.78 (br,

1 H), 1.54 (s, 3H); IR (CHC13) 3605, 3452, 3306, 3017, 1219, 1027, 733

cm-1; Anal. Calcd for C6H80 2: C, 63.98; H, 7.19. Found C, 64.27; H,

7.17. The alcohol (1.70 g, 15.1 mmol) was protected with ethoxyethyl

group to give (2S*, 3R*)-2,3-epoxy-1-(1-ethoxyethoxy)-3-methyl-4-

pentyne (2.46 g, 88%) as a colorless oil: 1H NMR (400 MHz, CDC13) 8

4.75 (q, 1H, J = 5.4 Hz), 3.68-3.44 (4H), 3.34 (t, 1H, J = 5.4 Hz), 2.29

(s, 1H), 1.51 (s, 3H), 1.312 and 1.309 (dx2, total 3H, J = 5.4 Hz), 1.19

(t, 3H, J = 7.1 Hz); IR (CHC13) 3306, 3015, 1384, 1229, 1133, 1083 cm-

1; Anal. Calcd for C10H160 3: C, 65.19; H, 8.75. Found C, 64.94; H, 8.66.

Lithiation of this acetylenic compound (0.89 g, 4.85 mmol) followed by

the reaction with 13 (1.14 g, 4.83 mmol) produced (4S*, SR*)-1-[2-(t

butyldimethylsilyloxy)phenyl]-4,5-epoxy-6-(1-ethoxy-ethoxy)-4-methyl-

2-hexyn-1-ol (1.41 g, 69% as a mixture of four diastereomeric isomers)

as a colorless oil: 1H NMR ( 400 MHz, CDC13) 8 7.49 (m, 1H), 7.18 (m,

1H), 6.95 (m, 1H), 6.80 (m, 1H), 5.70 (m, 1H), 4.74 (m, 1H), 3.70-3.45

(5H), 3.34 (m, 1H), 2.79 (m, 1H), 1.52 (m, 3H), 1.31 (m, 3H), 1.18 (m,

3H), 1.01 (s, 9H), 0.27 (m, 6H); IR (CHC13) 3020, 2933, 1488. 1258,

919 em-'; MS mle (o/o) 405 [(M-MerJ (0.3), 375 (2), 363 (5), 331 (9),

317 (14), 273 (85), 243 (100), 179 (100). This alcohol (0.97 g, 2.31

mmol) was oxidized to give 23 (0.73 g, 76o/o) as a colorless oil: 1H

NMR (400 MHz, CDCIJ 8 7.87 (dd, 1H, J = 1.8, 7.8 Hz), 7.39 (ddd,

lH, J = 1.8, 7.3, 8.2 Hz), 7.00 (m, 1H), 6.86 (dd, 1H, J = 0.9, 8.3 Hz),

4.75 and 4.74 (qx2, total1H, J = 5.3 Hz), 3.71-3.57 (3H), 3.48 (m, 1H),

3.45 (t, 1H, J = 5.4 Hz), 1.60 (s, 3H), 1.31 (dx2, total 3H, J = 5.4 Hz),

1.19 (t, 3H, J = 7.0 Hz), 0.99 (s, 9H), 0.21 (s, 6 H); IR (CHC13) 2933,

38

-

2210, 1649, 1479, 1254, 751 cm-1; MS mle (o/o) 403 [(M-Met] (2), 361

(67), 289 (31), 259 (25), 201 (100), 179 (54).

1- [2-(t-B u ty ldimethy lsily loxy )phenyl] -4- (2-tetrahydro

pyranyloxy)-2-butyn-1-one (24). To a solution of 1-(2-

tetrahydropyranyloxy)-2-propyne (0.33 g, 2.35 mmol) in THF (15 mL)

was added n-butyl lithium ( 1.5 mL, 1.62 M in hexane, 2.43 mmol) at -

78 OC and the mixture was stirred at -78 OC for 15 min. To the

resulting alkynyllithium solution, tetrahydrofuran ( 4 mL) solution of

13 (0.55 g, 2.35 mmol) was added at -78 OC and the mixture was

stirred for 2 h. The reaction mixture was quenched with sat. aq. NH4Cl

and extracted with ethyl acetate. The organic phase was washed with

brine, dried over anhydrous N ~SO 4 , filtered, and concentrated in

vacuo. The crude product was purified by flash chromatography (Si02 ,

15-50% ethyl acetate/hexane) to give 1-[2-(t-butyldimethylsilyloxy)

phenyl]-4-(2-tetrahydropyranyloxy)-2-butyn-1-ol (0.60 g, 68%) as a

colorless oil: 1H NMR (400 MHz, CDC13) 8 7.54 (dd, 1H, J = 1.6, 7.5

Hz), 7.18 (dt, 1H, J = 1.8, 7.8 Hz), 6.96 (dt, 1H, J = 1.1, 7.5 Hz), 6.81

(m, 1H), 5.73 (m, 1H), 4.79 (t, 1H, J = 3.3 Hz), 4.38-4.25 (2H), 3.81

(m, 1H), 3.49 (m, 1H), 2.73 (d, 1H, J = 5.8 Hz), 1.85-1.46 (m, 6H),

1.01 (s, 9H), 0.28 and 0.26 (sx2, total 3H); IR (CHC13) 3010, 2951,

2933, 2860, 1480, 1258, 1221, 1025, 919, 840, 753 cm-1; MS m/e (%)

275 (2), 236 (9), 217 (13), 179 (39), 85 (100); Anal. Calcd. for

C21 H320 4Si: C, 66.98; H, 8.57. Found C, 66.71; H, 8.71. Oxidation of

this alcohol (0.42 g, 1.12 mmol) with manganese dioxide produced 2 4

(0.29 g, 69%) as a yellow oil: 1H NMR ( 400 MHz, CDCI 3) 8 7.90 ( dd,

1H, J = 1.9, 7.9 Hz), 7.38 (ddd, 1H, J = 1.9, 7.3, 9.1 Hz), 7.00 (m, IH),

6.86 (dd, lH, J = 1.0, 8.4 Hz), 4.83 (t, IH, J = 3.2 Hz), 4.45, 4.47 (dx2,

each IH, J = 17 Hz), 3.83 (m, 1H), 3.53 (m, 1H), 1.85-1.48 (6H), 0.99

39

(s, 9H), 0.22 (s, 6H); IR (CHC13) 3011, 2952, 2932, 2859, 1649, 1479,

1234, 1028, 755 cm-1; MS m/e (o/o) 317 [(M-1Buf] (55), 233 (32), 217

(100), 189 (100), 85 (83); Anal. Calcd. for C21 H300 4Si: C, 67.34; H,

8.07. Found: C, 67.10; H, 7.96.

(E)-2- [1-(2-Tetrahydropyranyloxy) buten-3-yl]-4H -chromen-

4-one (25). The same procedure described for the synthesis of 19

was applied for 21 (25.5 mg, 61.5 ,umol) to afford 25 (15.3 mg, 83o/o)

as a yellow oil: 1H NMR (400 MHz, CDC13) 8 8.16 (dd, 1H, J = 1.5, 7.9

Hz), 7.65 (ddd, 1H, J = 1.7, 7.2, 8.6 Hz), 7.47 (dd, 1H, J = 0.7, 8.5 Hz),

7.36 (ddd, lH, J = 1.0, 7 .0, 8.0 Hz), 6.82 (m, 1H), 6.39 (s, 1H), 4.69

(m, lH), 4.53 (m, 1H), 4.29 (m, 1H), 3.89 (m, lH), 3.56 (m, 1H), 1.98

(d, 3H, J = 1.1 Hz), 1.90-1.51(6H); IR (CHC13) 3011, 2948, 2873, 2857,

1640, 1467, 1377, 1212, 1124, 1026, 775 cm-1; MS m/e (%) 216 [(M

THP+H) +] (79), 199 (58), 187 (88), 121 (66), 85 (100); Anal. Calcd for

C18

H20

04

: C, 71.98; H, 6.71. Found: C, 72.01; H, 6.66; HRMS calcd for

C13H

200

4 [(M-THP+H) +], 216.0787; found, 216.0804.

2- [(2R *, 3R *)-2,3-Epoxy-1-(1-ethoxyethyloxy)butan-3-yl]-

4H -chromen-4-one (26). The same procedure described for the

synthesis of 19 was applied for 22 (11.7 mg, 28.0 ,umol) to afford 26

(6.2 mg, 73o/o) as a yellow oil: 1H NMR (400 MHz, CDC13) 8 8.19 (dd,

lH, J = 1.6, 8.0 Hz), 7.66 (dddd, 1H, J = 0.3, 1.7, 7.2, 8.9 Hz), 7.45 (m,

lH), 7.40 (ddd, lH, J = 1.0, 7.2, 8.0 Hz), 6.380 and 6.379 (sx2, total

IH), 4.63 and 4.60 (qx2, total 2H, J = 5.3 Hz), 3.63-3.24 (5H), 1.73 (s,

3H), 1.21, 1.19 (dx2, total 3H, J = 5.4 Hz), 1.02, 1.01 (tx2, total 3H, J =

7.1 Hz); IR (CHC13) 3025, 1650, 1466, 1388, 1129 cm-1; MS m/e (%)

259 [(M-OEtr] (13), 232 (28), 162 (97), 73 (100).

40

2- [ (2S*, 3R *)-2,3-epoxy-1-(1-ethoxyethyloxy) butane-3-yl]-

4H -chromen-4-one (27). The same procedure described for the

synthesis of 19 was applied for 23 (53.5 mg, 0.13 mmol) to afford 2 7

(28.0 mg, 72o/o) as a yellow oil: 1H NMR (400 MHz, CDC13) 8 8.15 (dd,

1H, J = 1.6, 8.0 Hz), 7.65 (ddd, 1H, J = 1.7, 7.2, 8.7 Hz), 7.43 (m, 1H),

7.38 (m, 1H), 6.41 (s, 1H), 4.79, 4.78 (qx2, total 1H, J = 5.4 Hz), 3.97-

3.62 (3H), 3.50 (m, 1H), 3.37 (dd, 1H, J = 5.1, 5.5 Hz), 1.71 and 1.68

(sx2, total 3H), 1.34 (d, 3H, J = 5.4 Hz), 1.19 (t, 3H, J = 7.1 Hz); IR

(CHC13) 3011, 2990, 1609, 1466, 1383, 1131 cm-1; MS m/e (%) 259

[(M-OEtr] (5), 232 (16), 214 (100), 189 (54), 171 (50).

2- [ (2-tetrahydropyranyloxy )methy I] -4H -chromen-4-one ( 2 8).

The same procedure described for the synthesis of 19 was applied for

24 (40.7 mg, 0.11 mmol) to afford 28 (20.6 mg, 73o/o) as a yellow oil:

1H NMR (400 MHz, CDC13) 88.17 (dd, 1H, 1= 1.7, 7.9 Hz), 7.65 (ddd,

1H, J = 1.7, 7.2, 8.7 Hz), 7.42 (dd, 1H, J = 0.6, 8.5 Hz), 7.38 (ddd, 1H,

J = 0.8, 7.1, 8.1 Hz), 6.46 (t, 1H, J = 0.9 Hz), 4.78 (t, 1H, J = 3.3 Hz),

4.62 (dd, 1H, J = 1.0, 15.0 Hz), 4.44 (dd, 1H, J = 0.8, 14.9 Hz), 3.85

(m, 1H), 3.55 (m, 1H), 1.90-1.51 (6H); 13C NMR (50 MHz, CDC13) 8

178.5, 165.9, 156.5, 133.8, 125.9, 125.2, 124.2, 118.1, 109.4, 98.3,

64.6, 62.0, 30.0, 25.1, 18.6; IR (CHC13) 3020, 3008, 2949, 2885, 1651,

1467, 1215, 1122, 1036, 745 cm-1; MS mle (%) 205 (6), 176 [(M

THP+HrJ (37), 160 (100), 85 (79); Anal. Calcd. for C 15H 160 4: C, 69.22;

H, 6.20. Found: C, 69.09; H, 6.19; HRMS calcd for C10H80 3 [(M

THP+Hf], 176.0474; found, 176.0485. Hydrolysis of 28 (10.5 mg,

40.3 ,umol) in refluxing acetone (0.6 mL) and water (0.2 mL)

containing a catalytic amount of pyridinium p-toluenesulfonate gave 2-

hydroxymethyl-4H-chromen-4-one (6.0 mg, 84o/o) as a white powder: 1H NMR (200 MHz, DMSO-d6) 8 8.02 (dd, 1H, J = 1.8, 8.2 Hz), 7.79

41

(ddd , lH, J = 1.7, 7.0, 8.5 Hz) , 7.60 (dd, 1H, J = 1.0, 8.5 Hz), 7.47

(ddd, 1H, J = 1.1 , 7.1, 8.1 Hz), 6.33 (s, 1H), 5.79 (br, 1H), 4.44 (s, 2H); 13C NMR (50 MHz, DMSO-d6) 8 177.14. 170.09, 155.97. 134.40,

125.54, 125 .13. 123.62. 118.38 107.44. 59.81. These data were

identical with those reported.24

1- [2-( 1-t-B u tyldimethylsiloxy )na ph thy I] -4- (2-tetrahydro

pyranyloxy)-2-butyn-1-one (29). To a solution of

ethylmagnesium bromide (3 M in ethyl ether, 0.26 ml, 0.78 mmol) in

THF (4 mL) was added a solution of 1-(2-tetrahydropyranyloxy)-2-

propyne (98.8 mg, 0.70 mmol) in THF (1 mL) at 0 OC and the mixture

was stirred at 50 OC for 1 h. After the reaction mixture was cooled to

ambient temperature, a solution of 2-(1-t-butyldimethylsilyloxy)

naphthaldehyde (0.22 g, 0.77 mmol) in THF (1 mL) was added and the

whole mixture was stirred at ambient temperature for 20 min. The

mixture was diluted with sat. aq. NH4Cl and extracted with ethyl acetate.

The organic phase was washed with brine, dried over anhydrous

N~S04 , filtered, and concentrated in vacuo. The crude product was

purified by flash chromatography (Si02 , 15o/o ethyl acetate/hexane) to

give 1-[2-(1-t-butyldimethylsilyloxy)naphthyl]-4-(2-tetrahydropyranyl

oxy)-2-butyn-1-ol (162 mg, 54o/o) as a mixture of diastereoisomer as a

yellow oil: 1H NMR (400 MHz, CDC13) 8 8.07 (lH), 7.77 (lH), 7.73 (d,

lH, J = 8.6 Hz), 7.54 (d, 1H, J = 8.6 Hz), 7.47-7.41 (2H), 6.00 (t, 1H, J

= 1.7 Hz), 4.80 (t, 1H, J = 3.4 Hz), 4.38-4.28 (2H), 3.81 and 3.72

(mx2, total 1H), 3.50 (m, 1H), 2.20 (br, 1H), 1.85-1.47 (6H), 1.13 and

0.90 (sx2, total 9H), 0.19 and 0.08 (sx2, total 6H); IR (CHC13) 3020,

2953, 2860, 1374, 1260, 1213, 1088, 1024, 899, 841, 829, 785 cm- 1;

MS m/e (o/o) 426 (M+) (23), 285 (44), 267 (96), 193 (83), 85 (74), 69

(100); HRMS calcd for C25H340 4Si (M+) 426.2226; found, 426.2207. To

42

a solution of this alcohol (63.6 mg, 0.15 mmol) in dichloromethane was

added manganese dioxide( 100.0 mg) and the mixture was stirred for 36

h at ambient temperature. The reaction mixture was diluted with ethyl

ether, filtered and concentrated in vacuo. The crude product was

purified by flash chromatography (Si02, 15% ethyl acetate/hexane) to

give 29 (46.8 mg, 73%) as a yellow oil: 1H NMR (400 MHz, CDC13) 8 8.24 (m, 1H), 7.90 (d, 1H, J = 8.8 Hz), 7.78 (m, 1H), 7.56 (ddd, 1H, J =

1.3, 6.9, 8.1 Hz), 7.49 (dd, 1H, J = 1.5, 8.4 Hz), 7.48 (m, 1H), 4.85 (t,

1H, J = 3.2 Hz), 4.50 (s, 2H), 3.84 (ddd, 1H, J = 3.2, 9.3, 12.1 Hz), 3.53

(m, 1H), 1.84-1.50 (m, 6H), 1.12 (s, 9H), 0.10 (sx2, 6H, J = 0.7 Hz); IR

(CHC13) 3021, 2952, 2932, 1649, 1619, 1395, 1231, 1122, 1027, 903,

827, 725 cm-1; MS mle (%) 409 [(M-Met] (7), 367 (99), 283 (97), 267

(99), 239 (100); HRMS calcd for C24H290 4Si [(M-Met] 409.1835;

found, 409.1826.

2-[ (2-tetrahydropyranyloxy)methyl]-4H -naphtho[1,2-b ]pyran

(30). To a solution of 29 (89.5 mg, 0.21 mmol) and 18-crown-6 (137

mg, 0.52 mmol) in N,N-dimethylformamide (3 mL) was added a

potassium fluoride (24.6 mg, 0.42 mmol) at 0 OC and the mixture was

stirred at ambient temperature for 4 h. The reaction mixture was

diluted with sat. aq. NH4Cl, and extracted with ethyl acetate. The




mg, 81 %) as a yellow oil: 1H NMR (400 MHz, CDC13) 8 8.45 (m, 1H),

8.13 (d, lH, J = 8.6 Hz), 7.92 (m, lH), 7.76 (d, lH, J = 8.6 Hz), 7.69

(m, lH), 7.65 (m, 1H), 6.62 (s, 1H), 4.85 (t, 1H, J = 3.2 Hz), 4.78 (dd,

1H, J = 0.8, 14.7 Hz), 4.61 (d, 1H, J = 14.7 Hz), 3.90 (ddd, 1H, J = 3.1,

9.3, 12.4 Hz), 3.59 (m, 1H), 1.91-1.53 (6H); IR (CHC13) 3021, 2358,

43

1652, 1212, 774 cm-1; MS m/e (o/o) 310 (M+) (100), 254 (38), 226 (61),

210 (98), 181 (85); HRMS calcd for C19H180 4 (M+) 310.1203; found,

310.1183.

44

References and Notes


Antibiot. 1990,43, 1513-1518. (b) Yoshida, M.; Hara, M.; Saitoh, Y.; Sano, H. J. Antibiot. 1990,43, 1519-1523.

(2) For a review of the antibiotics of the pluramycin group, see: Sequin,

U. Fortschr. Chern. Naturst. 1986, 50, 57-122.

(3) (a) Kondo, S.; Miyamoto, M.; Naganawa, H.; Takeuchi, T.;

Umezawa, H. J. Antibiot. 1977, 30, 1143-1145. (b) Maeda, K.; Takeuchi, T.; Nitta, K.; Yagishita, K.; Utahara, R.; Osato, T.; Ueda, M.;

Kondo, S.; Okami, Y.; Umesawa, H. J. Antibiot., Ser. A 1956, 9, 75-81.

(4) (a) Sequin, U.; Bedford, C. T.; Chung, S. K.; Scott, A. I. Helv.

Chim. Acta 1977, 60, 896-906. (b) Zehnder, M.; Sequin, U.; Nadig, H. Helv. Chim. Acta 1979, 62, 2525-2533.

(5) Bennet, G. N. Nucleic Acids Res. 1982, 10, 4581-4594.

(6) Hara, M.; Yoshida, M.; Nakano, H. Biochemistry 1990, 29, 10449-10455.

(7) Chan, K.L.; Sugiyama, H.; Saito, I. Tetrahedron Lett. 1991, 52,

7719-7722.

(8) Hansen, M.; Hurley, L. J. Am. Chern. Soc. 1995, 117, 2421-2429. (9) (a) Hauser, F. M.; Rhee, R. P. J. Am. Chern. Soc. 1979, 1 OJ, 1628-1629. (b) Hauser, F. M.; Rhee, R. P. J. Org. Chern. 1980, 45, 3061-3068.

(1 0) For other synthetic studies for pluramycin antibiotics, see: (a)

Parker, K. A.; Koh, Y.-H. J. Am. Chern. Soc. 1994, 116, 11149-11150. (b) Dubois, E.; Beau, J.-M. J. Chern. Soc., Chern. Commun.

1990, 1191-1192. (c) Dubois, E.; Beau, J.-M. Carbohydr. Res. 1992, 228, 103-120.

45

(11) Hirao, I.; Yamaguchi, M.; Hamada, M. Synthesis 1984, 1076-

1078.

(12) For a general synthesis of 4H-chromen-4-ones, see: Hepworth, J.

D. Pyrans and Fused Pyrans: (iii) Synthesis and Application in

Comprehensive Heterocvclic Chemzstry, Katritzky, A. R., Rees, C. W.

Ed: Vol. 3, Pergamon Press. Oxford. 1984, pp 737-883.

(13) (a) Baldwin, J. K. J. Chern. Soc., Chern. Commun. 1976, 734-736.

(b) Baldwin, J. K.: Cutting, J.; Dupont, W.; Kruse, L.; Silberman, L.;

Thomas, R. C. J. Chern. Soc., Chern. Commun. 1976, 736-738.

(14) Garcia H.; Iborra, S.; Primo, J.; Miranda, M. A. J. Org. Chern.

1986, 51' 4432-4436.

( 15) For recent studies of a 6-endo-digonal addition to an unactivated

carbon-carbon triple bond, see: Padwa, A.; Krumpe, K. K.; Weingaten,

M. D. J. Org. Chern. 1995, 60, 5595-5603.

(16) For a related approach involving Pd-catalyzed carbonylation of o

iodophenol with terminal alkynes giving chromenoes, see Torii, S.;

Okumoto, H.; Sadakane, M.; Shostakovsky, M. V.; Ponomaryov, A. B.;

Kalinin, V. N. Tetrahedron 1993, 31, 6773-6784.

( 17) A preliminary account describing portions of this work has been

published. Nakatani, K.; Okamoto, A.; Yamanuki, M.; Saito, I. J. Org.

Chern. 1994, 59, 4360-4361.

( 18) All semiempilical and ab initio calculations were carried out using

Spartan molecular modeling software (version 3.1).

( 19) For related ab initio studies of the nucleophilic addition to the

acetylenic bonds, see: (a) Eisenstein, 0.; Procter, G.; Dunitz, J. D. Helv.

Chim. Acta 1978, 61, 2538-2541. (b) Dykstra, C. E.; Arduengo, A. J.;

Fukunaga, T. J. Am. Soc. Chern. 1978, 100, 6007-6012. (c) Strozier,

R. W.; Caramella, P.; Houk, K. N. J. Am. Soc. Chern. 1979, 101,

46

·-

1340-1343. (d) Houk, K. N.; Strozeiner, R. W.; Rozeboom, M. D.;

Nagase, S. J. Am. Soc. Chern. 1982, 104, 323-325.

(20) The energy gap between the two conformers calculated at the 3-

21 G(*) level became large ( 4. 72 kcal/mol), with the s-trans conformer

being more stable.

(21) For examples of syn addition, see: Bailey, W. F.; Ovaska, T. V. J.

Am. Chern. Soc. 1993, 115, 3080-3090 and references cited therein.

(22) Analyses of the vibrational frequencies were carried out using

Gaussian 92 program. Frisch, M. J.; Trucks, G. W.; Head-Gordon, M.;

Gill, P. M. W.; Wong, M. W.; Foresman, J. B.; Johnson, B. G.;

Schlegel, H. B.; Robb, M.A.; Replogle, E. S.; Gomperts, R.; Andres, J.

L.; Raghavachari, K.; Binkley, J. S.; Gonzalez, C.; Martin, R. L.; Fox,

D. J.; Defrees, D. J.; Baker, J.; Stewart, J. J. P.; Pople, J. A. Gaussian,

Inc. Pittsburgh P A, 1992.

(23) The interconversion was observed for chalcogenachromanones

upon the direct lithiation at the C2 position. Detty, M. Y.; McGarry, L.

W. J. Org. Chern. 1988, 53, 1203-1207.

(24) Payard, M.; Couquelet, J. Synthesis, 1979, 889.

47

-

CHAPTER 2

Essential Structure for Efficient DNA Alkylation by Kapurimycin A3

Abstract: An antibiotic kapurimycin A3 (1) is well-known to

efficiently alkylate DNA guanine bases. In order to know the origin of

effective guanine alkylation, we have investigated the relationship

between the structure of 1 and the activity for DNA guanine alkylation

by 1. One of the kapurimycin A3 analogs, ABC-ring analog 2, was

prepared through the coupling of BC-ring fragment with optically

active side chain fragment synthesized by the use of Sharpless

asymmetric dihydroxylation followed by 6-endo-dig selective

cyclization of o-silyloxyphenyl ethynyl ketone and intramolecular

Mitsunobu reaction for epoxide formation. Based on the results of

DNA cleavage by ABC-ring analog 2, diol 15, AB-ring analog 16 and

epoxyalcohol 17, we found that 2 has a minimum structure required for

effective DNA alkylation. Besides, the reactivity of 2 toward DNA was

about one tenth of that for 1. The enzymatic digestion of DNA

modified by 2 showed that DNA modification by 2 selectively occurred

at guanine bases. We also isolated and identified the 2-guanine adduct

by NMR and MS spectroscopy. In the cleaving assay of 32P-5' -end

labeled DNA, ABC-ring analog 2 showed a very similar sequence

selectivity to that for natural kapurimycin A3 (G*G > G* A > G*T >

G*C, G* denotes alkylated guanine).

49

Introduction

Kapurimycin A3

(1) is an antitumor antibiotic isolated by Kyowa

Hakko Kogyo group in 1990.1'2 Kapurimycin A3 consists of a

tetrahydroanthrapyranone ring and a vinyl epoxide side chain attached

to C2-position of the pyranone ring system. By our group and others, a

guanine-kapurimycin A3 adduct, obtained from the reaction of DNA

with 1, was isolated and identified, and it was shown kapurimycin A3

(1 ) alkylates guanine base in DNA duplex.3

Antibiotics which modify guanine base in DNA duplex are typically

represented by aflatoxin B 1 oxide4 and altromycin B. 5 In several studies

on these DNA alkylating antibiotics, the binding structure of DNA-drug

complex was investigated by NMR spectroscopy. All of these antibiotics

intercalate into DNA in their aromatic rings and subsequently generate

guanine-adduct with nucleophilic attack of guanine N7 to the reactive

epoxide group of the drugs. Kapurimycin A3 (1) has an aromatic ring

system and an epoxy subunit like other drugs and alkylates guanine base.

We were interested in the structure of DNA natural alkylating agents

and investigated a minimal structure required for efficient alkylation of

guanine base. In this study, we focused our attention on kapurimycin A3

(1) and have synthesized a diverse type of kapurimycin analogs. Herein

we report that ABC ring analog 2 effectively alkylates DNA guanine

base with a very similar sequence selectivity to that of natural

kapurimycin A., (1).

0 C02H

AcO kapurimycin A3 (1 )

50

(S, S)-ABC (2)

--


The synthetic route for ABC-ring analog 2 is outlined in Scheme 1.

Asymmetric dihydroxylation of p-methoxybenzyl ether 4 obtained from

a commercially available alcohol 3 with AD-mix-,lf' produced diol 5 ,

which was converted to alcohol 8. Oxidation of 8 and Wittig

olefination provided the alkene 10 (Z:E = 5: 1) 7 which was then

desilylated to 11. Addition of 1-tert-butyldimethylsilyloxy-2-

naphthaldehyde to lithiated 11 provided the corresponding coupling

product 12, which was subsequently oxidized with manganese dioxide to

ketone 13. Selective formation of pyranone ring was successfully

achieved by treating 13 with KF and 18-crown-6 in anhydrous DMF as

reported earlier8 to give tricyclic compound 14. Hydrolysis of the

acetal produced diol 15, which was subjected to intramolecular

Mitsunobu reaction9 to furnish (S, S)-2. Other kapurimycin analogs 16

and 17 were synthesized via a similar route.

We determined the absolute configuration of diol 15 by analysis of

NOESY spectrum of 14 and NMR analysis by the modified Mosher's

method. 10 In NOESY spectrum of 14, we observed the correlation

between H12 and H14. Besides, it was determined that the chirality of

C13 was (R), using MTPA esterification of diol15 and NMR analysis of

MTPA esters by the modified Mosher's method. Thus, the absolute

configuration of diol 15 is (llS, 13R).

51

Scheme 1.0

iOR1

~

,..3· R1 =H '1' 4;R1 =PMB

Yo 0 ~

OH HO ~ OPMB

I 5

k ____....

14

j ____....

15 2

a Reagents and Conditions: a) p-methoxybenzyl chloride, 94%; b) AD-mix-/3, 76% (91 % e.e.); c) Me

2C(0Meh, 95%; d) LHMDS then TMSCl, 91 %; e) DDQ, 99% f) Dess-Martin Periodinane, 94%; g) Ph3P=CHCH3, 78% (Z/£=1011); h) NaOMe, 80%; i) LHMDS, CeC13 then 1-t-butyldimethylsilyloxy-2-naphthaldehyde, 47%; j) Mn02, 76%; k) KF, 18-crown-6, DMF, 80%; 1) HCl, 98%; m) DEAD, PPh3, 76%.

OH

16 17

52

Next, we determined the absolute configuration of ABC analog 2 by

use of solvolysis of 2 and comparison of the structural data of solvolysis

product with those of 15. Hydrolysis of 2 in Tris-HCI buffer (pH 7 .6)

gave 11,13-dihydroxy compound (53%) and 11,15-dihydroxy

compound (7%) (eq. 1). It was indicated from the observation of 1H

NMR spectrum and optical rotation that the given 11, 13-dihydroxy

compoundwasidentifiedwith 15 (15: [af50 =-197.3, c 0.15, MeOH;

the hydrolysis product: [a] 250 = -135.7, c 0.07, MeOH). Methanolysis

of 2 also gave 11-hydroxy-13-methoxy compound (50o/o) and 11-

hydroxy-15-methoxy compound (15%) which relative structure was

confirmed by HSQC, HMBC and MS (eq. 2). This result shows that

solvolysis of the epoxide occurs at C13. Furthermore, in the NOESY

spectrum of 2 the correlation between H12 and H13 was observed. As

clear from above, the absolute configuration of ABC analog 2 is (liS,

13S). Intramolecular Mitsunobu reaction from (liS, 13R)-diol 15

served (liS, 13S)-ABC analog 2 through SN2 inversion at C13 and

hydrolysis of 2 mainly generated again (liS, 13R)-diol 15 through SN2

inversion at C 13.

0-MTPA (b) HO W

- 22

0

(Hz) aromatic protons : - 14- +7

.1[<X:15-(S)-MTPA) - <X:15-(R)-MTPA)]

Figure 1. NMR experiment ( 400 MHz) for structural identification. (a) NOESY observation of compound 14; (b) modified Mosher's method (The numbers in figure are the difference of chemical shift between (S)-MTPA ester and (R)MTPA ester of 15); (c) NOESY observation of ABC-ring analog 2.

53

2 (dl form)

l1 #IS - 14

2 (d/-forrn)

buffer

(pH 7.6)

Tris-HCI salt

MeOH

15 (53%)

19(50%)

OH

(1)

18(7%)

OMe

(2)

20 (15%)

The DNA cleavage activities of 2, 15, 16 and 17 were demonstrated

by relaxation assay of pBR322 supercoiled DNA cleaving assay (Figure

2). We investigated 10 pM of each drug with 40 pM of DNA in Tris

HCI (pH 7 .6) at 37 oc for 5 h. After ethanol precipitation for removal

of unreacted and hydrolyzed drugs, a half of the recovered DNA was

incubated again in water at 37 oc for 24 h. The different forms of

DNA were separated on agarose gel. Only the incubation with 2 clearly

converted supercoiled DNA (form I) into nicked-circular (form II) and

linear (form III) DNAs. The DNAs treated with 15 and 16 showed no

cleavage. Weak DNA cleavage was observed for DNA treated with 17.

This result suggests that epoxide, ABC-ring system and alkenyl group

attached to epoxide are the units essential for efficiently reaction with

DNA. The thermal treatment of DNA reacted with 2 resulted in distinct

formation of form II DNA. This result indicates that DNA cleavage by

2 proceeded through alkylation of DNA nucleobase.

54

form II

form I

Lane 1 2 3 4 5 6 7 8 9 10

2 16 17 15 cone ( 10 ,uM)

Figure 2. Supercoiled DNA cleavage assay. Supercoiled pBR322 DNA (40 ,uM) was incubated at 37 oc for 5 h in Tris-HCl (pH 7.6) in the presence of 10 ,uM of each drug and the DNA samples were precipitated in ethanol to remove unreacted drug (work 1, evennumbers lanes). The recovered DNA samples were dissolved in TrisHCl buffer (pH 7.6) and further incubated at 37 oc for 24 h (work 2, odd-numbers lanes). Conversion of supercoiled DNA was analyzed by agarose gel electrophoresis with ethidium staining.

We have investigated hydrolysis rates of epoxides of 1, 2, 16 and

17 to estimate the nucleophilicity of epoxide in reaction buffer (Figure

3). We incubated 100 J.LM of each drug in Tris-HCl (pH 7.6). The

hydrolysis reaction was monitored by HPLC at each time. The half

lifetimes of 1, 2 and 16 are 3.5 h, 5.6 h and 5.9 h, respectively, and the

hydrolysis of 17 was found to be very slow (t 112 = > 150 h). The

compounds containing a double bond directly attached to epoxide like 1,

2 and 16 completely decomposed in 48 h, whereas the compound

without double bond like 17 showed a very little hydrolysis rate. The

difference of DNA alkylation rate of 2 and 17 in DNA cleavage assay

arises from the nucleophilicity of their epoxides.

55

8v

~ 60 I;){)

--- 40

20

o £\BC 2

o AB 16 o epoxyalcoho 17

6 kapurimycin A3 1

igu Pl0ts fo ... th~ decrease of drugs by (J -+---r- b d o 1 o 20 30 40 so hydrolysis. 100 ,uM ot each drug was mcu ate

tll ,1, tr 1 n~-HCl (pl ' .5) a 7 °C. Tht.. amount of drug in a reaction buffer was monitored by HPLC .

ing the san c n1ethod as supercoiled DNA cleaving assay shown

above, we compared the DNA cleavage activity of 2 with that of

kapurimycin A3

(1). The result was shown in Figure 4. Kapurimycin

A3

has ca. 10-fold higher activity than 2, because the band of nicked

circular DNA by 10 ,uM of 2 is close to that for 1 ,uM of 1. The DNA

treated with 10 ,uM of 1 appeared as a more mobile smearing band due

to over-breakage of DNA.

form II form III form I

·~ir :·- . -: ---~, ~ t. } ~: "'1 t)l

~;..,.- . . ~' ~.~~\

Lane 2 3 4 5 6 7 8 9 10

cone (,UM) 10 10

2 1

Figure 4. Supercoiled DNA cleavage assay by 1 and 2. Supercoiled pBR322 DNA ( 40 ,uM) was incubated at 37 oc for 5 h in Tris-HCl (pH 7 .6) in the presence of 10 ,uM or 1 ,uM of drug and the DNA samples were precipitated in ethanol to remove unreacted and hydrolyzed drug (work 1, even-numbers lanes). The recovered DNA samples were dissolved in Tris-HCl buffer (pH 7 .6) and further incubated at 37 oc for 24 h (work 2, odd-numbers lanes). Conversion of supercoiled DNA was analyzed by agarose gel electrophoresis with ethidium staining.

56

Next, we investigated the reaction of 2 with calf thymus DNA to

know the nucleobase selectivity of the alkylation by 2. Calf thymus

DNA was incubated with 2 in Tris-HCl (pH 7.6) at 37 oc and digested

with snake venom phosphodiesterase and alkaline phosphatase.

Decreases of given nucleosides were monitored with HPLC (F1gure 5).

The content of dG decreased to 73o/a of that in intact DNA in 24 h,

while other nucleobases remained unchanged. This result shows that 2

is a good guanine alkylating agent.

(a) ; 0 h dG dA

dT

0 5 10 15 20 25 Retention Time (min)

(b)· 24 h dG dT ' 73% 101%

de OJn ~L) .) ~ dA 100 Uul95 %

30 Figure 5. HPLC profiles for nucleotide analysis of (a) intact calf thymus DNA and (b) calf thymus DNA treated with 2. The reaction of calf thymus DNA with 2 was carried out in 50 mM Tris-HCl (pH 7.5) at 37 °C for 24 h. The resulting DNA was digested with snake venom phosphodiesterase and alkaline phosphatase.

0 5 1 0 15 20 25 30 Retention Time (min)

Sequence selectivity for DNA alkylation by 2 was examined by using 32P-5' -end-labeled EcoRI/Rsal fragment of pBR322 DNA. The labeled

DNA was incubated in presence of drug at 37 oc for 24 hand recovered

by ethanol precipitation. Subsequently, a half of each DNA sample was

heated in water at 90 oc for 30 min and another half was treated with

hot piperidine. The result of the assay is shown in Figure 6. ABC

analog 2 exhibited guanine selective DNA cleavage like kapurimycin A3

(1) by treatment with hot piperidine. In the case of the heat-treated

DNAs, the cleavage bands did not comigrate with those of Maxam

Gilbert guanine bands, and slightly shifted to 3 '-side of Maxam-Gilbert

57

guanine bands. This result suggests that ABC analog 2 cleaves DNA via guanine alkylation mechanism. Furthermore, we found that 2 cleaved DNA in sequence selective manner. The selectivity of GN sequence decreased in the order, G*G > G* A > G*T > G*C (G* denotes alkylated guanine). In continuous guanine sequences, 5' -side guanine was more reactive than 3' -side guanine. The sequence selectivity of 2 was very similar to that of 1.

1

2 3

2

4 5

3' -~ A A G G G..-•G~ 'C :r 'T :r 'T : c •A : c I G I G~ : T I G :·G~ I 'A : : c i !E-.-I I c ::~~ I A

i ::: I c I T

T T G G~ T A A T A A T A IG~

,' T

,/ ~ T G..T A A T T G

---G~

T 5'-A

Figure 6. Cleavage assay of 32P-5'-end-labeled DNA fragment (513 bp EcoRI-Rsai fragment of pBR322 DNA) by 1 and 2. DNA was incubated in the presence of I ,uM of 1 (lanes 2 and 3) and 10 ,uM of 2 (lanes 4 and 5) for 24 h at 37 °C in Tris-HCI buffer (pH 7.6). Recovered DNA by ethanol precipitation was heated at 90 °C for 30 min in the absence (lanes 2 and 4 ), and the presence (lanes 3 and 5) of 10% piperidine, and analyzed by electrophoresis on 8% denatured polyacrylamide gel. lane 1, Maxam-Gilbert G+A; lane 2, 1 with heating; lane 3, 1 with piperidine treatment; lane 4, 2 with heating; lane 5, 2 with piperidine treatment.

58

ABC analog 2 was incubated with herring sperm DNA at 37 oc for 5 h in order to get 2-guanine adduct (Scheme 2). After incubation and ethanol precipitation, we heated the recovered DNA in water at 90 oc for 30 min and obtained the adduct by butanol extraction followed by column purification. The data obtained from 1H-NMR spectra in methanol-d4 and mass spectra were shown in Table 1. The mass spectrum shows a peak at m/e 444 ((M+Hr). In the NMR, the signal of H13 of ABC analog 2 considerably shifted to downfield, compared with the other signal of 2. The chemical shift of the signals of 2-guanine adduct is closely similar to those of reported 1-guanine adduct. 3 These data implies that 2-guanine adduct is produced from the nucleophilic attack of guanine N7 to C 13 of 2.

Scheme 2.

Herring Sperm DNA

A cO

2-guanine adduct

1-guanine adduct

59

Table 1. The 1H NMR (400 MHzJ chemical shift

assignments for the guanine adducts in methanol-d4

2 -guanine adduct 1-guanine adduct

8 (L18)a 8 (L18)b

H3 6.47 (--0.09) H3 6.28 ( -0.10)

.L 1 8 I (--0.06) Hl'i 1.81 (--0.10)

H/3 6.,29(+2.69) H/6 6.44 (+2.43)

H14 6.15 (+1.01) H17 6.15 ( + 1.00)

HI~ 6 05 (+0.25) Hl8 6.02 ( +0.20)

H16 1.94 (+0.12) H19 1.90 ( +0.08)

{2_JlQ !ill ~ SJM

a ..18 =-(chemical shi ft of 2-guanine adduct)- (chemical shift of 2)

b L18 = (chemical shift of 1-guanine adduct) - (chemical shift of 1)

Kapurimycin A3 was cytotoxic at pM level against HeLa S3 and T24

cells in vitro. 1 We investigated the cytotoxicities of synthetic analogs 2

and 16 against HeLa cell (Table 2). While 16 didn't show significant

cytotoxicity in 10 pM dose, 2 showed one tenth of IC50 for 1 (IC50 = 2.7

pM). This result is consistent with DNA cleaving experiment.

Conclusion

Table 2. In vitro cytotoxic activities

of 1 , 2 and 16 against HeLa cell

HeLa S3 1 2 16

ICsoa 0.28 ,LLM 2.7 ,LLM > 10 ,LLMb

a Drug concentration causing 50% inhibitior

of cell growth with 72 h drug incubation

b Highest dose tested

We investigated the structure-activity relationship for DNA

alkylation by antitumor antibiotic kapurimycin A3 (1) and its simplified

analogs. It was shown that the epoxide moiety for DNA alkylation, the

60

double bond for activation of epoxide and the tricyclic structure for

DNA binding were all required for the efficient alkylation of DNA.

Kapurirnycin ABC-ring analog 2 fulfills these requirements and

efficiently alkylates DNA with a similar sequence selectivity as that for

kapurirnycin A3 (G*G > G* A > G*T > G*C, G* denotes alkylated

guanine).

61


General Techniques. 1H NMR spectra were measured with JEOL

JNM a-400 (400 MHz) spectrometers. Coupling constants (J values) are

reported in Hz. 13C NMR spectra were measured with Varian GEMINI

200 (50 MHz) spectrometers. The chemical shifts are expressed in ppm

downfield from tetramethylsilane, using residual chloroform ( 8 = 7.24

in 1H NMR, 8 = 77.0 in 13C NMR) as an internal standard. The

following abbreviations were used to explain the multiplicities: s,

si nglet~ d, doublet; t, triplet; q, quartet; m, multiplet; br, broad. IR

spectra were recorded on JASCO FT/IR-5M spectrophotometers.

Melting points were obtained on a Yanagimoto Seisakusho micro

melting point apparatus and are uncorrected. Optical rotations were

recorded using a Perkin-Elmer 243 polarimeter. Electron impact mass

spectra (MS) and high-resolution mass spectra (HRMS) were recorded

on JEOL JMS-DX 300 or JEOL JMS-SX 102A. Microanalyses were

performed by Kyoto University Microanalytical Center.

All reactions were monitored by thin layer chromatography carried

out on 0.25-mm E. Merck silica gel plates (60F-254) using UV light,

5% ethanolic phosphomolybdic acid, or p-anisaldehyde solution and heat

as developing agent. Wako gel (C-200, particle size 75-150 pm, Wako)

was used for column chromatography. Plasmid pBR322 DNA was

purchased from Wako. Tetrahydrofuran and ethyl ether were distilled

over sodium-benzophenone. Dichloromethane, toluene and N,N

dimethylformamide was distilled over calcium hydride. All reagents

were purchased at highest commercial quality and used without further

purification unless otherwise stated.

All reactions were carried out under nitrogen atmosphere with

anhydrous solvents under anhydrous conditions, unless otherwise noted.

62

Yields refer to chromatographically and spectroscopically CH NMR)

homogeneous materials, unless otherwise stated.

(E)-1-( 4-Methoxy benzyloxy )-3-methyl-2-pen ten-4- yne ( 4).

To a suspension of sodium hydride (60 %, 4.5 g, 113.3 mmol) in N,N

dimethylformamide (10 mL) was added 3 (10.01 g, 104.1 mmol) at 0

DC, and the reaction mixture was stirred for 10 min. To this mixture

was added p-methoxybenzyl chloride (15.5 mL, 114.3 mmol) at 0 DC,

and the reaction mixture was stirred for 30 min. The mixture was

diluted with sat. aq. NH4Cl and extracted with ethyl acetate. The crude

product was purified by column chromatography on silica gel, eluting

with 5 % ethyl acetate in hexane to give 4 (20.6 g, 102.0 mmol, 98 o/o)

as a yellow oil: 1H NMR (CDC13, 400 MHz) 8 7.24 (d, 2 H, J = 8.8 Hz),

6.86 (d, 2 H, J = 8.7 Hz), 6.07 (dtq, 1 H, J = 0.6, 1.5, 6.6 Hz), 4.42 (s, 2

H), 4.05 (dt, 2 H, J = 0.7, 6.6 Hz), 3.79 (s, 3 H), 2.81 (s, 1 H); IR

(CHC13) 3304, 3021, 3009, 1612, 1513, 1249, 1217, 1173, 1035, 773,

769 cm- 1•

(2R ,3R )-1-( 4-Methoxybenzyloxy )-3-methyl-4-pentyn-2,3-diol

(5). The mixture of AD-mix-/3 (Aldrich, 12.0 g) and methanesulfon

amide (0.82 g, 8.62 mmol) in t-butanol ( 40 mL) and water ( 40 mL) was

stirred at 0 DC for 30 min. To this mixture was added 4 (1.73 g, 8.54

mmol) at 0 °C, and the reaction mixture was stirred for 2 days. To the

mixture was added sodium sulfite (15 g) and the reaction mixture was

stirred for 2 h. This mixture was extracted with ethyl acetate, and then

the combined organic layer extracted with 1 N sodium hydroxide. The

crude product was purified by column chromatography on silica gel,

eluting with 20 %-100 o/o ethyl acetate in hexane to give 5 (1.63 g, 6.49

mmol, 76 %) as a colorless oil: 1H NMR (CDC13, 400 MHz) 8 7.24 (d, 2

63

H, J = 8.7 Hz), 6.86 (d, 2 H, J = 8.7 Hz), 4.48 (dx2, 2 H, J = 11.4, 13.2

Hz), 3.79 (s, 3 H), 3.74 (dx2, 2 H, J = 4.7, 9.2 Hz), 3.59 (dd, 1 H, J = 6.3, 9.2 Hz), 3.26 (s, 1 H), 2.53 (d, 1 H, J = 5.0 Hz), 2.46 (s, 1 H), 1.47

(s, 3 H); IR (CHC13

) 3562, 3306, 3011, 2937, 1514, 1250, 1174, 1087,

1036, 734 cm-1; MS (El) m/e (%) 250 (M+, 21), 137 (100), 121 (100);

HRMS (El) calcd for C14

H80 4 250.1204 (M+), found 250.1201.

( 4R ,5R) -4-E th yny 1-5- ( 4-methoxy benzyloxymethy I) -2,2,4- tri

methyl-1,3,-dioxolane (6). To a solution of 5 (1.97 g, 7.87 mmol),

2,2-din1ethoxypropane (4.9 mL, 39.85 mmol) and anhydrous copper

sulfate (6.29 g, 39.41 mmol) in acetone (50 mL) was added dl

camphorsulfonic acid (92.4 mg, 0.40 mmol) at 0 °C, and the reaction

mixture was stirred for 3 h. The mixture was filtered, diluted with sat.

aq. NaHC03

, concentrated and extracted with ethyl acetate. The crude


with 15 o/o ethyl acetate in hexane to give 6 (2.17 g, 7.48 mmol, 95 o/o)

as a colorless oil: 1H NMR (CDC13, 400 MHz) 8 7.26 (d, 2 H, 1 = 8.8

Hz), 6.86 (d, 2 H, J = 8.7 Hz), 4.58 (d, 1 H, J = 11.7 Hz), 4.46 (d, 1 H, 1

= 11.7 Hz), 4.42 (dd, 1 H, J = 5.0, 6.5 Hz), 3.79 (s, 3 H), 3.555 (d, 1 H,

J = 5.0 Hz), 3.551 (d, 1 H, J = 6.4 Hz), 2.48 (s, 1 H), 1.46 (d, 3 H, 1 = 0.5 Hz), 1.42 (d, 3 H, J = 0.5 Hz), 1.34 (s, 3 H); IR (CHC13) 3020, 2993,

1513, 1375, 1249, 1088, 780 cm-1; MS (EI) m/e (o/o) 290 (M+, 13), 275

(47), 232 (81), 137 (97), 121 (100); HRMS (El) calcd for C17H220 4 (M+)

290.1518, found 290.1524.

( 4R ,5R)-5-( 4-Methoxybenzyloxymethyl)-2,2,4-trimethyl-4-

(2-trimethylsilylethynyl)-1,3,-dioxolane (7). To a solution of 6

(1.63 g, 5.62 mmol) in tetrahydrofuran (20 mL) was added lithium

hexadisilazide (l.OM solution in tetrahydrofuran, 6.2 mL, 6.20 mmol)

64

at -78 °C, and the reaction mixture was stirred for 10 minutes. To this

mixture was added trimethylchlorosilane (0. 78 mL, 6.16 mmol) at -78

°C, and the reaction mixture was warmed to 0 OC and stirred for 20

m1n. The mixture was diluted with sat. aq. NH4Cl and extracted with

ethyl acetate. The crude product was purified by colun1n

chromatography on silica gel, eluting with 15 o/o ethyl acetate in hexane

to give 7 (1.79 g, 4.94 mmol, 88 o/o) as a colorless oil: 1H NMR (CDC13,

400 MHz) 8 7.27 (d, 2 H, J = 8.8 Hz), 6.86 (d, 2 H, J = 8.7 Hz), 4.58 (d,

1 H, J = 11.6 Hz), 4.46 (d, 1 H, J = 11.6 Hz), 4.39 (dd, 1 H, J = 4.4, 7.0

Hz), 3.79 (s, 3 H), 3.57 (dd, 1 H, J = 4.4, 10.3 Hz), 3.52 (dd, 1 H, J = 7.0, 10.3 Hz), 1.46 (s, 3 H), 1.41 (s, 3 H), 1.32 (s, 3 H), 0.13 (s, 9 H);

IR (CHC13) 3020, 1513, 1250, 1180, 1087, 848 cm-1; MS (EI) m/e (%)

347 [(M-Met, 1], 258 (4), 137 (35), 121 (100).

( 4R ,5R)-5-Hydroxymethyl-2,2,4-trimethyl-4-(2' -trimethyl

silylethynyl)-1,3-dioxolane (8). To a solution of 7 (0.31 g, 0.85

mmol) in dichloromethane (18 mL) and water (1 mL) was added

dichlorodicyanoquinone (0.23 g, 1.01 mmol) at 0 OC, and the reaction

mixture was stirred at ambient temperature for 4 h. The mixture was

diluted with sat. aq. NaHC03, and extracted with ethyl acetate. The

crude product was purified by column chromatography on silica gel,

eluting with 60 % toluene in hexane to give 8 (0.18 g, 0.84 mmol, 99

%) as a colorless needle: mp. 45 OC; 1H NMR (CDC13, 400 MHz) 8 4.31

(dd, 1 H, J = 3.9, 7.4 Hz), 3.73 (2 H), 1.73 (m, 1 H), 1.49 (s, 3 H), 1.43

(s, 3 H), 1.37 (s, 3 H), 0.14 (s, 9 H); IR (CHC13) 3020, 1217, 847, 754

cm-1; Anal. Calcd for C12H220 3Si: C, 59 .46; H, 9.15. Found: C, 59 .20;

H, 9.30.

65

( 4R ,5S)-5-Formyl-2,2,4-trimethyl-4-(2' -trimethylsilyl

ethynyl)-1,3,-dioxolane (9). To a solution of 8 (0.81 g, 3.34

mmol) in dichloromethane (30 mL) was added Dess-Martin periodinane

(1.7 g, 4.01 mmol) at ambient temperature, and the reaction mixture

was stirred at ambient temperature for 1 h. The mixture was diluted

with sat. aq. N~S203 and sat. aq. NaHC03, and then extracted with ethyl

acetate. The crude product was purified by column chromatography on

silica gel, eluting with 25 % ethyl acetate in hexane to give 9 (0.76 g,

3.14 mmol, 94 o/o) as a colorless oil: 1H NMR (CDC13, 400 MHz) 8 9.61

(d, 2 H, J = 2.2 Hz), 4.54 (d, 2 H, J = 2.4 Hz), 1.56 (d, 3 H, J = 0.5 Hz),

1.53 (d, 3 H, J = 0.5 Hz), 1.43 (s, 3 H), 0.16 (s, 9 H); MS (EI) m/e (o/o)

235[(M-Me)+, 10], 153 (55), 100 (74), 85 (100).

( 4R ,5R ,1 'Z)-5-(1' -Propenyl)-2,2,4-trimethyl-4-(2' '-tri

methylsilylethynyl)-1,3,-dioxolane (10). To a solution of 9

(0.67 g, 2.80 mmol) in toluene (20 mL) was added 12 mL of

ethylidenylphosphorane toluene solution (prepared from

ethyltriphenylphosphonium bromide (1.86 g, 5.01 mmol) and potassium

hexamethyldisilazide (0.5 M in toluene, 10 mL, 5.0 mmol) in toluene

(10 mL) at reflux for 2 h) at -78 °C, and the reaction mixture was

stirred at -78 oc for 1 h and at 0 OC for 1 h. The mixture was diluted

with sat. aq. NH4Cl and extracted with ethyl acetate. The crude product

was purified by column chromatography on silica gel, eluting with 30 o/o

toluene in hexane to give 10 (0.55 g, 2.17 mmol, 78 o/o, Zl E = 10:1

n1ixture) as a colorless oil. (Z)-formed olefin was isolated with

recycling preparative HPLC (solvent, chloroform; Japan Analytical

Industry Co.,Ltd.): 1H NMR (CDC13, 400 MHz) 8 5.82 (ddq, 1 H, J = 1.3, 7.0, 11.1 Hz), 5.40 (m, 1 H), 5.05 (d, 1 H, J = 8.5 Hz), 1.78 (dd, 3

H, J = 1.8, 7.0 Hz), 1.4 7 (s, 3 H), 1.43 (s, 3 H), 1.30 (s, 3 H), 0.13 (s, 9

66

H); IR (CHC13) 3017, 1730, 1250, 723 cm-1; MS (El) mle (o/o) 237 l(M

Mer, 11], 167 (27), 83 (100).

( 4R,5R,1' Z)-4-Ethynyl-5-(1 '-propenyl)-2,2,4-trimethyl-1,3,

dioxolane (11). To a solution of 10 (0.21 g, 0.82 mmol) in methanol

(1 0 mL) was added a drop of sodium methoxide methanol solution at 0

°C, and the reaction mixture was stirred at ambient temperature for 2 h.

The mixture was diluted with sat. aq. NH4Cl and extracted with pentane,

ether and water. After its concentration, 11 (0.12 g, 0.66 mmol, 80

%) was afforded as a colorless oil: 1H NMR (CDC13, 400 MHz) 8 5.84

(ddq, 1 H, J = 1.2, 7.0, 11.1 Hz), 5.40 (ddq, 1 H, J = 1.8, 8.6, 11.1 Hz),

5.08 (dd, 1 H, J = 1.1, 8.6 Hz), 2.47 (s, 1 H), 1.78 (dd, 3 H, J = 1.8, 7.0

Hz), 1.48 (s, 3 H), 1.44 (s, 3 H), 1.34 (s, 3 H); IR (CHCl3) 3017, 2932,

1721, 1291, 1222, 1077, 788 cm-1•

( 4R ,5R ,1 'Z)-4-{3-[1-(t-Butyldimethylsilyloxy)-2-naphthyl]-3-

hydroxy-1-propynyl }-5-(1 '-propenyl)-2,2,4-trimethyl-1 ,3,

dioxolane (12). To a solution of 11 (28.0 mg, 0.16 mmol) and

anhydrous cerium chloride (228.7 mg, 0.93 mmol) in tetrahydrofuran

(1 mL) stirred at ambient temperature for a hour was added lithium

hexadisilazide (l.OM solution in THF, 0.31 mL, 0.31 mmol) at -78 °C,

and the reaction mixture was stirred at -78 OC for 30 min. After

addition of 1-tert-butyldimethylsiloxy-2-naphthaldehyde ( 48.5 mg, 0.17

mmol) in tetrahydrofuran (0.2 mL) at -78 OC, the reaction mixture was

stirred at -78 OC for 10 min. It was diluted with sat. aq. NH4Cl at -78

OC, and extracted with ethyl acetate. The crude product was purified by

column chromatography on silica gel, eluting with 10 % ethyl acetate in

hexane to give 12 (34.0 mg, 72.9 Jlmol, 47 %) as a pale yellow oil: 1H

NMR (CDC13, 400 MHz) 8 8.06 (1 H, naphthalene), 7.77 (1 H,

67

naphthalene), 7.68 (1 H, naphthalene), 7.52 (1 H, naphthalene), 7.44 (2

H, naphthalene), 5.99 (1 H, 1 = 4.5 Hz, naph-CH), 5.78 (1 H, CH3-

CH=CH), 5.38 (1 H, CH3-CH=CH), 5.04 (1 H, CH3-CH=CH-CH), 2.12

(1 H, OH), 1.70, 1.61 (3 H, 1 = 1.8, 7.0 Hz, CH3-CH=CH), 1.47-1.41 (6

H. ace toni de), 1.36 (sx2, total 3 H, 4-CH3), 1.12 (sx2, total 9 H, Si-13u),

0.18 (6H, Si-CH,); IR (CHC11) 3017,2932, 1721, 1291, 1222, 1077,737

cm-1; MS m/e (o/o) 466 (M+, 7), 281 (100), 266 (26), 229 (31); HRMS

(El) calcd for C28

H38

0 4Si (M+) 466.2540, found 466.2533.

( 4R ,5R ,1 'Z)-4-{3-[1-(t-Butyldimethylsilyloxy)-2-naphthyl]-3-

oxo-1-propynyl }-5-(1 '-propenyl)-2,2,4-trimethyl-1,3,

dioxolane (13). To a solution of 12 (30.2 mg, 64.7 pmol) in

dichloromethane (2 mL) was added manganese (IV) oxide (11.3 mg),

and the reaction mixture was stirred for 24 h at ambient temperature.

The mixture was diluted with ethyl ether, filtered, and concentrated in

vacuo. The crude product was purified by column chromatography on

silica gel, eluting with 10 o/o ethyl acetate in hexane to give 13 (22.9 mg,

49.3 pmol, 76 o/o) as a yellow oil: 1H NMR (CDC13, 400 MHz) 8 8.24 (d,

1 H, 1 = 8.5 Hz), 7.94 (d, 1 H, 1 = 8.7 Hz), 7.78 (d, 1 H, 1 = 8.0 Hz),

7.56 (ddd, 1 H, 1 = 1.2, 6.8, 8.1 Hz), 7.51-7.45 (2 H), 5.87 (ddq, 1 H, 1

= 1.2, 7.1, 11.0 Hz), 5.44 (ddq, 1 H, 1= 1.8, 8.7, 10.8 Hz), 5.18 (dd, 1

H, 1 = 1.1, 8.7 Hz), 1.78 (dd, 3 H, 1 = 1.8, 7.1 Hz), 1.49 (s, 3 H), 1.45

(s, 3 H), 1.10 (s, 9 H), 0.11 (s, 6 H); IR (CHC13) 3023, 2988, 1729,

1597, 1375, 1251, 844 cm-1; MS (EI) m/e (o/o) 407 [(M-13ur, 6], 293

(52), 280 (100), 238 (73), 222 (80), 149 (94); HRMS (EI) calcd for

C24H270 4Si [(M-13urJ 407.1679, found 407.1652.

( 4S ,5R ,1 'Z)-4-( 4H -naphtho[1,2-b ]pyran-2-yl)-5-(1'

propenyl)-2,2,4-trimethyl-1,3,-dioxolane (14). To a solution of

68

13 (20.1 mg, 43.3 pmol) and 18-crown-6 (23.2 mg, 87.8 pmol) in N,N

dimethylformamide (1 mL) was added a potassium fluoride (5.2 mg,

89.6 pmol) at 0 °C, and the reaction mixture was stirred at an1bient

temperature for 2 h. The mixture was diluted with sat. aq. NH4Cl, and

extracted with ethyl acetate. The crude product was purified by colutnn

chromatography on silica gel, eluting with 15 o/o ethyl acetate in hexane

to give 14 (12.1 mg, 34.5 pmol, 80 %) as a yellow oil: 1H NMR

(CDC13, 400 MHz) 8 8.28 (m, 1 H), 8.14 (d, 1 H, 1 = 8.8 Hz), 7.92 (d, 1

H, J = 11.4 Hz), 7.76 (d, 1 H, 1 = 8.6 Hz), 7.69 (ddd, 1 H, 1 = 1.4, 7.0,

8.2 Hz), 7.63 (ddd, 1 H, 1 = 1.4, 6.9, 8.3 Hz), 6.81 (s, 1 H), 5.99 (ddq, 1

H, 1= 1.0, 7.1, 11.1 Hz), 5.72 (ddq, 1 H, 1= 1.8, 9.0, 10.9 Hz), 5.02

(dd, 1 H, 1 = 1.0, 9.0 Hz), 1.61 (s, 3 H), 1.60 (s, 3 H), 1.53 (s, 3 H),

1.40 (dd, 3 H, 1 = 1.8, 7.1 Hz); IR (CHC13) 3021, 1649, 1390, 1230,

1109, 1015, 796 cm-1; MS (EI) mle (%) 350 (M+, 5), 335 [(M-Mer,

11], 293 (46), 280 (77), 239 (100); HRMS (El) calcd for C21 H190 4 [(M

MerJ 335.1283, found 335.1297.

2- [ (2S ,3R ,4Z)-2,3-Dihydroxy -4-hex en- 2-yl] -4H- na ph tho [ 1,2-

b ]pyran (15). To a solution of 14 (7 .0 mg, 20.0 pmol) in

tetrahydrofuran (0.5 mL) and acetic acid (0.5 mL) was added 0.2 M

hydrochloric acid (0.05 mL) at 0 °C, and the reaction mixture was

stirred at ambient temperature for 5 days. The mixture was poured

onto sat. aq. NaHC03 and ethyl acetate, and extracted. The crude


with 50 %ethyl acetate in hexane to give 15 (6.1 mg, 19.7 pmol, 98 %)

as a white solid: 1H NMR (CDC13, 400 MHz) 8 8.38 (m, 1 H), 7.93 (d, 1

H, J = 8.8 Hz), 7.87 (m, 1 H), 7.68 (2 H), 7.60 (d, 1 H, 1 = 8.6 Hz), 6.78

(s, 1 H), 5.86 (ddq, 1 H, 1 = 0.9, 7.0, 11.0 Hz), 5.67 (ddq, 1 H, J = 1.8,

9.3, 11.0 Hz), 5.04 (d, 1 H, 1 = 9.1 Hz), 3.40 (s, 1 H), 2.70 (s, 1 H),

69

1.76 (dd, 3 H, 1 = 1.7, 7.0 Hz), 1.57 (s, 3 H)~ IR (CHC13) 3020, 1647,

1630, 1391, 1217, 795 cm- 1 ~ MS (El) m/e (%) 310 (M+, 4), 240 (100),

197 (82); HRMS (EI) calcd for C19H180 4 (M+) 310.1205, found

310.1204.

2-[ (2S ,3S ,4Z)-2,3-Epoxy-4-hexen-2-yl]-4H -naphtho[1,2-b]

pyran (2). To a solution of 15 (39.7 mg, 0.13 mmol) and

triphenylphosphine (50.5 mg, 0.19 mmol) in toluene (2 mL) was added

diethyl azadicarboxylate (0.060 mL, 0.38 mmol) at 0 OC, and the

reaction mixture was stirred at ambient temperature for 3 h. The

mixture was filtered and concentrated in vacuo. The crude product was

purified by column chromatography on silica gel, eluting with 20 o/o

toluene in hexane to give optically active 2 (28.6 mg, 0.098 mmol, 76

o/o) as a White SOlid: mp. 118.0-119.2 oc~ [af5 D = -61.5° (C 0.26,

MeOH); 1H NMR (CDC13

, 400 MHz) 8 8.46 (m, 1 H), 8.13 (d, 1 H, 1 =

8.7 Hz), 7.93 (m, 1 H), 7.76 (d, 1 H, 1 = 8.6 Hz), 7.72-7.67 (2 H), 6.56

(s, 1 H), 5.80 (ddq, 1 H, 1= 1.0, 7.2, 11.1 Hz), 5.14 (ddq, 1 H, 1= 1.8,

8.6, 11.1 Hz), 3.90 (d, 1 H, 1 = 9.3 Hz), 1.87 (s, 3 H), 1.82 (dd, 3 H, 1 =

1.8, 7.1 Hz); IR (CHClJ 3015, 1650, 1633, 1441, 1391 cm-1

; MS (EI)

m/e (o/o) 292 (M+, 25), 250 (77), 235 (60), 171 (100), 135 (39), 105

(44); HRMS (EI) calcd for C10H160 3 (M+) 292.1098, found 292.1073.

2-[ (2R * ,3S * ,4Z)- and (2R * ,3S* ,4E)-2,3-Epoxy-4-hexen-2-yl]-

4H-chromen-4-one (16). 1H NMR (CDC13, 400 MHz) 8 8.19 (dd, E-

1 H, 1 = 1.7, 8.0 Hz), 8.18 (dd, Z-1 H, 1 = 1.7, 8.0 Hz), 7.66 (ddd, E-1

H, 1 = 1.7, 7.1, 8.4 Hz), 7.65 (ddd, Z-1 H, 1 = 1.7, 7.1, 8.4 Hz), 7.44

(m, E-1 H), 7.43 (m, Z-1 H), 7.39 (m, E-1 H), 7.38 (m, Z-1 H), 6.42 (s,

E-1 H), 6.40 (s, Z-1 H), 6.04 ( dq, E-1 H, 1 = 6.7, 15.4 Hz), 5.80 ( ddq,

Z-1 H, 1 = 1.0, 7.1, 11.2 Hz), 5.13 (ddq, E-1 H, 1 = 1.7, 8.6, 15.4 Hz),

70

5.17 (ddq, Z-1 H, 1 = 1.9, 8.7, 11.0 Hz), 3.82 (d, Z-1 H, 1 = 8.6 Hz),

3.52 (d, E-1 H, 1 = 8.6 Hz), 1.80 (dd, Z-3 H, 1 = 1.8, 7.1 Hz), 1.76 (s,

Z-3 H), 1.72 (s, E-3 H), 1.65 (dd, E-3 H, 1 = 1.7, 6.6 Hz)~ IR (CHC13)

3020, 1649, 1608, 1252, 849 cm- 1 ~ MS (EI) mle (%) 242 (M+, 13), 227

(15), 211 (24), 200 (92), 185 (73), 121 (100); HRMS (El) calcd for

C15H140 3 (M+) 242.0942, found 242.0920.

2-[(2S* ,3S*)-2,3-Epoxy-4-hydroxy-2-butyi]-4H -naphtho[1,2-

b ]pyran (17). mp. 175 OC; 1H NMR (CDC13, 400 MHz) 8 8.45 (m, 1

H), 8.11 (d, 1 H, 1 = 8.8 Hz), 7.92 (m, 1 H), 7.76 (d, 1 H, 1 = 8.6 Hz),

7.69 (2 H), 6.53 (s, 1 H), 3.72 (2 H), 3.44 (dd, 1 H, 1 = 4.9, 6.3 Hz),

1.86 (s, 3 H), 1. 72 (br, 1 H); IR (CHC13) 3022, 1650, 1441, 1391, 1228,

828 cm- 1; MS (EI) m/e (%) 282 (M+, 39), 212 (40), 170 (100), 114

(68); HRMS calcd for C17H140 4 (M+) 282.0892, found 282.0875.

Hydrolysis of 2. A solution of 2 (29.3 mg, 0.10 ,umol) in 10% (v/v)

acetonitrile/50 mM Tris-HCl buffer (10 mL, pH 7.6) was stirred at 37

oc for 5 days. After concentration, the mixture was extracted with

ethyl acetate and water. The crude product was purified by column

chromatography on silica gel, eluting with 20-60% ethyl acetate in

hexane to give 9 (16.5 mg, 53.2 ,umol, 53o/o) as a white solid, 12 (2.2

mg, 7.1 ,umol, 7%) and recovered 2 (11.0 mg, 37.6 ,umol, 38o/o).

Methanolysis of 2. To 9 mL of methanol solution of Tris-HCl salt,

which was obtained by concentration of 0.5 M Tris-HCl buffer (9 mL,

pH 7.6) in vacuo, was added 2 (14.6 mg, 49.9 ,umol) in acetonitrile (1

mL). After stirred at reflux for 5 days, the mixture was concentrated

and extracted with ethyl acetate and water. The crude product was

purified by column chromatography on silica gel, eluting with 20%

71

ethyl acetate in hexane to give 13 (8.1 mg, 25.0 ,Umol, 50%) as a white

solid, 14 (2.4 mg, 7.4 .umol, 15%) and recovered 2 (7.7 mg, 12.7

,Umol, 25o/o ).

Cleavage of pBR322 Supercoiled DNA by 2, 9, 10 and 11.

Kapurimycin analogs 2, 9, 10 or 11 ( 10 ,UM) was incubated with 40

,UM pBR322 supercoiled DNA (Nippon Gene) in 10o/o (v/v)

acetonitrile/50 mM Tris-HCl buffer (10 ,UL, pH 7.5) at 37 oc for 5 h.

The samples which require the thermal treatment were successively

precipitated in ethanol to remove the drug and the recovered DNA

pellets were dried and incubated again in 10 ,UL of 50 mM Tris-HCl

buffer at 37 oc for 24 h. To both thermal-treated and no treated

samples was added the 10 ,UL of loading buffer containing 0.05% (w/v)

bromophenol blue and 6o/o (v/v) glycerol for electrophoresis. The

different form of DNA were separated at ambient temperature on a 1%

(w/v) agarose gel involving ethidium bromide (0.5 ,Ug/mL). The gels

were placed on a UV transilluminator (313 nm) and photographed with

Polaroid 665 film. The result was shown in Figure 2.

Observation of Hydrolysis Rates of Drugs. The 100 ,UM of 1, 2,

10 or 11 was incubated in 10o/o (v/v) acetonitrile/50 mM Tris-HCl

buffer (200 pL, pH 7 .5) at 37 °C. The amount of drug in the reaction

buffer was monitored by HPLC analysis on a COSMOSIL 5C18-AR

column ( 4.6x 150 mm, elution with a solvent mixture (1: 1) of 0.05 M

ammonium formate and acetonitrile at a flow rate of 1.0 mL/min for 20

min). The amount of drug was determined by ratio of the peak areas of

original drug to all peak areas. The result was shown in Figure 3.

72

Cleavage of pBR322 Supercoiled DNA by 1 and 2.

Kapurimycin A3 1 or its analog 2 (1 0 ,UM or 1 ,UM) was incubated with

40 f.LM pBR322 supercoiled DNA (Nippon Gene) in 10o/o (v/v)

acetonitrile/50 mM Tris-HCl buffer (10 f.LL, pH 7.5) at 37 oc for 5 h.

The samples were successively precipitated in ethanol to remove the

drug and the recovered DNA pellets were dried and incubated again in

10 f.LL of 50 mM Tris-HCl buffer at 37 °C for 24 h. To thermal-treated

samples was added the 10 f.LL of loading buffer containing 0.05% (w/v)

bromophenol blue and 6o/o (v/v) glycerol for electrophoresis. The

different form of DNA were separated at ambient temperature on a 1%

(w/v) agarose gel involving ethidium bromide (0.5 f.Lg/mL). The gels

were placed on a UV transilluminator (313 nm) and photographed with


HPLC Analysis of Base-Selectivity of 2. The 50 f.LM of 2 was

incubated with 200 f.LM of calf thymus DNA in 10% (v/v) acetonitrile/50

mM Tris-HCl buffer (100 f.LL, pH 7.5) at 37 oc for 24 h. The resulting

DNA was digested with snake venom phosphodiesterase and alkaline

phosphatase. The concentrations of the produced four nucleotides were

confirmed by HPLC analysis on a COSMOSIL 5C 18-AR column

(4.6x150 mm, elution with a solvent mixture of 0.05 M ammonium

formate, linear gradient over 20 min from 2% to 5% acetonitrile at a

flow rate of 1.0 mL/min), and determined by comparison of the peak

areas of dG, dT and dA with that of dC. The result was shown in

Figure 5.

Preparation of 32P-5'-End-Labeled DNA Fragments of

pBR322 DNA. Digestion of supercoiled pBR322 plasmid DNA

(Nippon Gene) with EcoRI restriction endonuclease followed by

73

treatment with alkaline phosphatase gave a linearized pBR322 with

hydroxyl termini at its 5' -ends. Labeling at the 5' -end of the linearized

DNA was achieved by treatment with [ y-32P]ATP and T4 polynucleotide

kinase. The labeled DNA was further digested with Rsal restriction

endonuclease to yield two 5 '-end-labeled DNA fragments (167 and 513

base pair) which purified on 6% nondenaturing polyacrylamide gel.

The labeled DNA was recovered from the gel by a crush and soak

method. 11

Cleavage of 32P-S' -End-Labeled Fragment of pBR322 DNA.

The 1 ,uM of 1 or 10 ,uM of 2 was incubated with 10 ,uM of calf thymus

DNA and ca. l.Ox106 cpm 32P-5' -end-labeled DNA fragment in 20 mM

Tris-HCl buffer (100 ,uL, pH 7.6) at 37 oc for 24 h. The reaction

mixture was precipitated with ethanol and dried. The recovered DNA

was dissolved in 100 ,uL of water or 10% (v/v) piperidine and heated at

90 oc for 30 min. The mixture was concentrated in vacuo and

resuspended in 10 ,uL of 80% formamide loading buffer (80%

formamide, 1 mM EDT A, 0.1 o/o xylene cyanole and 0.1% bromophenol

blue). The samples (1 ,uL) were loaded onto 8o/o polyacrylamide and 7

M urea sequence gel and electrophoresesed at 1900 V for ca. 2 h. The

gel was dried and exposed to X-ray film with intensifying sheet at -70

°C. The result was shown in Figure 6.

2-Guanine Adduct. A solution of herring sperm DNA (10 mg) in

10o/o (v/v) acetonitrile/50 mM Tris-HCl buffer (2 mL, pH 7.5) was

stirred with 2 (2.9 mg, 0.01 mmol) at 37 oc for 5 h. The reaction

mixture was successively precipitated in ethanol to remove the drug and

the recovered DNA pellets were dried and incubated again in 50 mM

74

Tris-HCl buffer (2 mL) at 90 oc for 30 min. The mixture was

extracted with n-butanol and water and the organic layer was

concentrated in vacuo. The crude product was purified by reversed

column chromatography on ODS (W akogel LP-40C 18, 20--40 ,urn)

eluting with 30% water in acetonitrile to give 2-guanine adduct (0.8 mg,

1.8 jlmol, 18%) as white solid.

75


(1) (a) Hara, M.; Mokudai, T.; KobayashL E.: Gomi, K.; Nakano, H. 1.

Antibiot. 1990 ,43, 1513-1518. (b) Yoshida, M.; Hara, M.; Saitoh, Y.;

~ano, H. 1. Antibiot. 1990,43, 1519-1523.

(2) The absolute configuration of 1 has been determined to be 8S, 14S

and 16S. Uosaki, Y.; Saito, H. Abstract paper p 1013, 69th annual

meeting of the Chemical Society of Japan, Kyoto (1995).

(3) (a) Hara, M.: Yoshida, M.; Nakano, H. Biochemistry 1990, 29,

10449-10455. (b) Chan, K. L.; Sugiyama, H.; Saito, I. Tetrahedron

Lett. 1991, 52, 7719-7722. (c) Chan, K. L.; Sugiyama, H.; Saito, I.;

Hara, M. Phytochemistry 1995, 40, 1373-1374.

(4) (a) Stone, M. P.; Gopalakrishnan, S.; Harris, T. M.; Graves, D. E. 1.

Biomol. Struct. Dyn. 1988, 5, 1025-1041. (b) Gopalakrishnan, S.;

Byrd, S.; Stone, M. P.; Harris, T. M. Biochemistry 1989, 28, 726-734.

(c) Stone, M. P.; Gopalakrishnan, S.; Raney, K. D.; Raney, V. M.;

Byrd, S.; Harris, T. M. In Molecular Basis of Specificity in Nucleic

Acid-Drug Interactions; Pullman, B., Jortner, J., Eds.; Kluwer

Academic Publishers: 1990, p 451-480. (d) Raney, K. D.;


Toxicol. 1990, 3, 254-261. (e) Raney, K. D.; S.; Harris, T. M.; Stone,

M. P. Chern. Res. Toxicol. 1993, 6, 64-68. (f) Gopalakrishnan, S.;

Harris, T. M.; Stone, M. P. Biochemistry 1990, 29, 10438-10448.

(5) (a) Jackson, M.; Karwowski, J. P.; Theriault, R. J.; Hardy, D, J.;

Swanson, S, J.; Barlow, G, J.; Tillis, P. M.; McAlpine, J. B. 1. Antibiot.

1990, 43, 223-228. (b) Brill, G. M.; McAlpine, J. B.; Whisttern, D.

N.; Buko, A. M 1. Antibiot. 1990, 43, 229-237. (c) Sun. D.; Hansen,

M.; Clement, J. J.; Hurely, L. H. Biochemistry 1993, 32, 8068-8074.

(d) Hansen, M; Hurley, H. 1. Am. Chern. Soc. 1995, 117, 2421-2429.

76

(c) Sun, D.; Hansen, M.; Hurley, H. 1. Am. Chem. Soc. 1995, 117,

2430-2440.

(6) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.;

Hartung, J.; Jeong, K.-S.; Kwong, H.-L.; Morikawa, K.; Wand, z.-M.;

Xu, D.; Zhang, K.-L. 1. Org. Chern. 1992, 57, 2768-2771.

(7) E- and Z-mixtures are not separable throughout the synthesis of 2.

(8) (a) Nakatani, K.; Okamoto, A.; Yamanuki, M.; Saito, I. 1. Org.

Chern. 1994, 59, 4360-4361. (b) Nakatani, K.; Okamoto, A.; Saito, I.

Tetrahedron 1996, 52, 9427-9446.

(9) Mitsunobu, 0. Synthesis 1981, 1-28.

(10) Ohtani, I.; Kusurni, T.; Kashman, Y.; Kakisawa, H. 1. Am. Chern.

Soc. 1991,113,4092-4096.


77

CHAPTER 3

Guanine-Guanine Sequence Selectivity for DNA Alkylation by Kapurimycin Analogs

Abstract: In nature there has been known several guanine selective

DNA alkylating antitumor antibiotics. In order to estimate the

contribution of the DNA binding ability to the guanine-guanine (GG)

selective cleavage, we synthesized kapurimycin A3 analogs (S, S)-ABC 2

and (S, S)-ABCD' -ring analogs 3, and compared their sequence

selectivity for DNA cleavage. We found that 3 exceeded 2 in DNA

cleaving activity and in sequence selectivity of DNA cleavage (GG > GA

> GT > GC) as judged from the cleaving assay of 32P-5' -end-labeled

oligodeoxynucleotides. The tetracyclic aromatic ring system of 3

intercalates into DNA duplex more efficiently than 2. Besides, in the

cleaving assay of oligonucleotides having bulge sites, it was shown that

ABCD' -ring analog 3 selectively alkylated 5' -side guanine of its

intercalation site. The above results suggest the selective alkylation of

5' -side guanine of GG sequence by ABCD' -ring analog 3 was originated

by the intercalation of 3 into two guanines of GG sequence. We

speculate the reaction mechanism of GG selective DNA alkylation by

natural antibiotics that is the closely related to the mechanism for the

model system.

79

Introduction

In nature, there are several types of antitumor antibiotics that can

alkyl ate guanine base of duplex DN A.1

Aflatoxin B 1 oxide/

psorospermin3 and kapurimycin A34 are typical such DNA alkylating

agents of natural origin. These antibiotics commonly possess an

aromatic ring system for DNA binding and an epoxy subunit for DNA

alkylation. The mechanism of guanine alkylation by these compounds

involves two steps, a DNA binding step and a guanine alkylating step as

shown in eq 1. The DNA binding step is most important for the

efficiency and the selectivity for guanine alkylation.

0 OMe A cO

aflatoxin B 1 oxide psorospermin kapurimycin A3 (1)

drug+ DNA ~ [drug ... DNA] __.. drug-DNA (1)

Kapurimycin A3

(1) is an effective guanine alkylating agent, which

consists of tetrahydroanthrapyranone ring and a vinyl epoxide side

chain. In the previous study,5 we found that 1 alkylates guanine with

sequence selectivity of 5' -G*G > G* A> G*T > G*C, where G* denotes

alkylated guanine site. Besides, it was found that kapurimycin A3 ABC

ring analog 2 showed a similar sequence selectivity to 1. Known

guanine-guanine (GG) sequence selective alkylators are not only 1 and 2

but also aflatoxin B 1

oxide and psorospermin, which, however,

effectively alkylated 3' -side guanine of GG sequence.

80

We investigated the mechanism of GG selective DNA alkylation in

order to know the correlation of their DNA binding ability with the

sequence selective alkylation. We prepared kapurimycin ABCD ' -ring

analog 3 containing four aromatic rings, and compared the sequence

selectivity for DNA alkylation with that of 2. Herein, we report that the

intercalation of the aromatic ring into DNA is very important for GG

selective alkylation.

(S, S)-ABC (2) (S, S)-ABCD' (3)


The synthetic route for ABCD' -ring analog 3 is outlined in Schemes

1 and 2. The BCD' -ring unit was prepared starting from commercially

available 1-hydroxyanthraquinone ( 4 ). Anthracene 6 was obtained by

reduction of anthraquinone 5 which was then formylated to 7. The

protective group of phenol 7 was converted from methoxymethyl group

to tert-butyldimethylsilyl group to give 9. The side chain moiety 10

was prepared as shown in a previous paper. 5 Addition of 9 to lithiated

10 provided coupling product 11. The oxidation of 11 with manganese

dioxide produced the corresponding ketone which was subsequently

treated with KF and 18-crown-6 in DMF as reported earlier6 to give

tetracyclic compound 12. Hydrolysis of the acetal produced diol which

81

was subjected to intramolecular Mitsunobu reaction to furnish ( 13S,

155)-3.7

Scheme 1.0

b ~

0 4· R1 = IJ a( 5: R 1 =MOM

~ ~

6

a Reagents and Conditions: a) MOMCl, 96%; b) NaBH4, 87%; c) n-BuLi then DMF, 80%; d) HCl, 93%; e) TBDMSOTf, crude product.

Scheme 2.a

Yo 07

10 11 12 3

a Reagents and Conditions: a) LHMDS, CeC13 then 9, 35%; b) Mn02, 69%; c) KF, 18-crown-6, DMF, 75%; d) HCl, 91 %; e) DEAD, PPh3, 63%.

The DNA cleaving activity of 2 and 3 was demonstrated by

monitoring the conversion of supercoiled pBR322 DNA (form I) to

nicked circular (form II) and linear duplex (form III) DNAs. We

incubated 10 ,uM or 1 ,uM of 2 and 3 with 40 ,uM of DNA in Tris-HCl

(pH 7 .5) at 37 oc for 5 h. After ethanol precipitation for the removal

of unreacted and hydrolyzed drug, the DNA was incubated again in

water at 37 oc for 24 h. The different forms of DNA were separated

on agarose gel. The incubation of 3 with DNA converted form I DNA

to form II and form III DNAs in a dose-dependent fashion. ABCD'

ring analog 3 cleaved DNA more efficiently than ABC-ring analog 2.

82

form II

form III

form I

Lane cone (.uM)

2 3 4 5 10 1 10 1

'------'

2 3 Figure _1. Cleavage assay of supercoiled DNA by analogs 2 and 3. Superc01led pBR322 DNA (40 ,UM) was incubated at 37 oc for 5 h m Tris-HCl (pH 7.6) in the presence of 10 ,UM or 1 ,UM of drug and the DNA samples were precipitated with ethanol to remove unreacted drug. The recover~d DNA samples were dissolved in Tris-HCJ buffer (pH 7 6) and further mcubated at 37 oc for 24 h. Conversion of supercoiled DNA was analyzed by agarose gel electrophoresis with ethidium bromide staining.

In order to know the relationship between ring system of

kapurimycin A3 (1) and its analogs 2 and 3 and the sequence selectivity

for DNA cleavage, the reactions of DNA with 1, 2 and 3 were

examined by using 32P-5' -end-labeled oligodeoxynucleotides and

analyzed by electrophoresis on denatured polyacrylamide gel (Figure 2).

The 32P-5' -end-labeled oligodeoxynucleotide duplex was incubated with

these drugs at 37 oc for 24 hand heated in 10% piperidine at 90 oc for

30 min to induce strand breakage. DNA cleavage was observed at all

guanine sites in every case. The time required to reduce the intact DNA

to 60o/o is 30 min for 1, 7 h for 2 and 5 h for 3. Of special interest is

that cleavage by 3 was highly sequence selective at 5' -side guanine of

GG sequence. The 5' -side G selectivity for 5' -GN sequence decreased in

the order of 5' -G*G > G* A > G*T > G*C. It is worthwhile to note that

3 well exceeded natural kapurimycin A3 (1) in terms of GG sequence

selectivity. As a result of this experiment, it was confirmed that both

DNA cleaving activity and the sequence selectivity considerably

increased with increasing the number of the aromatic rings of their

synthetic drugs.

83

T ·· ... T ··· ... A . G T T G Ci T T

1 2

.. "

relative O.D. (%)

60

40

A: kapurimycin A3 (1 ), 30 min l B: (S, S)-ABC 2, 7 h

C: (S, S)-ABCD' 3, 5 h C

3

-< + 0

2

: - ' fl. " L ' - .:L rll .... - ' 5'- TGTT T GT T AGTT CGTT TGCT T G GTT GAT TOT TT GTorigin

Figure 2. Strand breakage assa~s depicting reactive sites for 1, 2 and 3 on the oligodeoxynucleotide duplex. 2P-5'-labeled oligodeoxynucleotide was treated with 1, 2 or 3 (50 J1M) in a reaction buffer (20 mM Tris-HCl, pH 7.6 and 10 ,UM calf thymus DNA) at 37 °C. After ethanol precipitation, the residue was heated in 10% piperidine at 90 OC for 30 min. The resulting DNAs were analyzed by electrophoresis on 15% denatured polyacrylamide gel. The height of bars in this histogram shows the percentage of strand breakage at a given site relative to the total strand breakage. The cleaving sites of DNA treated with 1 fm 30 min is shown by white bars; 2 for 7 h is shown by gray bars; 3 for 5 h is shown by black bars.

84

J

(a)

2 3 4

1

5 6

3

7 8

ongm (b)

I 30 (0.0)

G

f G T G G T T G T

G

G T

T

5'

20

10

0

Lane 7

Lane 8

5' T T G G T G T T G GT GT T G A T ongm

Figure 3. Cleavage of DNA containing 7-deazaguanine by 1 and 3. The oligodeoxynucleotide containing 7 -deazaguanine was prepared by elongation of primer in presence of exo- Klenow fragment. DNA cleaving assay was applied with 50 J1M of drug in Tris-HCl (pH 7.6) at 37 oc for 24 h. After ethanol precipitation, the recovered DNA was treated with 10% piperidine solution. (a) The result of polyacrylamide gel electrophoresis. lane 1, G+A sequencing reaction; lane 2, elongated DNA containing 7-deazaguanine sites (ODNl); lane 3, DNA with guanine instead of 7-deazaguanine (ODN2); lane 4, primer; lane 5, ODNl treated with 1; lane 6, ODN2 treated with 1; lane 7, ODNl treated with 3; lane 8, ODN2 treared with 3. The 0 in figure shows 7 -deazaguanine in ODN 1 or guanine in ODN2. Solid line in figure shows the primer part before elongation. (b) The densitometric analysis of lane 7 and lane 8 in (a). Lane 7 and lane 8 were shown as black bars and white bars, respectively.

We examined the DNA cleaving assay for 1 and 3 by the use of

oligonucleotide containing 7-deazaguanine in order to confirm whether

1 or 3 modifies guanine N7 -position. Oligodeoxynucleotides having 7-

deazaguanine were prepared by primer elongation method. The primer

was elongated by exo--Klenow fragment in presence of dATP, dCTP,

dTTP and deaza-dGTP. The result obtained from the cleaving assay of

DNA involving 7-deazaguanine bases was shown in Figure 3. As clear

from figures, the DNA cleavage never occurred at deaza G site. The

result of figures clearly shows that the cleavage bands were generated

85

by the alkylation of guanine N7. Drug 3 binds to DNA and efficiently

alkylate guanine N7 in the major groove. We next examined the DNA binding ability of the aromatic ring of

kapurimycin analogs 13, 14 and 15. We prepared compounds 13, 14 and 15 which are lacking in epoxide side chain, and examined unwinding assay of supercoiled DNA using topoisomerase I. Topoisomerase I is allowed to adjust DNA linking number in the presence of an intercalator that influences the DNA unwinding angle, resulting in the shift of the gaussian distribution of topoisomers. A solution of pBR322 plasmid DNA with each drug was incubated 1n presence of human topoisomerase I at 37 oc and the resulting DNAs were analyzed by agarose gel electrophoresis (Figure 4). As clear from Figure 4, the DNA treated with 15 was more strongly unwound than that treated with 14. Only little unwinding was observed for DNA treated with 13. This result clearly indicates that the aromatic moiety of kapurimycin analog 3 intercalates into DNA and that the binding ability of these drugs increases with increasing the number of aromatic

ring.

9~H ~0

13 14

OH

0

15

86

open-circular

supercoiled

12345678 ~~~~~~

control 13 14 15

open-circular DNA

(0.0 .) J " ' '·.) l} '\ ( \ l

. ....... • · ' ......... .... 'l. ... "'

. , ! J \ / ) , ... . ,1 ·•' \ ' ; fi I ' '. . . ' .•. ...,.

, ..... /'~·'\/,/\ .\r, I,

' \ (\ / l I l ·- .J _' ·.,.,I , I • ' J ,,

15 (lane 8)

14 (lane 6)

13 (lane 4) control

(lane 2) 0 I~

distance from the bands of open-circular DNA (mm)

Figure 4. Unwinding of supercoiled pBR322 DNA by topoisomerase I in the presence of kapurimycin analogs 13, 14 and 15. Supercoiled pBR322 DNA (250 ng) was first treated with topoisomerase I (topo I) for 30 min in a reaction buffer (10 mM Tris-HCl, pH 7.9, 1 mM EDTA, 150 mM NaCl, 0.1 mM spermidine, 5% glycerol, 0.1% BSA), and further incubated for 30 min in the absence (lane 2) and presence of 13, 14 and 15 ( 10 ,uM) (lanes 4, 6 and 8, respectively). The resulting DNAs were analyzed by electrophoresis on 1% native agarose gel at 1. 3 V /em for 18 h. D N As in lanes 1, 3, 5 and 7 were treated as for lanes 2, 4, 6 and 8, respectively, in the absence of topo I. lane 1, intact DNA; lane 2, topo I; lane 3, 13; lane 4, 13 with topo I; lane 5, 14; lane 6, 14 with topo I; lane 7, 15; lane 8, 15 with topo I.

In order to gain insight into the mechanism of GG selective DNA alkylation by ABCD' -ring analog 3, we have examined the guanine cleavage reaction of duplex oligodeoxynucleotide (ODN) having a bulge site opposite to guanine doublet and guanine triplet of the target strand (Figure 5). In DNA, the stabilization of bulge structures by intercalating agents has been suggested as a mutagenesis mechanism, and direct measurements of ethidium binding to synthetic DNAs support this conclusion. 8 Specific strand scission near a bulge can be monitored to determine if intercalation is the mode of binding of any drug that cleaves DNA. As clear from the Figure 5, the most effective cleavage site of ODNl is highly dependent on the location of the bulge site, with considerably increased alkylation selectivity at 5' -G of the guanine doublet, i.e., 5'-GG-3'/3'-CTC-5'. Since bulge site is well known to be the favorable site for intercalation, it is highly likely that ABCD' -ring

87

analog 3 intercalates two guanine bases of GG doublet and alkylates

selectively the 5' -side G.

2 3 4 5 lane 4

r'. T '

~ T G 23

G 22

t~ 21

G11 T

b G G T

5'- T

Conclusion

(ODN4)

(ODN 3) (ODN 5) lane 3 llane 5

10 15 t 20 t 25

ODN I 5' ... T TGGGT TG GTI G G GTI GGTI ... 3'

ODN 2 3' ... A ACCCA AC CAA C C CAA CCAA ... 5' (lane 2)

ODN 3 3' ... A ACCCA ACTCAA C C CAA CCAA ... 5' (lane 3)

ODN 4 3' ... A ACCCA AC C AA CTC CAA CCAA ... 5' (lane 4)

ODN 5 3' .. . A ACCCA AC CAA C C"ilCAA CCAA. .. 5' (lane 5)

Figure 5. Determination of alkylation site of the target 35-mer (ODN 1) by 3 in the presence of complementary strand ODN 2, 3, 4 and 5. Upon duplex formation with ODN 1, ODN 3, 4 and 5 produce the bulge structure at the complementary se~uence of G

17G

18, G21 G22 and G22G23 of ODN 1. 3 P-5'-end

labeled ODN 1 annealed with a complementary ODN was treated with 3 (50 ,UM). lane 1, G+A sequencing reaction; lane 2, ODN 2; lane 3, ODN 3; lane 4, ODN 4; lane 5, ODN 5.

We have focused on the contribution of the aromatic ring of DNA

alkylating antibiotics to the sequence selectivity for guanine alkylation,

and compared the DNA cleavage by kapurimycin A3, ABC-ring analog

2 and ABCD' -ring analog 3. ABCD' -ring analog 3 exceeded ABC-ring

analog 2 in both the reactivity and the sequence selectivity of DNA

cleavage (G*G > G* A > G*T > G*C). It was elucidated by the

experiments shown above that the intercalation of aromatic ring system

88

of 3 into DNA is a very important factor for effective GG selective

DNA alkylation. Investigation of the mechanism of DNA cleavage by

synthetic analog 3 enables us to explain why the GG selective DNA

alkylation occurs with naturally occurred antibiotics.

89



JNM a-400 (400 MHz) spectrometers. Coupling constants (J values) are

reported in Hz. The chemical shifts are expressed in ppm downfield


NMR) as an internal standard. The following abbreviations were used

to explain the multiplicities: s, singlet; d, doublet; t, triplet; q, quartet;

m, multiplet; br, broad. IR spectra were recorded on JASCO FT!IR-5M

spectrophotometers. Melting points were obtained on a Yanagimoto

Seisakusho micro melting point apparatus and uncorrected. Optical

rotations were recorded using a Perkin-Elmer 243 polarimeter.

Electron impact mass spectra (MS) and high-resolution mass spectra

(HRMS) were recorded on JEOL JMS-DX 300 or JEOL JMS-SX 102A.

Microanalyses were performed at Kyoto University Microanalytical

Center.



5o/o ethanolic phosphomolybdic acid, or p-anisaldehyde solution and heat

as developing agent. Wako gel (C-200, particle size 75-150 J.lm, Wako)



over sodium-benzophenone. Dichloromethane, toluene and N, N





anhydrou solvents under anhydrous conditions unless otherwise noted.

90


homogeneous materials unless otherwise stated.

1-Methoxymethoxy-9,10-anthraquinone (5). A suspension of 4

(6.0 g, 26.7 mmol) in chloroform (80 mL) was treated with N,N

diisopropylethylamine (41.9 mL, 240 mmol) and chloromethyl methyl

ether (12.2 mL, 160 mmol) at ambient temperature and subsequently

heated to reflux for 1 h. The mixture was allowed to cool to ambient

temperature and washed with IN aqueous sodium hydroxide for 1 h and

extracted with chloroform. The organic phase was concentrated and

then the resulting solid was washed successively with 1 N sodium

hydroxide, water and then ethanol. Two recrystallizations from ethyl

acetate produced 5 (6.87 g, 25.6 mmol, 96%) as a yellow solid: 1H

NMR (CDC13, 400 MHz) 8 8.26-8.22 (2H), 8.02 (dd, lH, J = 7.7, 1.1

Hz), 7.79-7.70 (2H), 7.68 (dd, lH, J = 8.4, 7.7 Hz), 7.56 (dd, lH, J = 8.4, 1.1 Hz), 5.39 (s, 2H), 3.56 (s, 3H); IR (CHC13) 1673, 1586, 1211,

769 cm-1; MS (EI) mle (%) 268 (M+, 49), 237 (87), 208 (100), 180

(46), 151 (38), 139 (61); HRMS (EI) calcd for C16H120 4 (M+) 268.0735,

found 268.0722; Anal. Calcd for C 16H 120 4: C, 71.64; H, 4.51. Found: C,

71.62; H, 4.63.

1-Methoxymethoxyanthracene ( 6). To a suspension of 5 (1.67 g,

6.2 mmol) in 2-propanol (30 mL) was added sodium borohydride (7 .0

g, 187 mmol) at 0 °C. The reaction mixture was heated to reflux for 18

h, and treated with 2N hydrochloric acid at 0 oc until the pH of the

mixture became 4-6. The mixture was extracted with ethyl acetate.

The crude product was purified by column chromatography on silica

gel, eluting with toluene to give 6 (1.29 g, 5.4 mmol, 87%) as a yellow

solid: 1H NMR (CDC13, 400 MHz) 8 8.83 (s, 1H), 8.36 (s, 1H), 8.04-

91

7.96 (2H), 7.63 (d, 1H, J = 8.6 Hz), 7.47-7.42 (2H), 7.34 (dd, 1H, J = 8 .6, 7.5 Hz), 7.00 ( d, 1 H, J = 7.5 Hz), 5.46 (s, 2H), 3.58 (s, 3H); IR

(CHC13

) 3056, 1221, 1141, 1045 cm-1;MS (EI) m/e (o/o) 238 (M+, 100),

208 (53), 193 (26), 178 (35), 165 (100), 163 (21); HRMS (El) calcd for

C 16H140 2 (M+) 238.0994, found 238.0971; Anal. Ca1cd for C16Ht402: C,

80.65; H, 5.84. Found: C, 80.55; H, 5.92.

1-Methoxymethoxy-2-anthraldehyde (7). To a solution of 6 (1.99

g, 8.35 mmol) in tetrahydrofuran (30 mL) was added n-butyl lithium

( 1.6M solution in hexane, 10.5 mL, 16.8 mmol) at 0 oc, and the mixture

was stirred for 1 h at 0 °C. To the mixture was added N,N

dimethylformamide (1.94 mL, 25.0 mmol) at 0 oc and stirred at 0 oc for 1 h. The mixture was diluted with sat. aq. NH4Cl at 0 oc, and

extracted with ethy 1 acetate. The crude product was purified by column

chromatography on silica gel, eluting with 10o/o ethyl acetate in hexane

to give 7 (1.78 g, 6.70 mmol, 80o/o) as a yellow solid: mp. 108-110 oc; 1H NMR (CDC1

3, 400 MHz) 8 10.56 (s, 1H), 8.81 (s, 1H), 8.42 (s, 1H),

8.08-8.00 (2H), 7.82-7.80 (2H), 7.58-7.51 (2H), 5.38 (s, 2H), 3.70 (s,

3H); IR (CHC13

) 1675, 1621, 1235, 923, 890 cm-1; MS (El) m/e (o/o) 266

(M+, 60), 236 (20), 220 (100), 206 (39), 199 (25), 165 (80), 164 (40);

HRMS (EI) calcd for C17

H140 3 (M+) 266.0943, found 266.0952; Anal.

Calcd for C17

H 140 3: C, 76.68; H, 5.30. Found: C, 76.21; H, 5.38.

1-Hydroxy-2-anthraldehyde (8). To a solution of 7 (0.70 g, 2.6

mmol) in methanol ( 40 mL) was added cone. hydrochloric acid (0.50

mL) at ambient temperature, and the reaction mixture was stirred for 4

h. The mixture was diluted with sat. aq. NaHC03 and extracted with

ethyl acetate. The crude product was purified by column

chromatography on silica gel, eluting with 5o/o ethyl acetate in hexane to

92

give 8 (0.54 g, 2.4 mmol, 93o/o) as a yellow solid: mp. 138-140 °C; 1H

NMR (CDC13, 400 MHz) 8 13.29 (s, 1H), 9.95 (s, lH), 9.07 (s, 1 H),

8.30 (s, 1H), 8.08 (d, 1H, J = 8.4 Hz), 7.98 (d, 1H, J = 8.2 Hz), 7.59-

7.48 (3H), 7.41 (d, 1H, 1= 9.0 Hz); IR (CHC13) 3550-3100, 1621, 1210,

735, 667 cm-1; MS (EI) m/e (%) 222 (M+, 100), 221 (20), 165 (50), 164

(20); HRMS (EI) calcd for C 15H100 2 (M+) 222.0681, found 222.0688;

Anal. Calcd for C 15H100 2: C, 81.06; H, 4.53. Found: C, 80.76; H,

4.278.08.

1-(tert-B utyldimethylsilyloxy)-2-anthraldehyde (9). To a

solution of 8 (432 mg, 1.94 mmol) and 2,6-lutidine (0.45 mL, 3.88

mmol) in dichloromethane (15 mL) was added tert-butyldimethylsilyl

trifluoromethanesulfonate (0.89 mL, 3.88 mmol) at -78 °C, and the

reaction mixture was stirred at -78 oc for 2 h. After diluted with sat.

aq. NaHC03 at -78 °C, the mixture was warmed to ambient temperature

and extracted with ethyl acetate. Short column chromatography on

silica gel with 5o/o ethyl acetate in hexane gave crude product 9 (0.57 g)

as an orange solid. This crude product was used for next reaction

without purification.

( 4R ,SR, 1 'Z)-4-{ 3- [ 1-(tert-B u tyldimethylsily loxy )-2-an thry I]-

3-hydroxy-1-propynyl}-5-(1 '-propenyl)-2,2,4-trimethyl-1,3-

dioxorane (11). To a solution of 10 (99.2 mg, 0.55 mmol) and

anhydrous cerium chloride (411 mg, 1.7 mmol) in tetrahydrofuran (5

mL) being stirred at ambient temperature for 10 min was added lithium

hexadisilazide (1.0 M in tetrahydrofuran, 1.1 mL, 1.1 mmol) at -78 °C,

and the reaction mixture was stirred at -78 °C for 15 min. After

addition of a solution of 9 (187 mg) in tetrahydrofuran (1 mL) at -78

°C, the reaction mixture was stirred at -78 oc for 15 min. It was

93

diluted with sat. aq. NH4Cl at -78 °C, and extracted with ethyl acetate.

The crude product was purified by column chromatography on silica

gel, eluting with 35o/o toluene in hexane to give 11 (98.8 mg, 0.19

mmol, 35o/o) as a yellow oil: 1H NMR (CDC13, 400 MHz) 8 8.67 (s, 1H),

8.34 (s, 1H), 7.99, 7.96 (sx2, total 2H), 7.73-7.63 (2H), 7.50-7.43 (2H),

6.06-6.04 (lH), 5.83-5.75 (lH), 5.42-5.36 (lH), 5.07, 5.03 (ddx2, total

lH, J = 8.6, 1.1 Hz), 2.16-2.12 (lH), 1.70, 1.60 (ddx2, total 3H, J =

7 .0, 1.7 Hz), 1.48-1.43 (6H), 1.37, 1.35 (sx2, total 3H), 1.20, 1.19 (sx2,

total 9H), 0.24-0.22 (6H); IR (CHC13) 3571, 3019, 1216 cm-1

; MS (EI)

m/e (o/o) 516 (M+, 2), 222 (70), 75 (100).

( 4S ,SR ,1 'Z)-4-( 4H -anthra[1,2-b ]pyran-2-yl)-5-(1 '-propenyl)-

2,2,4-trimethyl-1,3-dioxorane ( 12). To a solution of 11 (56.1

mg, 109 JLmol) in dichloromethane (3 mL) was added manganese

dioxide (50 mg), and the reaction mixture was stirred for 3 h at ambient

temperature. The mixture was diluted with ethyl ether, filtered, and

concentrated in vacuo. The crude product was purified by column

chromatography on silica gel, eluting with 13o/o ethyl acetate in hexane

to give the corresponding ketone (38.7 mg, 75.2 JLmol, 69o/o) as a

yellow oil: MS (EI) m/e (%) 457 [(M-13ur, 10], 368 (18), 294 (37),

149 (100)~ HRMS (EI) calcd for C32H380 4Si [(M-13ur] 457.1836, found

457.1840. To solution of ketone (29.8 mg, 57.9 Jlmol) and 18-crown-6

(30.7 mg, 116 JLmol) in N,N-dimethylformamide (3 mL) was added

potassium fluoride (6.8 mg, 117 JLmol) at 0 °C, and the reaction mixture

was stirred at ambient temperature for 1 h. The mixture was diluted

with sat. aq. NH4Cl, and extracted with ethyl acetate. The crude product

was purified by column chromatography on silica gel, eluting with 13o/o

ethyl acetate in hexane to give 12 (17.3 mg, 43.2 JLmo1, 75o/o) as a

yellow oil: 1H NMR (CDC13, 400 MHz) 8 8.84 (s, lH), 8.46 (s, 1H),

94

J

8.08-8.04 (3H), 7.87 (d, lH, J = 9.0 Hz), 7.63-7.56 (2H), 6.85 (s, 1H),

6.12-6.04 (m, lH), 5.85-5.78 (m, 1H), 5.10 (d, 1H, J = 10.0 Hz), 1.68

(s, 3H), 1.63 (s, 3H), 1.56 (s, 3H), 1.43 (dd, 3H, J = 7 .0, 1.8 Hz)~ IR

(CHC13) 3263, 2995, 1650, 1386, 1105 cm-1; MS (EI) mle (%) 400 (M+,

59), 289 (100), 288 (75), 149 (43); HRMS (EI) calcd for C26

H24

04

(M+)

400.1674, found 400.1658.

2-[ (2S ,3S ,4Z)-2,3-Epoxy-4-hexen-2-yi]-4H -anthra[1,2-b]

pyran (3). To solution of 12 (14.4 mg, 36.0 JLmol) in tetrahydrofuran

(1 mL) and acetic acid (1 mL) was added 0.2 M hydrochloric acid (0.1

mL) at 0 °C, and the reaction mixture was stirred at ambient

temperature for 3 days. The mixture was poured onto sat. aq. NaHC01

and ethyl acetate, and extracted. The crude product was purified by

column chromatography on silica gel, eluting with 50% ethyl acetate in

hexane to give the corresponding diol (11.8 mg, 32.7 Jlmol, 91 %) as a

white solid: 1H NMR (CDC13, 400 MHz) 8 8.92 (s, 1H), 8.39 (s, 1H),

8.12-8.04 (2H), 7.90 (d, 1H, J = 9.0 Hz), 7.72 (d, 1H, J = 9.0 Hz), 7.64-

7.57 (2H), 6.82 (s, 1H), 5.93-5.85 (m, 1H), 5.74-5.67 (m, 1H), 5.13 (d,

1H, J = 9.3 Hz), 3.34 (br, 1H), 2.53 (br, 1H), 1.80 (dd, 3H, J = 7.0, 1.8

Hz), 1.64 (s, 3H); MS (FAB) (NBA) m/e 361 [(M+Ht]; HRMS (FAB)

calcd for C23H21 0 4 [(M+Hr] 361.1440, found 361.1446. To a solution

of the diol (6.9 mg, 19.1 Jimol) and triphenylphosphine (7.0 mg, 26.7

Jlmol) in toluene (1 mL) was added diethyl azadicarboxilic acid (6 JLL,

37.9 Jlmol) at 0 °C, and the reaction mixture was stirred at ambient

temperature for 24 h. The mixture was concentrated in vacuo and

purified by column chromatography on silica gel, eluting with 13 o/o

toluene in hexane to give 3 ( 4.1 mg, 12.0 JLmol, 63o/o) as a white solid: 1H NMR (CDC13, 400 MHz) 8 9.00 (s, lH), 8.47 (s, lH), 8.14 (m, 1H),

8.08-8.04 (2H), 7.88 (d, lH, J = 9.0 Hz), 7.63-7.56 (2H), 6.61 (s, 1H),

95

5.81 (ddd, 1H,J= 11.2, 7.1, 1.1 Hz), 5.18 (dq, 1H, 1= 8.6, 1.7 Hz),

3.95 (d, 1H, J = 8.6 Hz), 1.96 (s, 3H), 1.85 (dd, 3H, J = 7.1, 1.7 Hz); IR

(CHC13) 3129, 1651, 1223, 909 cm-1

; MS (EI) m/e (o/o) 342 (M+, 100),

300 (33), 220 (76); HRMS (EI) calcd for C23H 180 3 (M+) 342.1256, found

342.1245.

Cleavage of pBR322 Supercoiled DNA by 2 and 3 ·

Kapurimycin ABC analog 2 or ABCD' analog 3 (1 0 pM or 1 pM) was

incubated with 40 pM pBR322 supercoiled DNA (Nippon Gene) in 10o/o

(v/v) acetonitrile/50 mM Tris-HCl buffer (10 pL, pH 7 .5) at 37 oc for 5

h. The samples were successively precipitated with ethanol to remove

the drug and the recovered DNA pellets were dried and incubated again

in 10 pL of 50 mM Tris-HCl buffer at 37 oc for 24 h. To heat treated

samples was added 10 pL of loading buffer containing 0.05% (w/v)

bromophenol blue and 6o/o (v/v) glycerol for electrophoresis. Different

forms of DNA were separated at ambient temperature on a 1% (w/v)

agarose gel invoving ethidium bromide (0.5 pg/mL). The gels were

placed on a UV transilluminator (313 nm) and photographed with


Preparation of 32P-5'-End-Labeled Oligodeoxynucletide

Duplex. 400 pmol of single-strand oligodeoxynucleotide, purchased

from Greiner Japan Co. Ltd. was 5' -end-labeled by phosphorylation

with 4 pL of [y-32P]ATP (Amersham, 370 MBq/,uL) and 4 pL T4

polynucleotide kinase (Takara, 10 units/pL) using standard procedures.

The 5' -end-labeled DNA was recovered by ethanol precipitation and

further purified by 15% nondenatured gel electrophoresis and isolated

by the crush and soak method.9 The isolated DNA was incubated with

96

equimolar of complementary DNA in 100 pL of water at 90 oc for 5

min and cooled slowly to ambient temperature for forming the duplex.

Cleavage of 32P-5'-End-Labeled Oligodeoxynucleotide. Single

stranded 43-mer DNA oligomers 5' -d(TTTTTGTTTGTT AGTTC

GTTTGCTTGGTTGA TTGTTTGTTTTT)-3' and the corresponding

complementary oligomer were purchased fron1 Greiner Japan Co. Ltd.

The 32P-5' -end-labeled ODN duplex was prepared as shown above. 50

pM of 1, 2 or 3 was incubated with 10 pM of calf thymus DNA and ca.

l.Ox106 cpm 32P-5' -end-labeled ODN duplex in 20 mM Tris-HCl buffer

(100 pL, pH 7.6) at 37 °C. At each time, the sample (10 pL) was

separated from the reaction mixture, and precipitated with methanol.

The recovered DNA was dissolved in 100 pL of 10% (v/v) piperidine

and heated at 90 oc for 30 min. The mixture was concentrated in vacuo

and resuspended in 10 pL of 80% formamide loading buffer (80%

formamide, 1 mM EDT A, 0.1% xylene cyanole and 0.1% bromophenol

blue). The samples (1 ,uL) were loaded onto 15% polyacrylamide and 7




Preparation of Oligodeoxynucleotide with 7-Deazaguanine.

Both 16-mer DNA oligomer 5'-d(TTTTGCTGATTGGTGT)-3' as

primer for DNA elongation and the 32-mer DNA oligomer 5'-

d(AAAAGCA TCAACACCAACACCAA TCAGCAAAA)-3' as a

template for DNA elongation were purchased from Greiner Japan Co.

Ltd. Before oligomer elongation, the primer was 32P-end-labeled as

shown above and annealed with template oligomer. To a solution of the

annealed oligomer complex in reaction buffer (1 0 mM Tris-HCl, pH

97

7.3, 10 mM MgC12

, 1 mM DTT) was added dATP, dCTP, dTTP (1 mM,

Takara), 7-deaza-dGTP (1 mM, Boehringer Mannheim) and exo-

Klenow Fragment (5 units, Ambion). The reaction mixture was

incubated at 37 oc for 2 h. The reaction mixture was precipitated with

ethanol and purified by 20o/o nondenatured gel electrophoresis and

isolated by the crush and soak mothod.9

Cleavage of 32P-5' -End-Labeled Oligodeoxynucleotide with 7-

Deazaguanine Sites. 32P-5' -End-labeled ODN duplex was prepared as

shown above. Kapurimycin A3 (1) (5 pM) or its analog 3 (50 pM) was

06 32p 5' incubated with 10 pM of calf thymus DNA and ca. l.Ox1 cpm - -

end-labeled ODN duplex in 20 mM Tris-HCl buffer (100 J.LL, pH 7.6) at

37 oc for 24 h. The sample was precipitated with methanol and dried.

The recovered DNA was dissolved in 100 pL of 10% (v/v) piperidine


and resuspended in 10 pL of 80o/o formamide loading buffer (80%

formam ide, 1 mM EDT A, 0.1% xylene cyanole and 0.1% bromophenol

blue). The samples (1 pL) were loaded onto 15o/o polyacrylamide and 7




Supercoiled DNA Unwinding Assay. To a solution of pBR 322

plasmid DNA (250 ng, Nippon Gene) in topo I reaction buffer (10 mM

Tris-HCI, pH 7.9, 1 mM EDTA, 150 mM NaCl, 0.1 mM spermidine,

5% glycerol, 0.1 o/o BSA) was added human topoisomerase I (TopoGEN,

4 units). The reaction mixture was incubated for 30 min at 37°C. After

the drug ( 100 pM) was added, the incubation was continued for another

30 min at 37°C. The reaction was terminated by addition of SDS to 1 o/o.

98

After proteinase K was added to 50 pg/mL, the 1nixture was digested

for 20 min at 56°C. Addition of 0.1 vol. of 1 Ox gel loading buffer wa~

followed by chloroform extraction. Different forms of DNA were

separated at room temperature on a 1% agarose gel. The gel was

stained for 30 min with ethidium bromide (0.5 pM/mL) and destained

for 20 min in water. It was placed on a UV transilluminator (313 nm)

and photographed with Polaroid 665 film. The result was shown in

Figure 4.

Cleavage of 32P-5'-End-Labeled Oligodeoxynucleotide Duplex

with Bulge Sites. Each of single-stranded DNA oligomers in Figure

8 (ODNs 1-5) were purchased from Greiner Japan Co. Ltd. The 32P-5'

end-labeled ODN duplex was prepared as shown above. 50 pM of 3

was incubated with 10 pM of calf thymus DNA and ca. l.Ox106 cpm 32P-

5' -end-labeled ODN duplex with or without the bulge site in 20 mM

Tris-HCl buffer (100 pL, pH 7.6) at 37 oc for 5 h. The sample was

precipitated with methanol and dried. The recovered DNA was

dissolved in 100 pL of 10o/o (v/v) piperidine and heated at 90 oc for 30

min. The mixture was concentrated in vacuo and resuspended in 10 pL

of 80% formamide loading buffer (80o/o formam ide, 1 mM EDT A,

0.1% xylene cyanole and 0.1 o/o bromophenol blue). The samples (1 pL)

were loaded onto 15o/o polyacrylamide and 7 M urea sequence gel and

electrophoresesed at 1900 V for ca. 2 h. The gel was dried and exposed

to X-ray film with intensifying sheet at -70 °C. The result was shown

in Figure 5.

99

References

(1) Hartley, J. A. In Molecular Basis of Specificity in Nucleic Acid

Drug Interactions; Pullman, B., Jortner, J., Eds.; Kluwer Academic

Publishers: 1990, p 513-530.

(2) (a) Stone, M. P.; Gopalakrishnan, S.; Harris, T. M. ; Graves, D. E. J.


Byrd, S.; Stone, M. P. ; Harris, T. M. Biochemistry 1989, 28, 726-734.

(c) Stone, M. P. ; Gopalakrishnan, S.; Raney, K. D.; Raney, V. M.;

Byrd, S. ; Harris, T. M. In Molecular Basis of Specificity in Nucleic





M. P. Chern. Res. Toxicol. 1993 , 6, 64-68. (f) Gopa1akrishnan, S.;

Harris, T. M.; Stone, M.P. Biochemistry 1990, 29, 10438-10448.

(3) (a) Kupchan, S. M.; Streelman, D. R.; Sneden, A. T. J. Nat. Prod.

1980, 43, 296-301. (b) Habib, A. M.; Ho, D. K.; Masuda, S.;

McCloud, T.; Reddy, K. S.; Aboushoer, M.; McKenzie, A.; Byrn, S. R.;

Chang, C. J.; Cassady, J. M. J. Org. Chern. 1987, 52, 412-418. (c)

Hansen, M.; Lee, S.-J.; Cassady, J. M.; Hurley, L. H. J. Am. Chern. Soc.

1996, 118, 5553-5561.



Sano, H. J. Antibiot. 1990, 43, 1519-1523. (c) Hara, M.; Yoshida, M.;

Nakano, H. Biochemistry 1990, 29, 10449-10455. (d) Chan, K. L.;

Sugiyama, H.; Saito, I. Tetrahedron Lett. 1991, 52, 7719-7722. (e)

Chan, K. L.; Sugiyama, H.; Saito, I.; Hara, M. Phytochemistry 1995,

40, 1373- 1374.

100

(5) Refer to the previous chapter.

(6) (a) Nakatani, K.; Okamoto, A.; Yamanuki, M.; Saito, I. 1. Org.

Chern. 1994, 59, 4360-4361. (b) Nakatani, K.; Okamoto, A.; Saito, I.

Tetrahedron 1996, 52, 9427-9446.

(7) Mitsunobu, 0. Synthesis 1981, 1-28.

(8) (a) Nelson, J. W.; Tinoco, I. Jr. Biochemistry 1985, 24, 6416-6421.

(b) White, S. A.; Draper, D. E. Nucleic Acids Res. 1987, 15, 4049-

4064. (c) Williams, L. D.; Goldberg, I. H. Biochemistry 1988, 27,

3004-3011. (d) Caceres-Cortes, J.; Wang, A. H.-J. Biochemistry 1996,

35, 616-625. (e) Stassinopoulos, A.; Ji, J.; Gao, X.; Goldberg, 1. H.

Science 1996, 272, 1943-1946.


101

CHAPTER 4

Effect of Absolute Configuration of Epoxy Subunit on Guanine-Guanine Sequence Selective Alkylation

Abstract: We prepared two kapurimycin analogs, (S, S)-ABCD' 2 and

enantiomeric (R, R)-ABCD' ent-2, and compared their sequence

selectivity for DNA alkylation. (S, S)-Analog 2 exhibited a high

sequence selectivity (G*G > G* A > G*T > G*C), whereas (R, R)-analog

ent-2 showed a lower reactivity toward DNA than (S, S)-isomer 2 and

equally reacted at all guanine sites without sequence selectivity of its

alkylation. DNA unwinding assay demonstrated that both analogs have

very similar binding ability. The model of drug-DNA complex showed

that the stereochemical orientation of the epoxy subunit considerably

affects the reactivity toward DNA as well as the sequence selectivity for

DNA alkylation.

103

Introduction

The study of interactions of DNA with ligands of differing

chiralities provides a rich ground for applications to mechanisms of

molecular recognition, probing of polymorphism of DNA, and

structure-biological activity correlations. Comparing the DNA

interaction chemistry between individual members of enentiomeric or

diastereomeric pairs of chiral metal complexes, 1 antitumor agents, and

carcinogens has provided sharp insights into such phenomena. Recent

notable examples are the alkylation of DNA by the stereoisomers of the

carcinogenic benzo[a]pyrenediol epoxides (BPDE)/ the natural and

unnatural enantiomers of CC-I 065, and enantiomeric pairs of CC-1 065

functional derivatives3 and of duocarmycin SA,4 antibiotics mitomycin

C5 and staurosporin6 as well as the enantiospecific recognition of DNA

for cleavage by bleomycin.7 NMR studies of the DNA adducts of the

BPDE enantiomers ( + )- and (-)-anti-BPDE revealed an extraordinary

relationship among the stereochemistry of the drug-DNA linkage,

adduct orientation in the minor groove, and tumorigenicity. The two

enantiomers of CC-I 065 were shown to be potent cytotoxins, which

however displayed distinct DNA sequence selectivities and opposite

groove orientation of the adducts. 3·5

An antitumor antibiotic kapurimycin A3 (1) 8 is known to selectively

alkylate guanine N7 of guanine-guanine (GG) sequence. In the previous

study,9 using the analog of 1, 2, we investigated the relationship between

the aromatic ring system and the sequence selectivity for DNA

alkylation and elucidated that the intercalation of the planar aromatic

portion into two guanine bases of GG sequence was very important for

their GG elective alkylation. The binding of the aromatic ring system

to DNA facilitates the approach of the epoxy subunit close to guanine

104

N7. Thus, the sequence selectivity of DNA binding directly reflects the

sequence selective DNA alkylation. The guanine N7 alkylation step by

the epoxy group also seems to be important for the sequence selective

alkylation.

The use of the enantiomer of the DNA binding antibiotic was the

promissing methods to get insight into the action mechanisn1 of

antibiotics with DNA and provided very useful information on the

structure of drug-DNA complex and the subsequent alkylation. In order

to understand the mechanism of sequence selectivitive guanine

alkylation, we prepared the (S, S)-ABCD' analog 2 having the same

absolute configuration of the epoxy subunit as that of natural

kapurimycin A3, and the enantiomer (R, R)-ABCD' analog ent-2, and

compared their sequence selectivities for guanine alkylation. Herein, we

report the relationship between the absolute configuration of the epoxy

subunit and the sequence selectivity for guanine alkylation.

kapurimycin A3 (1) (S, S)-ABCD' 2 (R, R)-ABCD' ent-2


(R, R)-ABCD' analog ent-2 with an opposite configuration to (S,

S)-ABCD' analog 2 was prepared from optically active side chain unit 3

as described previously. 9 Asymmetric dihydroxylation of p-

105

methoxybenzyl ether 4 obtains from a commercially available (E)-3-

methyl-2-penten-4-yn-1-ol with AD-mix-a10 produced optically active

diol 5 which was converted to alcohol 3 (91 o/o e.e.). We optically

purified alcohol 3 by means of optical resolution. In the purification of

3, we condensed 3 with (-)-(S)-2-methoxy-2-trifluoromethylphenyl

acetic (MTPA) acid and removed the undesired isomer by use of HPLC

with chiral column (Scheme 1 and Figure 1). Reduction of collected

MTPA ester by DIBAL gave optically pure 3 again (>99% e.e.). (R,

R)-ABCD' analog ent-2 was synthesized from optically pure 3

according to the synthetic route of 2.

Scheme 1.

OPMB OH

{ AD-mix-a rOPMB _re_f. 4__..

4 5

Yo Yo ~0-(S)-MTPA _D_IB_A_L ~OH _re_f._4_

ent-2

TMS TMS

3

106

Yo 0 = ~0-(S)-MTPA

TMS

Yo ~0-(S)-MTPA

TMS

_)

0 5 10 15 20 (rmn)

Figure 1. HPLC profile for purification of 3-(S)MTP A ester. The diastereomeric mixture was separated by HPLC on Daicel CHIRALCEL OJ (10x250 mm, elution with 2% hexane in 2-propanol at a flow rate of 1.0 mL/min).

In order to know the relationship between the epoxide absolute

configuration and the sequence selectivity, the reaction of DNA with 2

and ent-2 was examined using 32P-5' -end-labeled oligodeoxynucleotides

and analyzed by electrophoresis on denatured polyacrylamide gel

(Figure 2). The DNA cleavage by treatment with 2 or ent-2 followed

by hot piperidine treatment was observed at all guanine sites. The DNA

cleavage experiment indicated that (S, S)-ABCD' analog 2 alkylated

DNA much more effectively than (R, R)-isomer ent-2. Of special

interest is that the alkylation by 2 was highly selective for 5' -side

guanine of 5' -GG sequences (G*G > G* A > G*T > G*C), but ent- 2

alkylated at guanine with no sequence selectivity.

107

T ·· .. T ·· ...

A G T T G

-G T

relative 0 .0. (%) 60

2

A: (S, S)-ABCD' 2, 5 h

40 B: (R, R)-ABCD' ent-2, 24 h

ent-2

l 20 ' AL

0 } n n .n ..n .- ~ L t, 1~ Jl_ __ .,[L .. 5'- T GT T TGT T AGT T CGT T TGC TT G GT T GAT TGT T T G~ongm

Figure 2. Strand breakage assays de~icting reactive sites for 2 and ent-2 on the oligodeoxynucleotide duplex. 3 P-5'-end-labeled oligodeoxynucleotide previously prepared was treated with 2 or ent-2 (50 ,UM) in a reaction buffer (20 rnM Tris-HCl, pH 7.6 and 10 ,UM calf thymus DNA) at 37 OC. After ethanol precipitation, the residue was heated in 10% piperidine at 90 OC for 30 min. The resulting DNAs were analyzed by electrophoresis on 15% denatured polyacrylamide gel. The height of bars in this histogram showed the percentage of strand breakage at a given site relative to the total strand breakage. The cleaving sites of DNA treated with 2 for 5 his shown by black bars; ent-2 for 24 his shown by white bars.

Next, we examined the binding ability of kapurimycin analogs 2 and

ent-2 to DNA by means of DNA unwinding assay using topoisomerase

I. The solution of pBR322 supercoiled DNA treated with 2 or ent-2

was incubated with topoisomerase I at 37 oc for 30 min (Figure 3). The

DNAs treated with ent-2 exhibited closely similar mobility to those

with 2. It is indicated that ABCD' analog ent-2 of (R, R)-configuration

108

had almost same intercalative DNA binding ability as (S, S)-isomer 2.

Thus, the lack of reactivity and sequence selectivity observed for e nt- 2

is completely different from the DNA binding by its planar aromatic

portion.

open-circular

supercoiled

I 2 3 4 5 6

control 2 ent-2

open-circular DNA

(0.0.) +

.... ...,.. ·-- ent-2 (lane 5)

_,.._. ..... ..,_ 2 (lane 3)

control (lane I)

0 2 4 6 8 10 distance from the bands of open-circular DNA (mm)

Figure 3. Unwinding of supercoiled pBR322 DNA by topoisomerase I in the presence of kapurimycin analogs 2 and ent-2 . Supercoiled pBR322 DNA (250 ng) was first treated with topoisomerase I (topo I) for 30 min in a reaction buffer (10 mM Tris-HCl, pH 7.9, 1 mM EDTA, 150 rnM NaCl, 0.1 mM spermidine, 5% glycerol, 0.1% BSA), and further incubated for 30 min in the absence (lane 1) and presence of 2 and ent-2 (10 ,UM) (lanes 3 and 5, respectively). The resulting DNAs were analyzed by electrophoresis on 1% native agarose gel at 1.3 V/cm for 18 h. DNAs in lanes 2, 4 and 6 were treated as for lanes 1, 3 and 5, respectively, in the absence of topo I. lane 1, topo I; lane 2, intact DNA; lane 3, 2 with topo I; lane 4, 2; lane 5, ent-2 with topo I; lane 6, ent-2.

(a) (b) (c)

Figure 4. DNA-drug binding models. (a) exo-aflatoxin B 1 oxide-GG/CC complex; (b) endo-aflatoxin B 1 oxide-GG/CC complex; (c) (S, S)-ABCD' 2-GG/CC complex. These models were obtained from optimization of drug-5'd(TGGT)-3 '/5'-d(ACCA)-3' complex by Amber* force field in water by mean of Macromodel. DNA backbone and A/C base pairs were removed and guanine N9 and cytosine N 1 was substituted by methyl group, respectively.

109

The (S, S)-analog 2 reacts efficiently with DNA, whereas (R, R)

analog ent-2 has only little reactivity toward DNA in spite of its

effective DNA binding. A rational reason for the different efficiency of

their DNA alkylation was proposed in the DNA alkylation by aflatoxin

B 1

oxide. 11 Carcinogen aflatoxin B 1 is converted to biologically active

form by chemical oxidation with dimethyldioxirane or by enzymatically

by cytochrome P450 mixed-function oxidases. Both processes gave rise

to a mixture of the exo- and endo-8,9-epoxides of aflatoxin B 1• While

exo-aflatoxin B 1

oxide is a potent mutagen, endo isomer fails to react

with DNA. Harris et al. have previously proposed that the reaction of

the exo epoxide of aflatoxin B 1 with DNA involves an intercalated

transition state. 11j · l They have obtained a number of lines of evidence

which indicate that aflatoxin B 1 and many of its derivatives strongly

associate with DNA and that the association involves intercalation of the

planer portion of the aflatoxin moiety into the DNA. According to their

proposal, we built a binding model of aflatoxin B 1 oxide with DNA in

order to rationalize the remarkable difference in the reactivity of the

aflatoxin B 1 oxide stereoisomers with DNA (Figure 4). 12 Whereas the

intercalation of the planar portion of exo-aflatoxin epoxide suitably

oriented for the oxirane moiety for SN2 attack to the epoxy group

(Figure 4a), a similar intercalation of the endo epoxide places the

epoxide ring inaccessible for SN2 attack (Figure 4b ). The binding model

of kapurimycin (S, S)-ABCD' analog 2 with DNA also showed that the

orientation of 2 is highly suitable for backside attack of guanine N7 on

the epoxide (Figure 4c). Furthermore, the distance from epoxide

carbon of 2 to 5' -side guanine N7 of the intercalation site is only ca. 2.9

A 11 and close enough for SN2 reaction. Thus, when (S, S)-ABCD'

analog 2 intercalates into DNA, the epoxide is oriented at the favorable

site for SN2 attack by guanine N7. However, (R, R)-isomer ent-2 was

110

less reactive and less sequence selective than (S, S)-isomer 2. According

to these models, the lack of the reactivity and the sequence selectivity in

ent-2 would indicate that the intercalation mode of the aromatic planar

portion of ent-2 is not optimally suited for SN2 attack of guanine N7 to

the epoxide.

Conclusion

We have synthesized kapurimycin analogs, (S, S)-ABCD' 2 and its

enantiomeric (R, R)-ABCD' ent-2 and compared their sequence

selectivities for DNA alkylation. While (S, S)-analog 2 showed a higher

reactivity and the sequence selectivity for the guanine alkylation, (R, R)

isomer ent-2 reacted with DNA more slowly with no sequence

selectivity. When (S, S)-ABCD' analog 2 intercalates into DNA, the

orientation of (S, S)-epoxide is optimally suited for the SN2 attack from

the back side of the epoxide by 5' -side guanine N7 at its intercalation

site. The (S, S) of epoxide showed 5' -side guanine selectivity in the

alkylation of GG sequence. The mechanistic study of sequence selective

DNA alkylation using (S, S)-ABCD' analog 2 would be very useful for

explaining the mechanism of the GG sequence selective alkylation by an

antibiotic kapurimycin A3 as well as 3 '-side guanine selectivity in GG

selective alkylation by aflatoxin B 1 oxide and psorospermin.

1 11



JNM a-400 (400 MHz) spectrometers. Coupling constants (1 values) are

reported in Hz. The chemical shifts are expressed in ppm downfield


NMR) as an internal standard. The following abbreviations were used

to explain the multiplicities: s, singlet; d, doublet; t, triplet; q, quartet;

m, multiplet; br, broad. IR spectra were recorded on JASCO FT/IR-5M

spectrophotometers. Melting points were obtained on a Yanagimoto

Seisakusho micro melting point apparatus and are uncorrected. Optical




Microanalyses were performed by Kyoto University Microanalytical

Center.



5o/o ethanolic phosphomolybdic acid, or p-anisaldehyde solution and heat

as developing agent. Wako gel (C-200, particle size 75-150 J.lm, Wako)



over sodium-benzophenone. Dichloromethane, toluene and N,N






112



Purification of Alcohol 3 for the Synthesis of (R, R)-ABCD'

Analog ent-2. To a solution of alcohol 3 (196 mg, 0.92 mmol) in

dichloromethane (5 mL) was added (-)-(S)-2-methoxy-2-trifluoro

methylphenylacetic (MTPA) acid (215 mg, 0.92 mmol), dicyclohexyl

carbodiimide (191 mg, 0.92 mmol) and N,N-dimethylaminopyridine

(12.0 mg, 0.098 mmol) at ambient temperature, and the resulting

solution mixture was stirred at ambient temperature for 1 h. The

mixture was diluted with ethyl acetate, filtered and concentrated in

vacuo. The crude product was purified by column chromatography on

silica gel, eluting with toluene to give diastereomeric mixture of MTPA

ester (354 mg). The mixture was separated by HPLC on Daicel

CHIRALCEL OJ (10x250 mm, elution with 2% hexane in 2-propanol at

a flow rate of 1.0 mL/min). The solution of given optically pure MTPA

ester in elution solvent was concentrated in vacuo. 1H NMR (CDCL~,

400 MHz) 8 7.55-7.53 (2 H), 7.42-7.37 (3 H), 4.47-4.37 (3 H), 3.55 (d,

3 H, 1 = 1.1 Hz), 1.44 (s, 3 H), 1.40 (s, 3 H), 1.36 (s, 3 H), 0.13 (s, 3 H).

Subsequently, to the solution of optically pure MTPA ester (1 01 mg,

0.22 mmol) in dichloromethane (2 mL) was added diisobutylaluminum

hydride (1.0 M in toluene, 0.5 mL, 0.5 mmol) at -78 °C, and the

solution was stirred at -78 oc for 15 min and at 0 °C for 15 min. The

mixture was diluted with methanol (1 0 mL) and stirred at ambient

temperature for 10 min, and further diluted with each 5 mL of ethyl

acetate and sat. aq. Rochelle salt and stirred at ambient temperature for

30 min. The mixture was extracted with ethyl acetate. The crude


with 30% ethyl acetate in hexane to give diastereomeric mixture of

113

MTPA ester (41.0 mg, 0.19 mmol): [a]25

0 = -19.17 (c 0.120, MeOH);

1H NMR (CDC13, 400 MHz) 8 4.31 (dd, 1 H, J = 3.9, 7.4 Hz), 3.73 (2

H), 1.73 (m, 1 H), 1.49 (s, 3 H), 1.43 (s, 3 H), 1.37 (s, 3 H), 0.14 (s, 9

H); IR (CHC13) 3020, 1217, 847, 754 cm-1; Anal. Calcd for C12H2203Si:

C, 59.46; H, 9.15. Found: C, 59.20; H, 9.30.

Preparation of 32P-5'-End-Labeled Oligodeoxynucleotide

Duplex. The 400 pmol of single-stranded oligodeoxynucleotide,

purchased from Greiner Japan Co. Ltd., was 5' -end-labeled by

phosphorylation with 4 f.1L of [ y-12P]ATP (Amersham, 370 MBq/ f.1L)

and 4 f.1L T4 polynucleotide kinase (Takara, 10 units/f.1L) using standard

procedures. The 5' -end-labeled DNA was recovered by ethanol

precipitation and further purified by 15% nondenatured gel

electrophoresis and isolated by crush and soak method. 14

The isolated

DNA was incubated with equimolar of the complementary strand in 100

f.1L of water at 90 oc for 5 min and cooled slowly to ambient

temperature for forming duplex.


stranded 43-mer DNA oligomers 5'-d(TTTTTGTTTGTTAGTT

CGTTTGCTTGGTTGATTGTTTGTTTTT)-3' and the corresponding

complementary oligomer were purchased from Greiner Japan Co. Ltd.

The 32P-5' -end-labeled ODN duplex was prepared as shown above. 50

f.1M of 1, 2 or 3 was incubated with 10 f.1M of calf thymus DNA and ca.

l.Ox106 cpm 32P-5'-end-labeled ODN duplex in 20 mM Tris-HCl buffer

( 100 f.1L, pH 7 .6) at 37 °C. At each time, the sample (1 0 f.1L) was

separated from the reaction mixture and precipitated with methanol.

The recovered DNA was dissolved in 100 pL of 10o/o (v/v) piperidine


114

and resuspended in 10 f.lL of 80% formamide loading buffer (80%

formamide, 1 mM EDTA, 0.1 o/o xylene cyanole and 0.1 o/o bron1ophenol

blue). The samples (1 f.lL) were loaded onto 15% polyacrylamide and 7




Supercoiled DNA Unwinding Assay. To a solution of pBR 322

plasmid DNA (250 ng, Nippon Gene) in topo I reaction buffer (10 rnM

Tris-HCl, pH 7.9, 1 mM EDTA, 150 mM NaCl, 0.1 mM spermidine,

5% glycerol, 0.1% BSA) was added human topoisomerase I (4 units,

TopoGEN). The reaction mixture was incubated for 30 min at 37°C.

After the test drug (1 00 f.1M) was added, the mixture was incubated for

another 30 min at 37 °C. The reaction was terminated by addition of

SDS to 1%. After proteinase K was added to 50 pg/mL, the mixture

was digested for 20 min at 56 °C. Addition of 0.1 vol. of 1 Ox gel

loading buffer was followed by chloroform extraction. Different forms

of DNA were separated at room temperature on a 1% agarose gel. The

gel was stained for 30 min with ethidium bromide (0.5 pM/mL) and

destained for 20 min in water. It was placed on a UV transilluminator

(313 nm) and photographed with Polaroid 665 film. The result was

shown in Figure 3.

115


(1) Barton, J. K. Science 1986, 233, 727-734.

(2) Geacintov, N. E.; Cosman, M.; Hingerty, B. E.; Amin, S.; Broyde,

S.; Patel, D. J. Chem. Res. Toxicol. 1997, 10, 111-146.

(3) (a) Boger, D. L.; Machiya, K.; Hertzog, D. L.; Kitos, P. A.; Holmes,

D. J. Am. Chem. Soc. 1993, 115, 9025-9036. (b) Boger, D. L.;

McKie, J. A.; Nishi, T.; Ogiku, T. J. Am. Chem. Soc. 1997, 119, 311-

325. (c) Boger, D. L.; Hertzog, D. L.; Bollinger, B.; Johnson, D. S.;

Cai, H.; Goldberg, J.; Turnbull, P. J. Am. Chem. Soc. 1997, 119,

4977-4986. (d) Boger, D. L.; Bollinger, B.; Hertzog, D. L.; Johnson,

D. S.; Cai, H.; Mesini, P.; Garbaccio, R. M.; Jin, Q.; Kitos, P. A. 1.

Am. Chem. Soc. 1997, 119, 4987-4998.

(4) Hurley, L. H.; Warpehoski, M. A.; Lee, C.-S.; McGovren, J. P.;

Scahill, T. A.; Kelly, R. C.; Mitchell, M. A.; Wicnienski, N. A.;

Gebhard, I.; Johnson, P. D.; Bradford, V. S. J. Am. Chem. Soc. 1990,

112, 4633-4649.

(5) (a) Gargiulo, D.; Musser, S. S.; Yang, L.; Fukuyama, T.; Tomatz,

M. J. Am. Chern. Soc. 1995, 117, 9388-9398.

(6) Link, J. T.; Raghavan, S.; Gallant, M.; Danishefsky, S. J.; Chou, T.

C.; Ballas, L. M. J. Am. Chem. Soc. 1996, 118, 2825-2842.

(7) Urata, H.; Ueda, Y.; Usami, Y.; Akagi, M. J. Am. Chem. Soc.

1993,115,7135-7138.



Sano, H. J. Antibiot. 1990,43, 1519-1523. (c) Hara, M.; Yoshida, M.;



116

Chan, K. L.; Sugiyama, H.; Saito, 1.; Hara, M. Phytochemistry 1995,

40, 1373-1374.

(9) Refer to chapter 3.

(10) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.;

Hartung, J.; Jeong, K.-S.; Kwong, H.-L.; Morikawa, K.; Wand, Z.-M.;

Xu, D.; Zhang, K.-L. J. Org. Chem. 1992, 57, 2768-2771.

(11) (a) Muench, K. F.; Misra, R. P.; Humayun, M. Z. Proc. Nat!.

Acad. Sci. USA 1983, 80, 6-10. (b) Stone, M. P.; Gopalakrishnan, S.;

Harris, T. M.; Graves, D. E. J. Biomol. Struct. Dyn. 1988, 5, 1025-

1041. (c) Benasutti, M.; Ejadi, S.; Whitlow, M. D.; Loechler, E. L.

Biochemistry 1988, 27, 472-481. (d) Baertschi, S. W.; Raney, K. D.;

Stone, M.P.; Harris, T. M. J. Am. Chem. Soc. 1988, 110, 7929-7931.

(e) Gopalakrishnan, S.; Byrd, S.; Stone, M. P.; Harris, T. M.

Biochemistry 1989, 28, 726-734. (f) Stone, M. P.; Gopalakrishnan, S.;

Raney, K. D.; Raney, V. M.; Byrd, S.; Harris, T. M. In Molecular Basis

of Specificity in Nucleic Acid-Drug Interactions; Pullman, B., Jortner,

J ., Eds.; Kluwer Academic Publishers: 1990, p 451-480. (g) Raney, K.

D.; Gopalakrishnan, S.; Byrd, S.; Stone, M. P.; Harris, T. M. Chem.

Res. Toxicol. 1990, 3, 254-261. (h) Gopalakrishnan, S.; Harris, T. M.;

Stone, M. P. Biochemistry 1990, 29, 10438-10448. (i) Baertschi, S.

W.; Raney, K. D.; Stone, M.P.; Harris, T. M. J. Am. Chem. Soc. 1991,

113, 4092-4096. (j) Raney, K. D.; Coles, B.; Guengerrich, F. P.;

Harris, T. M. Chem. Res. Toxicol. 1992, 5, 333-335. (k) Raney, K.

D.; S.; Harris, T. M.; Stone, M. P. Chem. Res. Toxicol. 1993, 6, 64-

68. (l) Iyer, R. S.; Coles, B. F.; Raney, K. D.; Thier, R.; Guengerich,

F. P.; Harris, T. M. J. Am. Chern. Soc. 1994, 116, 1603-1609.

(12) We performed energy minimizations starting from idealized B

form DNA helices to which kapurimycin ABCD' -ring analog had been

intercalated. We used the AMBER* set of force field parameters on

117

Macromodel: (a) Weiner, S. 1.; Kollman, P. A.; Case, D. A.; Singh, U.

C.; Ghio, C.; Alagona, G.; Profeta, S., 1r; Weiner, P. 1. Am. Chern.

Soc. 1984, 106, 765. (b) Weiner, S. 1.; Kollman, P. A.; Nguyen, D. T.;

Case, D. A. J. Comput. Chern. 1986, 7, 230. (c) Mohamadi, F.;

Richards, N. G. 1.; Guida, W. C.; Liskamp, R.; Caufield, C.; Chang, G.;

Hendrickson, T.; Still, W. C. J. Comput. Chern. 1990, 11, 440.

(13) The mean value of the results obtained from three calculations of

drug-DNA complex.


118

CHAPTER 5

Sequence Selective Alkylation of Continuous Guanine Sequences by DNA Intercalators Possessing Epoxy Side Chain

Abstract: In order to elucidate the origin of sequence selectivity for the intercalation into DNA by aromatic ring of DNA alkylating agents, we synthesized DNA alkylating agents with various intercalators such as anthraquinone 4, anthracene 5 and methoxyanthracene 6. In the cleavage assay using 32P-5' -end-labeled DNA, the anthraquinone 4 having the most electron-poor ring system among these three synthetic intercalators, showed the highest DNA cleavage activity. The most reactive site of DNA cleavage in this experiment was 5' -side guanine of GG sequence. From DNA unwinding assay and CD titration, anthraquinone 4 intercalated into DNA and its intercalation ability was the highest among three synthetic intercalators. The ab initio calculation of LUMO energy of anthraquinone, anthracene and methoxyanthracene provided that anthraquinone has a lowest LUMO energy. Further, anthraquinone has the lowest reduction potential among these intercalators. We investigated the alkylation of GGG sequence whose HOMO have already been calculated and showed that anthraquinone 4 intercalated between 5' -side guanine and central guanine of GGG sequence. This result indicates that the sequence selective intercalation is created by interaction of HOMO localized at 5' -side guanine in GGG sequence with LUMO of the intercalator.

121

Introduction

In several previous papers it has so far been reported to modelize

the interaction between DNA and DNA interacting agents, such as

ethidium bromide, sulfur mustard and nitrogen mustard with a goal to

access the interaction mechanism. 1 Their calculations suggest that the

intercalated complexes of drugs with DNA are stabilized by frontier

orbital interactions between the lowest unoccupied molecular orbital

(LUMO) of drug and the highest occupied molecular orbital (HOMO)

of DNA. Our group has also examined the most reactive sites in DNA

toward one electron photooxidation by the use of ab initio calculation of

the stacked base pair systems.2 Comparing HOMO energies obtained

from ab initio 6-31 G* calculations, we found that the GG/CC system

had the highest HOMO among the seven possible guanine-containing

base pairs and that the susceptibility of GN sequences is in the following

order, GG > GA > GT > GC. When two guanines are stacked each

other in a B-form geometry, the HOMO is localized only on the 5' -side

guan1ne.

An antitumor antibiotic kapurimycin A3 (1) 3 intercalates into DNA

and alkylate guanine N7 as well as aflatoxin B 1 oxide (2 )4 and

psorospermin (3 )5 do. In our study using kapurimycin analogs,6 we

found that DNA cleavage by these analogs was highly sequence selective

at 5' -side guanine of GG sequence. The 5' -side G selectivity for 5' -GN

sequence decreased in the order of 5' -GG > GA > GT > GC.

122

0 0

A cO 0 OMe kapurimycin A3 (1) aflatoxin B 1 oxide (2) psorospermin (3)

The order of sequence selectivity for guanine alkylation by

kapurimycin analogs is in good agreement with the results of our

calculation. We assumed that the key to solve the reason for GG

sequence selective alkylation was the interaction of guanine with HOMO

on the guanine and LUMO of intercalated subunit of DNA alkylating

agents. In order to elucidate the contribution of the LUMO energy of

intercalators in such HOMO-LUMO interactions, we synthesized DNA

alkylating agents possessing various intercalators such as 4, 5 and 6, and

compared their reactivity toward DNA and their sequence selectivity for

DNA alkylation. Herein, we report the relationship between the LUMO

energy of intercalators and the sequence selectivity of guanine

alkylation.

0 r r 0 or 0 0

o¢6 oc6 QC6 ~ ~ ~

0 MeO

4 5 6

123


We synthesized DNA alkylating agent 4 from 1-

hydroxyanthraquinone which was coupled with (25)-( + )-glycidyl 3-

nitrobenzenesulfate as shown in Scheme 1. 7 The other compounds 5 and

6 were also prepared from the corresponding 1-hydroxyanthracene

derivatives in a similar fashion.

Scheme 1.

0 0 ,, 0 _;'j S' ~~ 0

4 0 2) q 'b

(25% yield, 75% recovered) N02

First, we examined the DNA cleaving assay by the use of the

synthetic alkylating agents to know the relationship between the ring

system of intercalators and their sequence selective DNA cleavage

activity. The reaction of DNA with 4, 5 and 6 was examined using 32

P-

5' -end-labeled oligodeoxynucleotide and analyzed by electrophoresis on

denatured polyacrylamide gel. The labeled oligodeoxynucleotide duplex

was incubated with agents at 37 oc for 24 h and heated in 1 Oo/o

piperidine at 90 oc for 30 min to induce strand breakage. The result of

the assay was shown in Figure 1. DNA cleavage was observed at all

guanine sites in all these three cases. The reaction rates increased in the

order of 4 > 5 > 6 and the DNA cleaving ability of 6 was very weak.

Of special interest is that the cleavage by 4 was sequence selective at 5'

side guanine of GG sequence.

124

In order to know the correlation between the DNA binding modes

and sequence selectivity, we examined DNA unwinding assay using

topoisomerase I. Topoisomerase I is allowed to adjust DNA linking

number in the presence of intercalator that influences the DNA

unwinding angle, resulting in the shift of the gaussian distribution of

topoisomers. The solution of pBR322 plasmid DNA and each drug was

incubated with human topoisomerase I at 37 oc and the resulting DNAs

(b) relative O.D. (%) 50

40

(a)

T··· ... T A G T T G

- rG T / T:/

30 j 20 }Q

< I

o _l____j_ _ _lo.L_.l_ ]_...lm l I I "L _L _L__j~ L 1 JitrL J ~ 1 _, ~' th 1 ~ __,_c , , • I 5'- T GT T T GT TAG T T CGT T T GC T T Gi GT T GAT TG T T T GTorigin

- - --- --- - - - I

Figure 1. Strand breakage assays depicting reactive sites of oligodeoxynucleotide duplex for 4, ~ and 6. 32P-5'-end-labeled oligodeoxynucleotide previously prepared was treated with 4, 5 or 6 (500 pM) in a reaction buffer (20 mM Tris-HCl, pH 7.6 an~ 10 pM calf thymus DNA) at 37 OC for 24 h. After ethanol precipitation, the residue was heated in 10% piperidine at 90 OC for 30 min. The resulting DNAs were analyzed by electrophoresis on 15% denatured polyacrylamide gel. (a) Result of autoradiography. (b) DNA cleavage sites by 4. The height of bars in the histogram showed the percentage of strand breakage at a given site relative to the total strand breakage.

125

were analyzed by agarose gel electrophoresis. As clear from Figure 2,

anthraquinone 4 intercalated into DNA much more strongly than 5 and

6 did. Furthermore, we observed the structural change of duplex DNA

by addition of anthraquinone 4 by means of CD spectroscopy. The

structure of calf thymus DNA (0.1 mM) in the presence of 4 (0.05 mM)

was compared with that of calf thymus DNA in the absence of 4 in a

buffer (10 mM Tris-HCl, pH 7.3, 10% (v/v) acetonitrile). The addition

of 4 to a DNA solution led to the decrease of the characteristic Cotton

effects ascribed to B-form DNA, the positive CD at 275 nm and the

negative CD at 245 nm, together with the increase of the negative

induced CD at 450 nm derived from the formation of 4-DNA complex.

This result shows the structural change of DNA induced by intercalation

of 4 to DNA duplex.

open-ctrcular

\ #

supercoiled ''' 12345678 ~~~~~~

control 6 5 4

Figure 2. Unwinding of supercoiled pBR322 DNA by topoisomerase I in the presence of kapurimycin analogs 4, 5 and 6. Supercoiled pBR322 DNA (250 ng) was first treated with topoisomerase I (topo I) for 30 min in a reaction buffer (10 mM Tris-HCl, pH 7.9, 1 mM EDTA, 150 mM NaCl, 0.1 mM spermidine, 5% glycerol, 0.1% BSA), and further incubated for 30 min in the absence (lane 2) and presence of 4, 5 and 6 (10 ,uM) (lanes 4, 6 and 8, respectively). The resulting DNAs were analyzed by electrophoresis on 1% native agarose gel at 1.3 V/cm for 18 h. DNAs in lanes 1, 3, 5 and 7 were treated as for lanes 2, 4, 6 and 8, respectively, in the absence of topo I. lane 1, intact DNA; lane 2, topo I; lane 3, 6; lane 4, 6 with topo I; lane 5, 5; lane 6, 5 with topo I; lane 7, 4; lane 8, 4

with topo I.

126

5 I 4 . 3 2 ~

b 1 1 X 0 i'--rr--;IJ-~r--:7'.aw~~.;::..:..::...~~.:::::...

~ -1 ~ -2 -3 -4 r-

-5 -'---- ~ - -

200 250 300 350 400 450 500

wave length (nm)

Figure 3. CD spectral change due to the interaction of 4 with calf thymus DNA. (A) A solution of calf thymus DNA (0.1 mM) in the absence of 4 in the buffer (10 mM Tris-HCl, pH 7.3, 10% (v/v) acetonitrile), plain line; (B) A solution of calf thymus DNA (0.1 mM) in the presence of 4 (0.05 mM) in the buffer (10 mM Tris-HCl, pH 7.3, 10% (v/v) acetonitrile), bold line.

Next, in order to know the reason for the different binding abilities

of the intercalators, we investigated the electronic structure of

intercalator by means of ab initio calculations. In this study, we

performed the MO calculation of the molecules 1n which

oxiranylmethoxy group was replaced with methyl group. The

calculated LUMO energies of intercalators were shown in Table 1. The

LUMO energy of anthraquinone was, the lowest among the three

intercalators and the order of the heights of LUMO energies of aromatic

compounds was comparable to the order of the reactivities of the

corresponding alkylating agents to DNA. Then, the reduction potentials

of 4, 5 and 6 were measured by cyclic voltammetry. The reduction

potentials were shown in Table 1. The reduction potential of 4 was the

smallest among the three intercalators and the order of their potentials

was in agreement with LUMO calculation. These results suggest that

there is a correlation between the LUMO energy of the intercalators and

the observed sequence selectivity for DNA alkylation.

127

Figure 4. LUMO of anthraquinone 4. MO c~lculation was performed at the HF/6-31G* level utilizing Spartan (verswn 4.0.2 GL).

Table 1. Reduction Potentials and LUMO Energies of Alkoxy-substituted Anthraquinone and Anthracene.

1-( methox ymethoxy )anthraquinone

1-(methoxymethoxy)anthracene

1 ,5-bis(methoxymethoxy)anthracene

Reduction LUMO Potential (V)a Energy ( e V)b

-D.70

-1.99

-2.02

0.095lc

0.2118d

0.2307e

a Reduction potential given from the cyclic voltammograms of DNA intercalators in 0.1 M LiC104 solution in acetonitrile. Scan rate: 100 mV/s . Working electrode: Pt wire. Reference electrode: Ag/ Agel (1 .0 M KCl). Counter electrode: glassy carbon. b LUMO energies was calculated by ab initio (HF/6-31 G*). c 1-Methoxyanthraquinone was calculated. d 1-Methoxyanthracene was calculated. e 1 ,5-Dimethoxyanthracene was calculated.

Next, we have examined the cleaving assay of the DNA containing

GGG sequence by the use of 4 to elucidate that the interaction between

DNA HOMO and intercalator LUMO. Our previous calculations have

indicated that stacking of three guanine bases significantly lowered the

HOMO energy and that the HOMO of the stacked 5'-GGG-3' was

localized mainly on the 5' -side guanine in a B-form DNA duplex

128

(Figure 5). If there is a HOMO- LUMO interaction between GGG

sequence and intercalators, intercalators would not equally insert into

two intercalation sites, i.e. the site between 5' -side guanine and central

guanine (Figure 6, site A) and the site between central guanine and 3'

side guanine (Figure 6, site B). When the HOMO-LUMO interaction

prevails, then the intercalators predominantly insert into site A.

Each drug was incubated with 32P-5' -end-labeled oligodeoxy

nucleotide as shown in Figure 7 and the DNA sample was subsequently

separated from the reaction mixture at each time and treated with 10%

piperidine solution at 90 oc for 30 min after ethanol precipitation. The

result analyzed by polyacrylamide gel electrophoresis was shown in

Figure 7. Anthraquinone 4 cleaved DNA preferentially at 5' -side

guanine site of continuous guanine sequence. The ratio of cleaving

intensity at each guanine in guanine triplet (G 1G2G3) by 4 in 5 h was

about 6:2:1 and the ratio in guanine doublet G1G2 sequence was about

10:1. The selective alkylation of 5' -side guanine in GG sequence shows

that 4 selectively alkylates 5' -side guanine at its intercalation site. Thus,

judging from the efficiency of DNA cleavage at 5' -side guanine and

central guanine in GGG sequence, 4 predominantly intercalates into site

A and alkylates 5' -side guanine of the intercalation site. This result

shows that the interaction between the HOMO on 5' -side guanine of

GGG sequence and the LUMO of intercalators determines the sequence

selectivity for the intercalation into DNA. In other words, this result

suggests that the 5' -side guanine of continuous guanines is extremely

important in HOMO-LUMO interactions of B-form DNA with electron

deficient intercalators.

129

) 3' : T

G

G

G

5'

Figure 5. HOMO of 5'-d(TGGGT)-3'/5'-d(ACC~_A_)-3'. MO calculation was performed at the HF/6-310* level ut1hzmg Spartan (version 4.0.2 GL).

(a) 5' 3'

9 ?3 r-- r2-

1 I 4 ]I site A

s;···············S31-3' 5'

Figure 6. Possible intercalation sites in GOG sequence. (a) Intercalati~n of 4 between G

1 and G2 (site A). (b) Intercalation of 4 betwee~ 0 2 and ?3 (s1te B).

The length of arrow shows the probabilities of G alkylatiOn ?Y 4 mtercalated into GG doublet being estimated on the basis of the result of F1gure 1.

130

(a)

T •• ••• ..-G

_...G ~G

T ----

T -..-Q ~G

T ---

4

(b)

12 ' 10

8 I 6 I 4

2

0

relative 0 .0. (%)

~ 0 ~ m D 5'- T G G T ···· T ,c; G G T -3' -

Figure 7. Strand breakage assays depicting reactive sites for 4 on the oligodeoxynucleotide duplex. 32P-5'-end-labeled oligodeoxynucleotide previously prepared was treated with 4 (500 pM) in a reaction buffer (20 mM Tris-HCl, pH 7.6 and 10 .uM calf thymus DNA) at 37 °C. After ethanol precipitation, the residue was heated in 10% piperidine at 90 oC for 30 min . The resulting DNAs were analyzed by electrophoresis on 15% denatured polyacrylamide gel. (a) The result of autoradiography. (b) DNA cleavage sites by 4 in 5 h. The height of bars in the histogram shows the percentage of strand breakage at a given site relative to the total strand breakage.

Conclusion

We synthesized DNA alkylating agents possessing various

intercalators such as anthraquinone 4, anthracene 5 and methoxy

anthracene 6 to investigate the sequence selectivity of their intercalation.

It was shown that the ability of sequence selective DNA intercalation

increased in the order of 4 > 5 > 6, which were reverse to the order of

their LUMO energies. In 5' -G 1G2G3 sequence, anthraquinone 4 having

the lowest LUMO energy selectively intercalated between G, and G2 .

This result shows that the sequence selectivity for intercalative DNA

131

alkylators is originated primarily from the interaction between the

HOMO of GGG sequence and the LUMO of intercalators.

132


General Techniques. 1H NMR spectra were measured with Varian

Mercury (400 MHz) spectrometers. Coupling constants (J values) are

reported in Hz. 13C NMR spectra were measured with Varian Mercury

(100 MHz) spectrometers. The chemical shifts are expressed in ppm

downfield from tetramethylsilane, using residual chloroform ( 8 = 7.24

in 1H NMR, 8 = 77.0 in 13C NMR) as an internal standard. The

following abbreviations were used to explain the multiplicities: s,

singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad. IR

spectra were recorded on JASCO FTIIR-5M spectrophotometers. UV

visible spectra were recorded on JASCO V -550 UV Nis

spectrophotometers. Melting points were obtained on a Y anagimoto

Seisakusho micro melting point apparatus and are uncorrected. Optical




Cyclic voltammetry was performed using BAS 1 OOB/W electrochemical

analyzer. Circular dichromism spectra were recorded on JASCO J-700

spectrophotometer. Microanalyses were performed by Kyoto

University Microanalytical Center.



5% ethanolic phosphomolybdic acid, or p-anisaldehyde solution and heat

as developing agent. Wako gel (C-200, particle size 75-150 Jlm, Wako)

was used for column chromatography. Tetrahydrofuran and ethyl ether

were distilled over sodium-benzophenone. Dichloromethane, toluene

and N,N-dimethylformamide were distilled over calcium hydride. All

133

reagents were purchased at highest commercial quality and used without

further purification unless otherwise stated.





1-(2-0xiranylmethoxy)anthracene-9,10-dione (4). To a

suspension of sodium hydride (60o/o, 20.2 mg, 0.51 mmol) 1n N,N

dimethylformamide (2 mL) was added 1-hydroxyanthraquinone (99.5

mg, 0.44 mmol) at 0 °C, and the reaction mixture was stirred at 80 oc for 30 min. After cooling down to ambient temperature, to this mixture

was added a solution of (2S)-( + )-glycidyl 3-nitrobenzenesulfate ( 15.5

mL, 114.3 mmol) in N,N-dimethylformamide (1 mL) at ambient

temperature, and the reaction mixture was stirred at 100 oc for 30 min.

After cooling down to ambient temperature, the mixture was diluted

with sat. aq. NH4Cl and extracted with ethyl acetate. The crude product

was purified by column chromatography on silica gel, eluting with 30 %

ethyl acetate in hexane to give 4 (31.0 mg, 0.11 mmol, 25o/o) as a yellow

solid and the recovered 1-hydroxyanthraquinone (72.4 mg, 0.32 mmol,

73%): mp. 174-175°C; 1H NMR (CDC13, 400 MHz) 8 8.27-8.21 (2H),

7.99 (dd, 1H, J = 7.7, 1.1 Hz), 7.77-7.67 (3H), 7.36 (dd, 1H, J = 8.5,

1.1 Hz), 4.48 (dd, 1H, J = 11.3, 2.7 Hz), 4.23 (dd, 1H, J = 11.2, 4.6 Hz),

3.50-3.48 (m, 1H), 3.13 (dd, 1H, J = 5.1, 2.7 Hz), 2.98 (dd, 1H, J = 5.1,

4.0 Hz); MS (El) m/e (%) 280 (M+, 22), 237 (31), 224 (100), 139 (54);

HRMS (El) calcd for C17

H12

0 4 (M+) 280.0736, found 280.0733; Anal.

Calcd for C17

H120

4: C, 72.85; H, 4.32. Found: C, 72.56; H, 4.45.

134

1-(2-0xiranylmethoxy)anthracene (5). The same procedure

described for the synthesis of 4 was applied for 1-hydroxyanthracene

(143.6 mg, 0.74 mmol) to afford 5 (143.9 mg, 77%) as a pale yellow

solid: 1H NMR (CDC13, 400 MHz) 8 8.86 (s, lH), 8.36 (s, lH), 8.06-

8.03 (m, lH), 7.98-7.96 (m, lH), 7.59 (d, 1H, J = 8.6 Hz), 7.48-7.44

(2H), 7.33 (t, 1H, J = 8.5 Hz), 6.71 (d, 1H, J = 7.5 Hz), 4.47 (dd, 1H, J

= 11.0, 3.1 Hz), 4.19 (dd, lH, J = 11.1, 6.7 Hz), 3.59-3.57 (m, lH),

3.02 (t, lH, J = 4.5 Hz), 2.90 (dd, 1H, J = 4.9, 2.6 Hz); MS (El) m/e

(%) 250 (M+, 100), 208 (27), 193 (37), 165 (96); HRMS (EI) calcd for

C17H1402 (M+) 250.0994, found 250.0999.

1-Methoxy-5-(2-oxiranylmethoxy)anthracene (6) Th . e same

procedure described for the synthesis of 4 was applied for 1-hydroxy-5-

methoxyanthracene (10.6 mg, 47.3 pmol) to afford 6 (6.1 mg, 46%) as

a yellow solid: 1H NMR (CDC13, 400 MHz) 8 8.80 (s, 1H), 8.76 (s, 1H),

7.64-7.61 (2H), 7.37-7.28 (2H), 6.75-6.71 (2H), 4.46 (dd, 1H, J = 11.0,

3.1 Hz), 4.20 (dd, 1H, J = 11.0, 5.7 Hz), 4.06 (s, 3H), 3.59-3.55 (m,

1H), 3.00 (t, 1H, J = 4.5 Hz), 2.89 (dd, 1H, J = 4.8, 2.6 Hz); MS (EI)

m/e (%) 280 (M+, 100), 223 (60), 195 (35), 152 (34); HRMS (EI) calcd

for C 1sH 1603 (M+) 280.1100, found 280.1093.

Preparation of 32P-5' -End-Labeled Oligodeoxynucleotide

Duplex. 400 pmol of single-strand oligodeoxynucleotide, purchased

from Greiner Japan Co. Ltd., was 5' -end-labeled by phosphorylation

with 4 pL of [y- 32P]ATP (Amersham, 370 MBq/pL) and 4 pL T4

polynucleotide kinase (Takara, 10 units/pL) using standard procedure.

The 5' -end-labeled DNA was recovered by ethanol precipitation and

further purified by 15 o/o nondenatured gel electrophoresis and isolated

by crush and soak method. 8 The isolated DNA was incubated with

135

equimolar of the complementary DNA in 100 J.1L of water at 90 oc for

5 min and cooled slowly to ambient temperature for forming duplex.

Cleavage of 32P-5' -End-Labeled Oligodeoxynucleotide. Single

stranded 43-mer DNA oligomers 5' -d(TTTTTGTTTGTTAGTTCG

TTTGCTTGGTTGATTGTTTGTTTTT)-3' and the corresponding

complementary oligomer were purchased from Greiner Japan Co. Ltd.

:nP-5' -end-labeled ODN duplex was prepared as shown above. A

solution of 4, 5 or 6 (each 500 J.1M) was incubated with 10 J.1M of calf

thymus DNA and ca. l.Ox106 cpm 32P-5' -end-labeled ODN duplex in 20

mM Tris-HCl buffer (100 J.1L, pH 7.6) at 37 oc for 24 h. The samples

were precipitated with methanol, and the recovered DNA was dissolved

in 100 J.1L of 10o/o (v/v) piperidine and heated at 90 oc for 30 min. The

mixture was concentrated in vacuo and resuspended in 10 J.1L of 80%

formamide loading buffer (80o/o formamide, 1 mM EDT A, 0.1% xylene

cyanole and 0.1% bromophenol blue). The samples (1 J.1L) were loaded

onto 15% polyacrylamide and 7 M urea sequence gel and

electrophoresesed at 1900 V for ca. 2 h. The gel was dried and exposed

to X -ray film with intensifying sheet at -70 °C. The result was shown

in Figure 1.


stranded 33-mer DNA oligomers 5'-d(CGTTATCATTGGTTATCATT

GGGTT A TCA TTCG)-3' and the corresponding complementary

oligomer were purchased from Greiner Japan Co. Ltd. The 32P-5' -end

labeled ODN duplex was prepared as shown above. A solution of 4

(500 J.1M) was incubated with 10 J.LM of calf thymus DNA and ca.

l.Ox106 cpm 32P-5' -end-labeled ODN duplex in 20 mM Tris-HCI buffer

(100 f.lL, pH 7 .6) at 37 °C. At each time, the sample (30 J.1L) was

136

separated from the reaction mixture and precipitated with methanol.

The recovered DNA was dissolved in 100 J.1L of 10% (v/v) piperidine


and resuspended in 10 J.1L of 80% formamide loading buffer (80%

formamide, 1 mM EDT A, 0.1% xylene cyanole and 0.1% brotnophenol

blue). The samples (1 J.1L) were loaded onto 15% polyacrylamide and 7




137


(1) (a) Broch, H.; Hamza, A.; Vasilescu, D. J. Biomol. Struct. Dyn.

1996, 13, 903-914. (b) Broch, H.; Hamza, A.; Vasilescu, D. J. Biomol.

Struct. Dyn. 1996, 13, 915-924. (c) Patterson, S. E.; Coxon, J. M.;

Strekowski, L. Bioorg. Med. Chern. 1997, 5, 277-281.

(2) Sugiyama, H.; Saito, I. J. Am. Chern. Soc. 1996, 118, 7063-7068.



Sano, H. J. Antibiot. 1990, 43, 1519-1523. (c) Hara, M.; Yoshida, M.;



Chan, K. L.; Sugiyama, H.; Saito, I.; Hara, M. Phytochemistry 1995,

40, 1373-1374.

(4) (a) Stone, M.P.; Gopalakrishnan, S.; Harris, T. M.; Graves, D. E. J


Byrd, S.; Stone, M. P.; Harris, T. M. Biochemistry 1989, 28, 726-734.

(c) Stone, M. P.; Gopalakrishnan, S.; Raney, K. D.; Raney, V. M.;

Byrd, S.; Harris, T. M. In Molecular Basis of Specificity in Nucleic





M. P. Chern. Res. Toxicol. 1993, 6, 64-68. (f) Gopalakrishnan, S.;

Harris, T. M.; Stone, M. P. Biochemistry 1990, 29, 10438-10448.

(5) (a) Kupchan, S. M.; Streelman, D. R.; Sneden, A. T. J. Nat. Prod.

1980, 43, 296-301. (b) Habib, A. M.; Ho, D. K.; Masuda, S.;

McCloud, T.; Reddy, K. S.; Aboushoer, M.; McKenzie, A.; Bym, S. R.;

Chang, C. J.; Cassady, J. M. J. Org. Chern. 1987, 52, 412-418. (c)

138

Hansen, M.; Lee, S.-J.; Cassady, J. M.; Hurley, L. H. J. Am. Chern. Soc.

1996, 118, 5553-5561.

( 6) Refer to the previous chapter.

(7) Klunder, J. M.; Onami, T.; Sharpless, K. B. J. Org. Chern. 1989,

54, 1295-1304.


139

List of Publications

Chapter 1 Highly Efficient Synthesis of 2-Substituted 4H-Chromen-4-

ones by means ofF-Induced 6-Endo-Digonal Cyclization

of o-(Silyloxy)phenyl Ethynyl Ketone Derivatives.

Nakatani, K. ; Okamoto, A.; Yamanuki, M.; Saito, I. 1.

Org. Chern. 1994, 59, 4360-4361.

6-Endo- and 5-Exo-Digonal Cyclizations of a-Hydroxy

phenyl Ethynyl Ketones: A Key Step for Highly Selective

Benzopyranone Formation.

Nakatani, K.; Okamoto, A.; Saito, I. Tetrahedron 1996,

52, 9427-9446.

Chapter 2 Synthesis of ABC Ring Analog of Kapurimycin A3 as an

Effective DNA Alkylating Agent.

Nakatani, K.; Okamoto, A.; Saito, I. Angew. Chern. 1997,

109, 2881-2883.

Truncated Analogs of Kapurimycin A3 and their DNA

Alkylation Mechanism.

Nakatani, K.; Okamoto, A.; Okuda, S.; Saito, I. Nucleic

Acid Symp. Ser. 1996, 35, 83-84.

Chapter 3 Studies on the Mechanism of DNA Sequence Selective

Alkylation by Kapurimycin A3 Analogs.

Okamoto, A.; Nakatani, K.; Saito, I. Nucleic Acid Symp.

Ser. 1997, 37, 27-28.

140

Chapter 4 Highly Selective Alkylation at S'G of 5 'GG3' Sequence by

an Aglycon Model of Pluramycin Antibiotics through

Preferential Intercalation into GG Step.

Nakatani, K.; Okamoto, A.; Matsuno, T.; Saito, I.

Submitted to 1. Am. Chern. Soc ..

Chapter 5 Okamoto, A.; Nakatani, K.; Saito, I. To be submitted.

141

List of Oral Presentations

1. "Synthetic Studies on Kapurimycin A3 and its Analogs: Indispensable

Subunit for Efficient DNA Alkylation"

K. Nakatani, A. Okamoto, M. Yamanuki, M. Takayama, I. Saito.

67th Annual Meeting of Chemical Society of Japan, Tokyo, Japan,

March, 1994.

2. "Synthetic Studies on Kapurimycin A3 and its Analogs"

K. Nakatani, A. Okamoto, I. Saito. 69th Annual Meeting of

Chemical Society of Japan, Kyoto, Japan, March, 1995.

3. "Synthesis of Kapurimycin A3 Analogs and their Reaction with

DNA"

K. Nakatani, A. Okamoto, I. Saito. 67th Symposium on Organic

Synthesis, Tokyo, Japan, May, 1995.

4. "Synthesis of Analogs of Antitumor Antibiotic Kapurimycin A3 and

the Analysis of the Reaction with DNA"

K. Nakatani, A. Okamoto, I. Saito. 37th Symposium on the

Chemistry of Natural Products, Tokushima, Japan, October, 1995.

5. "Studies on DNA Alkylation by Analogs of Antitumor Antibiotic

Kapurimycin A3"

K. Nakatani, A. Okamoto, I. Saito. 70th Annual Meeting of

Chemical Society of Japan, Tokyo, Japan, March, 1996.

142

6. "DNA Alkylation Mechanism of ABCD-ring Analogs of

Kapurimycin A3

"

K. Nakatani, A. Okamoto, T. Matsuno, I. Saito. 72th Annual

Meeting of Chemical Society of Japan, Tokyo, Japan, March, 1 997.

7. "Mechanism of Guanine-Guanine Sequence-Selective DNA

Alkylation by Kapurimycin A3

Analogs"

A. Okamoto, K. Nakatani, I. Saito. 214th American Chemical

Society National Meeting, Las Vegas, USA, September, 1997.

8. "Studies on the Mechanism of DNA Sequence Selective Alkylation

by Kapurimycin A3 Analogs"

A. Okamoto, K. Nakatani, I. Saito. 24th Symposium on Nucleic

Acids Chemistry, Tokyo, Japan, November, 1997.

Other Oral Presentation

1. "Intramolecular Silylstannylation to Acetylene"

K. Tamao, A. Okamoto, T. Kobayashi, A. Kawachi, Y. Ito. 65th

Annual Meeting of Chemical Society of Japan, Tokyo, Japan,

March, 1993.

143

design, synthesis and evaluation of novel dna alkylating...

Documents