Hypoxia-activated prodrugs: Substituent effects on the properties of nitro seco-1,2,9,9a-tetrahydrocyclopropa[c]benz[e]indol-4-one (nitroCBI) prodrugs of DNA minor groove alkylating agents

7258 J. Med. Chem. 2009, 52, 7258–7272

DOI: 10.1021/jm901202b

Hypoxia-Activated Prodrugs: Substituent Effects on the Properties of

Nitro seco-1,2,9,9a-Tetrahydrocyclopropa[c]benz[e]indol-4-one (nitroCBI) Prodrugs of

DNA Minor Groove Alkylating Agents

Moana Tercel,*^,†Graham J. Atwell,^†Shangjin Yang,^†Ralph J. Stevenson,^†K. Jane Botting,^†Maruta Boyd,^†Eileen Smith,^†

Robert F. Anderson,^†,‡William A. Denny,^†William R. Wilson,^†and Frederik B. Pruijn^†

^†Auckland Cancer Society Research Centre, Faculty of Medical and Health Sciences, The University of Auckland, Private Bag 92019,

Auckland 1142, New Zealand, and ^‡Department of Chemistry, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand

Received August 11, 2009

Nitrochloromethylbenzindolines (nitroCBIs) are a new class of hypoxia-activated prodrugs for antitumor

therapy. The recently reported prototypes undergo hypoxia-selective metabolism to form potent DNA

minor groove alkylating agents and are selectively toxic to some but not all hypoxic tumor cell lines. Here we

report a series of 31 analogues that bear an extra electron-withdrawing substituent that serves to raise the

one-electron reduction potential of the nitroCBI. We identify a subset of compounds, those with a basic side

chain and sulfonamide or carboxamide substituent, that have consistently high hypoxic selectivity. The best

of these, with a 7-sulfonamide substituent, displays hypoxic cytotoxicity ratios of 275 and 330 in Skov3 and

HT29 human tumor cell lines, respectively. This compound (28) is efficiently and selectively metabolized to

the corresponding aminoCBI, is selectively cytotoxic under hypoxia in all 11 cell lines examined, and

demonstrates activity against hypoxic tumor cells in a human tumor xenograft in vivo.

Introduction

(1-methyl-2-nitro-1H-imidazol-5-yl)methyl N,N-bis(2-bromo-

ethyl)phosphordiamidate (TH-302¹⁴), the latter of which

are both prodrugs of nitrogen mustard DNA cross-linking

agents.

The disorganized vasculature of solid tumors frequently

gives rise to regions of chronic or acute hypoxia. Cells within

these regions show resistance to radiotherapy and some forms

of chemotherapy and experience selective pressures that pro-

mote a more malignant phenotype.^1,2One possible approach

to treating hypoxic cells involves the development of hypoxia-

activated prodrugs (HAPs^a). These are relatively nontoxic

compounds that are activated by metabolic reduction in the

absence of oxygen. Since severe hypoxia is rarely found in

normal tissues, HAPs represent an avenue to target the most

refractory cells of solid tumors in a selective fashion.³

Successful exploitation of the HAP concept requires careful

prodrug design.^4,5Important considerations include (i) the

toxicity differential between prodrug and its reduction

product(s), (ii) the selectivity for activation by one-electron

reductases rather than two-electron (oxygen insensitive) re-

ductases,⁶(iii) tissue penetration properties that allow the

prodrug to reach hypoxic cells distant from the vasculature,⁷

and (iv) whether the reduction product is capable of diffusing

from the site of activation to kill surrounding cells. This last

property (the bystander effect) may allow for effective killing

of hypoxic cells without requiring metabolic activation in

every one.⁸Examples of HAPs in recent or current clinical

trial include the N-oxides tirapazamine^9,10(1, Figure 1) and

banoxantrone^11,12and the nitro compounds 2 (PR-104¹³) and

We have investigated a class of DNA monoalkylating

agents as possible “effectors” for various tumor selective

prodrug strategies. These compounds are related to a family

of antitumor antibiotics (including duocarmycin SA and

yatakemycin) that alkylate in a sequence-selective manner at

the N3 position of adenine in the minor groove of DNA and

are exceptionally potentcytotoxins.^15,16Four analogues of the

natural products were evaluated clinically on the basis of

excellent antitumor activity in preclinical models,^17,18but all

proved to be myelotoxic in patients and lacking in antitumor

activity at the low doses that could be administered systemi-

cally.^19,20Such observations have prompted efforts to target

similar compounds to tumor tissue, for example, by the use of

antibody-drug conjugates²¹or the formation of prodrugs

such as glycosides²²or metal complexes.²³

We reasoned that amino analogues of the alkylating subunit

of these compounds (see, for example, structure 5 in Figure 2)

would undergo spirocyclization and DNA alkylation in a

similar fashion to the natural products, which carry a phenol

at this position, and further that such amino compounds would

be amenable to the formation of a range of new prodrugs. We

have reported the syntheses of three types of amino alkylating

agents (CI,²⁴CBI,²⁵and DSA²⁶) along with their DNA

alkylation properties^27,28and cytotoxicity data. The last shows

that when the amino substituent is replaced by an electron

withdrawing group (a carbamate²⁹or nitro group^25,26,30for

instance), the cytotoxicity is dramatically reduced, presumably

because the compounds can no longer undergo spirocyclization

and DNA alkylation. We recently reported two particular

*To whom correspondence should be addressed. Phone: 64-9-373-

7599. Fax: 64-9-373-7502. E-mail: m.tercel@auckland.ac.nz.

^aAbbreviations: CBI, chloromethylbenzindoline; DEI, 5-[(di-

methylamino)ethoxy]indole; EDCI, N-ethyl-N⁰-(3-dimethylaminopro-

pyl)carbodiimide hydrochloride; EWG, electron-withdrawing group;

HAP, hypoxia-activated prodrug; HCR, hypoxic cytotoxicity ratio;

TMI, 5,6,7-trimethoxyindole.

r

pubs.acs.org/jmc

Published on Web 10/30/2009

2009 American Chemical Society

Article

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22 7259

examples of nitroCBI 4 that show promise as HAPs:³¹

(Table 1), bearing the neutral 5,6,7-trimethoxyindole (TMI)

6

side chain as found in duocarmycin SA, and 19 (Table 2) with a

basic 5-[(dimethylamino)ethoxy]indole (DEI) side chain. 6 was

found to be up to 300-fold more cytotoxic to cultured tumor

cells under hypoxic compared to aerobic conditions, while

19 was found to generate the corresponding aminoCBI (45,

Table 3) selectively in hypoxic cell culture. 45 was demonstrated

to be a diffusible metabolite capable of bystander killing of

surrounding cells in multicellular layer cultures. In addition 19

proved to be considerably less toxic to mice than 45 by ip

administration and provided selective killing of hypoxic cells in

murine RIF-1 tumors in vivo.³¹

Some shortcomings in the properties of 6 and 19 as HAPs

were also noted.³¹The hypoxic cytotoxicity ratio (HCR) was

only high in some cell lines and was generally weaker when the

TMI side chain of 6 was replaced by the somewhat more

soluble DEI side chain of 19 (to the point where the HCR for

19 in HT29 cells was only 2.5 as measured by clonogenic

assay). In addition, significant activity in vivo was only

achieved at or above the maximally tolerated dose. One

limitation may be that the reduction potential of 4 is too

low for efficient bioreduction, which could contribute to the

slow rates of cellular metabolism observed for 19 and explain

why 6, although clearly reduced under hypoxia, fails to

become as cytotoxic as the corresponding aminoCBI (39,

Table 3) under the cell culture conditions used. Here we

describe the synthesis of a number of new analogues of

4 where the A-ring bears an extra electron-withdrawing

group (EWG) at any one of the available positions and the

effect of this substituent on the reduction potential and

cytotoxicity of the compounds under oxic and hypoxic con-

ditions. We identify a small group of A-ring substituents that

confer consistently high hypoxic selectivity across a panel

of human tumor cell lines and a particular compound

with activity against hypoxic cells in SiHa human tumor

xenografts.

Figure 1. Structures of some known HAPs.

Figure 2. Proposed mechanism of action of nitroCBI hypoxia-

activated prodrugs. If 4 is reduced to 5 via an initial one-electron

step, this process can be reversed in the presence of oxygen. R is a

side chain that can bind in the minor groove of DNA.

Table 1. Structure and in Vitro Cytotoxicity Data for NitroCBIs Bearing a TMI (5,6,7-Trimethoxyindole) Side Chain

Skov3

HT29

IC₅₀(μM)^a

compd

R

oxic

hypoxic

HCR^c

oxic

hypoxic

HCR^c

b

σ_p

6

H

6-NO₂

0.00

4.2 ( 1.3

>20^d

1.7 ( 0.2

>20^d

2.8 ( 0.9

3.4 ( 0.4

>20^d

1.6 ( 0.7

>20^d

2.5 ( 0.7

7

8

6-COMe

7-NO₂

0.50 ( 0.07

16 ( 1

17 ( 1

6.4 ( 2.5

37 ( 3

5.1 ( 1.9

1.1 ( 0.3

5.8 ( 1.0

11 ( 5

0.14 ( 0.05

0.24 ( 0.02

9.3 ( 1.6

1.0 ( 0.4

3.7 ( 0.4

0.44 ( 0.15

0.51 ( 0.20

1.6 ( 0.5

3.9 ( 1.1

4.9 ( 0.6

>50^d

3.4 ( 0.7

70 ( 8

1.9 ( 0.4

13 ( 9

10 ( 1

19 ( 8

4.0 ( 2.8

6.1 ( 0.4

4.2 ( 2.3

2.6 ( 0.6

<1.1

0.68 ( 0.11

14 ( 2

14 ( 1

5.3 ( 1.4

43 ( 5

4.1 ( 1.4

1.9 ( 0.5

7.9 ( 0.7

14 ( 5

0.33 ( 0.01

1.3 ( 0.1

7.2 ( 1.7

2.4 ( 0.5

12 ( 1

0.93 ( 0.21

0.81 ( 0.27

3.6 ( 0.7

7.2 ( 1.6

14 ( 2

2.0 ( 0.6

12 ( 2

9

0.78

0.72

0.66

0.57

0.50

0.45

0.36

10

11

12

13

14

15

16

17

18

7-SO₂Me

7-CN

2.0 ( 0.3

2.5 ( 0.6

3.6 ( 0.2

5.2 ( 2.1

3.2 ( 1.8

2.8 ( 0.4

2.5 ( 1.4

1.9 ( 0.1

1.2 ( 0.1

7-SO₂NH₂

7-COMe

7-CO₂Me

7-CONH₂

8-SO₂Me

8-SO₂NH₂

9-NO₂

12 ( 2

55 ( 3

24 ( 2

34 ( 7

33 ( 10

^aDrug concentration to reduce cell density to 50% of that of the controls following 4 h of exposure. Values are the mean ( SEM for two to nine

experiments. ^bσ_pfrom ref 37. ^cHypoxic cytotoxicity ratio = IC₅₀(oxic)/IC₅₀(hypoxic). Values are the mean of intraexperiment ratios ( SEM for two to seven

experiments. ^dIC₅₀greater than solubility limit.

7260 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22

Tercel et al.

Table 2. Structure and in Vitro Cytotoxicity Data for NitroCBIs Bearing a DEI [(5-Dimethylaminoethoxy)indole] Side Chain

Skov3

IC₅₀(μM)^a

HT29

IC₅₀(μM)^a

compd

R

E(1) (mV)^b

oxic

hypoxic

HCR^c

oxic

hypoxic

HCR^c

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

H

6-NO₂

6-CN

-512 ( 8

-467 ( 8

-491 ( 8

-497 ( 7

-489^d

1.4 ( 0.5

4.1 ( 1.1

0.17 ( 0.01

3.7 ( 0.7

0.64 ( 0.11

50 ( 2

1.1 ( 0.1

1.3 ( 0.1

0.81 ( 0.18

6.9 ( 1.5

0.64 ( 0.25

0.44 ( 0.18

3.9 ( 0.5

1.9 ( 0.5

1.2 ( 0.3

4.9 ( 0.4

0.20 ( 0.04

0.52 ( 0.05

3.4 ( 0.6

1.2 ( 0.1

0.87 ( 0.12

1.4 ( 0.5

0.10 ( 0.01

0.80 ( 0.04

0.052 ( 0.016

3.2 ( 0.9

0.12 ( 0.01

0.12 ( 0.03

0.19 ( 0.08

0.028 ( 0.002

0.20 ( 0.05

0.22 ( 0.07

0.047 ( 0.007

0.68 ( 0.17

0.88 ( 0.34

0.25 ( 0.03

0.15 ( 0.01

0.21 ( 0.03

0.12 ( 0.03

0.84 ( 0.03

1.9 ( 0.6

2.9 ( 0.1

1.7 ( 0.1

4.6 ( 0.6

12 ( 2

18 ( 5

9.3 ( 0.6

11 ( 3

4.2 ( 1.3

275 ( 57

4.0 ( 2.5

1.9 ( 0.7

77 ( 10

2.7 ( 0.1

1.5 ( 0.2

21 ( 4

1.3 ( 0.3

2.4 ( 0.1

34 ( 11

1.5 ( 0.1

0.47 ( 0.05

2.8 ( 0.3

0.12 ( 0.02

2.7 ( 0.1

0.20 ( 0.03

23 ( 3

0.89 ( 0.06

0.97 ( 0.13

0.70 ( 0.16

4.6 ( 0.6

0.17 ( 0.04

0.32 ( 0.15

4.1 ( 0.5

0.98 ( 0.02

1.4 ( 0.3

0.48 ( 0.11

1.3 ( 0.1

0.10 ( 0.01

1.2 ( 0.2

0.035 ( 0.004

1.5 ( 1.0

0.093 ( 0.035

0.27 ( 0.07

0.43 ( 0.08

0.018 ( 0.004

0.12 ( 0.03

0.26 ( 0.04

0.14 ( 0.06

0.88 ( 0.20

1.2 ( 0.5

0.36 ( 0.07

0.14 ( 0.06

0.36 ( 0.10

0.20 ( 0.07

1.1 ( 0.2

1.1 ( 0.1

2.2 ( 0.2

1.1 ( 0.3

2.3 ( 0.4

5.2 ( 0.7

43 ( 5

9.6 ( 0.7

4.0 ( 1.1

1.7 ( 0.4

330 ( 110

1.5 ( 0.5

1.2 ( 0.4

60 ( 25

1.2 ( 0.3

1.4 ( 0.3

26 ( 6

1.1 ( 0.2

1.6 ( 0.2

22 ( 10

6-SO₂NH₂

6-COMe

6-CONH₂

7-NO₂

-481 ( 8

-353^d

7-SO₂Me

7-CN

-362 ( 8

-385 ( 8

-390 ( 9

-409^d

7-SO₂NH₂

7-COMe

7-CO₂Me

7-CONH₂

8-SO₂Me

8-CN

-429 ( 8

-422 ( 10

-420 ( 8

-419 ( 8

-456 ( 8

-446^d

8-SO₂NH₂

8-COMe

8-CO₂Me

8-CONH₂

9-NO₂

9.0 ( 1.7

0.15 ( 0.04

0.55 ( 0.09

2.8 ( 0.2

-452^d

-447 ( 9

-477 ( 8

1.4 ( 0.3

1.3 ( 0.3

^aDrug concentration to reduce cell density to 50% of that of the controls following 4 h of exposure. Values are the mean ( SEM for two to ten

experiments. ^bOne-electron reduction potential measured by pulse radiolysis. ^cHypoxic cytotoxicity ratio = IC₅₀(oxic)/IC₅₀(hypoxic). Values are the

mean of intraexperiment ratios ( SEM for two to six experiments. ^dCalculated from linear regression analysis.

Results

Chemistry. The new nitroCBIs prepared in this paper

(85). The bromination procedure at the beginning of this

sequence, which provides 86 in one step and 90% yield, is

considerably superior to that previously reported (four steps

and 49% overall yield).^34,35Nitration of 89 yielded the

7-nitro and 9-nitro analogues 90 and 91 (note that with an

unsubstituted A-ring only 3% of the 5-nitro isomer was

produced in this reaction), while acetylation and chlorosul-

fonylation provided the separable 6- and 7-substituted pro-

ducts 92-95. The isomer distribution in the acetylation

reaction was quite sensitive to the conditions, with reaction

in CS₂at 70 °C giving the 6-acetyl isomer as the major

product, while reaction in nitrobenzene at room temperature

favored the 7-acetyl isomer. The position of the substituents

was determined on the basis of 2D NMR experiments

(details are given in the Supporting Information). The

sulfonyl chloride 95 was further converted to the 6-cyano

and 6-amide nitration substrates 97 and 98.

Nitration reactions were either carried out on the trifluoro-

acetyl- (58) or Boc-protected indoline (57) (with the latter

undergoing deprotection under the acidic nitration con-

ditions) as shown in Scheme 1. The reactions were generally

conducted using concentrated H₂SO₄with a slight excess

of KNO₃or, for some of the less reactive 6-substituted

examples, using fuming HNO₃. Where the substituent was

8-SO₂NBn₂, pretreatment with concentrated H₂SO₄was

used to deprotect the dibenzylsulfonamide to give the pri-

mary sulfonamide nitration substrate. In almost all cases the

5-NO₂isomer was the major product, with yields generally

are listed in Tables 1 and 2, and the general route for

their preparation is shown in Scheme 1. In order to obtain

the correct substitution pattern, our previous synthesis

of nitroCBIs (unsubstituted in the A-ring) introduced the

nitro group early, followed by a multistep sequence involv-

ing malonate displacement and Curtius rearrangement to

construct the indoline ring.²⁵However, when the A-ring

bears an EWG, as shown in Scheme 1, this directs nitration

predominantly to the desired 5-position of intermediate 59

or 60. Late introduction of the nitro group allows the use

of previously described radical chemistry to construct the

indoline ring.³²The four-step sequence shown in Scheme 1

(from 53 to 57), involving a modified Curtius rearrange-

ment, bromination, chloroallylation, and radical ring clo-

sure, proved to be consistently high yielding and reliable

and was used in the synthesis of all the nitroCBIs described

here.

Suitable substituted 2-naphthoic acids employed as start-

ing materials in Scheme 1 were known (6-cyano, 6-methoxy-

carbonyl)³³or prepared as shown in Scheme 2 (6-methylsul-

fonyl, 7-methylsulfonyl, 7-cyano, 7-dibenzylaminosulfonyl,

7-acetyl, and 7-methoxycarbonyl).

An alternative route to nitration substrate 57 or 58

employed electrophilic aromatic substitution of the benzin-

doline 89 (Scheme 3), itself prepared by the same radical

ring closure method from tert-butyl 2-naphthylcarbamate

Article

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22 7261

Table 3. Structure and in Vitro Cytotoxicity Data for AminoCBIs

Skov3

HT29

IC₅₀(nM)^a

compd

R

R₁

oxic

hypoxic

HCR^b

oxic

hypoxic

HCR^b

39

40

41

42

43

44

45

46

47

48

49

50

51

52

H

TMI

DEI

1.3 ( 0.1

32 ( 6

10 ( 1

41 ( 5

34 ( 1

19 ( 1

6.8 ( 0.6

14 ( 4

13 ( 4

23 ( 2

11 ( 2

7.7 ( 2.0

13 ( 1

18 ( 4

1.7 ( 0.2

25 ( 1

9.0 ( 1.9

33 ( 6

27 ( 5

14 ( 2

6.6 ( 0.6

8.2 ( 0.1

12 ( 2

1.0 ( 0.1

1.2 ( 0.2

1.3 ( 0.2

1.3 ( 0.1

1.2 ( 0.2

1.4 ( 0.5

0.97 ( 0.05

1.7 ( 0.5

1.1 ( 0.2

1.4 ( 0.1

1.4 ( 0.2

0.91 ( 0.21

1.0 ( 0.1

1.5 ( 0.2

0.86 ( 0.31

32 ( 3

15 ( 2

86 ( 18

59 ( 7

25 ( 2

5.0 ( 0.6

11 ( 1

17 ( 1

2.5 ( 0.3

36 ( 5

11 ( 4

91 ( 26

28 ( 11

28 ( 5

3.5 ( 0.8

7.3 ( 2.0

8.6 ( 2.0

23 ( 11

5.6 ( 1.7

5.7 ( 1.8

48 ( 15

16 ( 1

0.76 ( 0.07

0.86 ( 0.14

1.9 ( 0.7

1.1 ( 0.5

2.3 ( 0.7

1.0 ( 0.2

1.3 ( 0.1

1.6 ( 0.4

2.0 ( 0.4

1.8 ( 0.6

1.5 ( 0.3

1.7 ( 0.7

1.4 ( 0.3

1.9 ( 0.2

7-SO₂Me

7-CN

7-SO₂NH₂

7-CONH₂

8-SO₂Me

H

7-SO₂Me

7-CN

7-SO₂NH₂

7-COMe

7-CO₂Me

7-CONH₂

8-SO₂Me

16 ( 1

35 ( 6

8.2 ( 2.7

8.3 ( 0.3

13 ( 1

7.6 ( 0.7

8.6 ( 0.8

61 ( 13

29 ( 1

12 ( 4

^aDrug concentration to reduce cell density to 50% of that of the controls following 4 h of exposure. Values are the mean ( SEM for two to seven

experiments. ^bHypoxic cytotoxicity ratio = IC₅₀(oxic)/IC₅₀(hypoxic). Values are the mean of intraexperiment ratios ( SEM for two to four

experiments.

higher for 7-substituents or for more electron withdrawing

substituents. Some other nitro isomers were also isolated: the

7-NO₂byproduct for 8-SO₂Me, 8-SO₂NH₂, and 9-NO₂and

the 9-NO₂byproduct for 6-CONH₂and 7-CO₂Me. The

position of the nitro group was firmly established by X-ray

crystallography (for 60 where EWG = 8-SO₂Me), 2D NMR

experiments, and comparison of the NMR spectra of the

remaining compounds to these reference examples (details in

Supporting Information).³⁶

Where the nitration was carried out on the trifluoroace-

tamide, this was subsequently cleaved in practically quanti-

tative yield by mild base treatment (Scheme 1, 59 to 60).

For the two sulfonyl chlorides (59, 6- and 7-SO₂Cl) prior

treatment with concentrated NH₃served to generate the

corresponding sulfonamides. Two nitrile examples (99 and

100) were further hydrolyzed to the amides (101 and 102) by

exposure to aqueous H₂SO₄(Scheme 4).

The only exception to this general route to intermediates

60 was for the 6-NO₂example 105 (Scheme 5), which was

prepared by nitration of 104, itself formed by deprotection of

the previously reported 103. The structure of the dinitro

product 105 was established by X-ray crystallography (see

Supporting Information).

Two different minor groove binding side chains were

employed, TMI and DEI as previously incorporated in the

A-ring unsubstituted nitroCBIs 6 and 19. The side chains

were introduced using the indole-2-carboxylic acids and

EDCI [N-ethyl-N⁰-(3-dimethylaminopropyl)carbodiimide

hydrochloride] as a coupling reagent (or for some of the

TMI compounds, using the corresponding acid chloride).

The EDCI reactions were found to proceed best under acidic

conditions, which were achieved by preforming the indoline

hydrochloride salt or by adding anhydrous TsOH to the

reaction mixture. Excess side chain acid could be removed

following the reaction by washing with dilute aqueous

NaHCO₃, but care is required in this step, since the products

are susceptible to elimination of HCl.

Several of the nitroCBIs were reduced to the correspond-

ing aminoCBIs (Scheme 1) as listed in Table 3. A PtO₂

catalyst was employed to avoid hydrogenolysis of the chlor-

ide.

Reduction Potentials. One-electron reduction potentials,

E(1), were determined in aqueous solution by measuring the

equilibrium constants, established within a few tens of

microseconds, for electron transfer between the radical

anions of the compounds and reference standards. Only

the more water-soluble DEI analogues of Table 2 were

investigated, and even some of these, including all acetyl

substituted examples, were too insoluble to permit reduction

potential measurement.

The experimentally determined values are plotted in Fig-

ure 3 as a function of σ_p³⁷of the A-ring substituent. While all

substituents raised E(1) above that of 19 (-512 mV), the

effect was strongest when the substituent was in the 7-

position [4E(1) = 15-93 mV for 6-, 8-, 9-substituents but

83-150 mV for 7-substituents]. The measured reduction

potentials exhibited a reasonable correlation with σ_pwhen

the substituent was in either the 7- or 8-position (r²= 0.97

and 0.88, respectively), and the regression lines shown in

Figure 3 were used to calculate E(1) for the poorly soluble

analogues.³⁸The reduction potential was much less sensitive

to the electron-withdrawing nature of a substituent in the 6-

position (or for the single example of a 9-substituted ana-

logue), with even the most electron-deficient nitro substitu-

ent only raising the reduction potential by 35-45 mV

compared to 19.

In Vitro Cytotoxicity. Our previous investigation of

the cytotoxicity of 6 and 19 under oxic and hypoxic condi-

tions used a clonogenic survival assay, but in order to

compare a large number of analogues in the present study,

7262 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22

Tercel et al.

Scheme 1^a

Scheme 2^a

^aConditions: (a) Me₂NC(S)Cl, DABCO; (b) 225 °C; (c) KOH, then

Me₂SO₄; (d) NaBO₃, AcOH; (e) BuLi, (MeS)₂, -78 °C; (f) BuLi, CO₂,

-78 °C; (g) CuCN, py, NMP; (h) Pd(OAc)₂, DPPP, CO(g), Et₃N,

MeOH; (i) NaOH; (j) BuLi, SO₂(g), -78 °C, then NCS; (k) Bn₂NH,

Et₃N; (l) KOH; (m) Ac₂O, AlCl₃.

hydrolysis of the “preprodrug” 2) exhibit HCRs of 75 and

5-10, respectively, in these two cell lines.¹³

For the new nitroCBI analogues both oxic and hypoxic

cytotoxicity were found to be highly variable and sensitive to

the nature of the A-ring substituent and side chain. The DEI

side chain confers greater cytotoxicity than TMI under both

oxic and hypoxic conditions (on the order of a 10-fold

differential under oxic conditions). Further, oxic and hy-

poxic cytotoxicity both vary over about 2 orders of magni-

tude depending on the substituent, with some compounds

being significantly more toxic (e.g., 6-COMe) and some

significantly less toxic (e.g., 6-CONH₂) than the unsubsti-

tuted parent. Neither oxic nor hypoxic cytotoxicity corre-

lates with reduction potential (illustrated in Supporting

Information Figure S1 for the DEI compounds).

These trends can be seen from the data in Tables 1 and

2 and also in Figure 4 where IC₅₀values under oxic versus

hypoxic conditions are plotted for both cell lines and both

side chains. Figure 4 also depicts HCR as displacement from

the solid line which indicates equivalent toxicity under both

conditions. For the TMI compounds, the HCR values are

generally modest, with only the 7-NO₂substituent generat-

ing an HCR greater than 10 (indicated by the dotted line) in

^aConditions: (a) DPPA, Et₃N, ^tBuOH; (b) NBS; (c) NaH, 1,3-

dichloropropene; (d) Bu₃SnH, AIBN; (e) HCl, then TFAA, py; (f) conc

H₂SO₄, KNO₃, or fuming HNO₃(for 6-CN, 6-COMe, 6-CONH₂);

(g) Cs₂CO₃; (h) 5,6,7-trimethoxyindole-2-carbonyl chloride, py, DMAP,

or 5,6,7-trimethoxyindole-2-carboxylic acid or 5-[2-(dimethylamino)-

ethoxy]indole-2-carboxylic acid, EDCI, TsOH; (i) H₂, PtO₂.

an antiproliferative assay using a 96-well plate format was

used instead. Oxic and hypoxic cytotoxicities were deter-

mined as IC₅₀values after 4 h of drug exposure in two human

tumor cell lines: the ovarian carcinoma Skov3 and the colon

carcinoma HT29. 6 and 19 had previously demonstrated

only modest hypoxia-selective cytotoxicity in HT29, and a

similar result was noted using the IC₅₀-based assay (Tables 1

and 2, HCR=2.5, 1.1). The Skov3 cell line, which had not

previously been examined, also exhibited little hypoxia-

selective sensitivity to these two compounds. In compari-

son, 1 and 3 (PR-104A, the alcohol derived by phosphate

Article

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22 7263

Scheme 3^a

Figure 3. Measured one-electron reduction potentials E(1) as a

function of σ_pfor the compounds of Table 2. Equations for the

regression lines for 6-, 7-, and 8-substituents are given in the

Experimental Section.

Scheme 5^a

^aConditions: (a) HCl; (b) conc H₂SO₄, KNO₃.

HCRs of 275 and 330 in Skov3 and HT29, respectively. For

this subset the observed HCR appears to correlate with E(1),

as shown for the Skov3 data in Figure 5, r²= 0.96 (for HT29,

r²= 0.75; see Supporting Information Figure S2), with the

highest HCR being observed for the substituent with the

highest one-electron reduction potential.

The cytotoxicity of several amino compounds was also

determined, and the results are collected in Table 3. For all of

these compounds the addition of an A-ring substituent

decreases the cytotoxicity compared to the unsubstituted

parent but leads to little variation in effect; all the A-ring

substituted aminoCBIs investigated remain potent cytotox-

ins with IC₅₀values in the 5-90 nM range. None of them

display any hypoxic selectivity. The superior hypoxic selec-

tivity of nitroCBIs bearing amide and sulfonamide sub-

stituents does not appear to be related to an intrinsically

more toxic reduction product, since the two substituents

that generate the highest HCRs as nitroCBIs (7-SO₂NH₂,

7-CONH₂) are associated with the weakest cytotoxicity as

aminoCBIs. This can also be seen in Figure 6 where cyto-

toxicities of nitroCBI and aminoCBI compounds are com-

pared in the HT29 cell line for all examples where the

aminoCBI-DEI compounds were available. In this compar-

ison the 7-SO₂NH₂and 7-CONH₂substituents stand out as

being those where the nitroCBI under hypoxia becomes

approximately as cytotoxic as the corresponding aminoCBI.

While there is a very large variation in nitroCBI cytotoxi-

city dependent on the A-ring substituent, the effect is con-

sistent for the two cell lines examined, with correlation

coefficients of 0.95 for IC₅₀values obtained under both oxic

and hypoxic conditions (Supporting Information Figure S3).

This suggests that there is a common mechanism of toxicity

for nitroCBIs in these two cell lines. A weaker correlation,

^aConditions: (a) NBS; (b) NaH, 1,3-dichloropropene; (c) Bu₃SnH,

AIBN; (d) HCl, then TFAA, py; (e) fuming HNO₃, 20 °C, 5 min;

(f) AcCl, AlCl₃, CS₂, 70 °C, 3 h (gives 47% 93) or PhNO₂, 0 °C, 16 h

(gives 20% 92 and recovered 89); (g) ClSO₃H; (h) ZnI₂, LiCl, PdCl₂-

(PhCN)₂, Ti(OiPr)₄; (i) KCN, Pd(PPh₃)₄, CuI; (j) 90% H₂SO₄.

Scheme 4^a

^aConditions: (a) 90% H₂SO₄, 65 °C, 1 h.

both cell lines. For the DEI compounds a different pattern

emerges. While most of the substituents again display mod-

est hypoxic selectivity, a subset containing the primary

amides and sulfonamides gives large HCRs in both cell lines.

The greatest effect is seen for the 7-SO₂NH₂substituent with

7264 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22

Tercel et al.

Figure 4. Comparison of nitroCBI cytotoxicity measured under oxic and hypoxic conditions. The TMI compounds are from Table 1 and

DEI compounds from Table 2. The solid line represents equivalent toxicity under both conditions (HCR = 1) while the dotted line represents

HCR = 10. Arrows indicate that the IC₅₀is above the solubility limit.

Figure 5. Relationship between hypoxic cytotoxicity ratio (HCR)

in Skov3 and one-electron reduction potential [E(1)] for the com-

pounds of Table 2. The A-ring substituents are divided into two

groups: (b) 6-, 7-, and 8-SO₂NH₂and 6-, 7-, and 8-CONH₂(linear

regression gives r²= 0.96); (O) all other substituents (linear regres-

sion gives r²= 0.23).

Figure 6. Comparative cytotoxicity of nitroCBI-DEI compounds

(19, 26-30, 32) under oxic and hypoxic conditions and the matching

aminoCBI-DEI compounds (45-52) under hypoxic conditions in

HT29.

r²=0.75, is observed for the smaller group of aminoCBIs

investigated, suggesting that they also have a common

mechanism of toxicity in the two cell lines.

Cytotoxicity in a panel of human tumor cell lines was

examined for the most hypoxia-selective agent 28 as shown in

Table 4. All cell lines exhibited an oxic IC₅₀in the low

micromolar range (0.5-7 μM) which was significantly en-

hanced under hypoxia (Figure 7), reaching low nanomolar

levels in several cell lines. The panel includes a cell line pair

that differ only in the level of expression of NADPH/

cytochrome P450 oxidoreductase,²³with the overexpressing

cell line being more sensitive to 28 under hypoxic but not oxic

conditions. All 11 cell lines examined display significant

hypoxic selectivity (HCR 19-330) to this compound. When

compared to 1 and 3 (Supporting Information Figure S4), it

is clear that 28 is a significantly more potent cytotoxin under

Article

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22 7265

Table 4. Cytotoxicity and Hypoxic Selectivity of NitroCBI 28 in a Panel

of Human Tumor Cell Lines

IC₅₀(μM)^a

tissue

colon

cervix

cell line

oxic

hypoxic

HCR^b

HCT 116

C33A

SiHa

1.7 ( 0.4

0.015 ( 0.001

110 ( 20

0.50 ( 0.02 0.0030 ( 0.0011 240 ( 110

1.7 ( 0.4

0.95 ( 0.14 0.035 ( 0.013

0.034 ( 0.008

71 ( 8

30 ( 7

NSCLC

A549

0.86 ( 0.22 0.0060 ( 0.0017 140 ( 10

P450R^{puro c}

H1299

H460

6.3 ( 0.2

3.8 ( 0.4

2.9 ( 0.3

2.9 ( 0.6

0.11 ( 0.08

0.11 ( 0.01

0.037 ( 0.026

0.15 ( 0.01

180 ( 100

36 ( 4

180 ( 80

19 ( 3

LCLC

prostate

melanoma

PC3

A375

^aDrug concentration to reduce cell density to 50% of that of the

Figure 8. HPLC chromatograms following incubation of 10 μM

nitroCBI 19 (A and C) and 28 (B and D) with mouse liver S9 (protein

concentration of 8 mg/mL) at 37 °C for 30 min under either hypoxic

(A and B) or oxic (C and D) conditions. HPLC conditions were

slightly different between the oxic and hypoxic samples. The peaks

marked with an asterisk are tentatively identified as the products of

N-oxidation of the dimethylamino side chain (see text and Support-

ing Information).

controls following 4 h of exposure. Values are the mean ( SEM for two

to six experiments. ^bHypoxic cytotoxicity ratio = IC₅₀(oxic)/IC₅₀

-

(hypoxic). Values are the mean of intraexperiment ratios ( SEM for

two to four experiments. ^cStable transfectant of A549 overexpressing

human NADPH/cytochrome P450 oxidoreductase (9-fold higher rate of

cytochrome c reduction than parent cell line) (ref 23).

Figure 7. Cytotoxicity of nitroCBI 28 under oxic and hypoxic

conditions in a panel of human tumor cell lines. A549* is the

A549 P450R^purocell line of Table 4.

hypoxic conditions than either of these established HAPs (up

to 300-fold more potent than 3 and 500-fold more potent

than 1). In addition, 28 exhibits hypoxic cell selectivity equal

to or greater than 3 in each cell line examined.

Figure 9. Clonogenic survival curves for HT29 single cell suspen-

sions (circles) and intact spheroids (squares) incubated with 28 for 4

h under either oxic (open symbols) or hypoxic (closed symbols)

conditions. Each point is a separate cell or spheroid suspension, and

the data are pooled from four experiments.

In Vitro Metabolism. To investigate whether the superior

HCR of 28 compared to 19 is due to enhanced reduction

under hypoxic conditions, the metabolism of these two

compounds was examined using subcellular fractions pre-

pared from mouse livers and HT29 xenografts. Postmito-

chondrial S9 fractions were supplemented with NADPH and

incubated with the nitroCBIs under either oxic or hypoxic

conditions. LC/MS analysis showed that under hypoxic

conditions the only significant metabolite was the corre-

sponding aminoCBI (illustrated for liver S9 in Figure 8A,

B), which was identified by mass spectrometry, UV-visible

spectroscopy, and comparison with authentic samples (see

Supporting Information Figure S5). Under otherwise iden-

tical incubation conditions there was clearly more extensive

reduction of the sulfonamide-substituted nitroCBI 28 com-

pared to the A-ring unsubstituted 19. Qualitatively similar

results were observed using S9 derived from HT29 tumors

(data not shown), although higher protein concentrations

and longer incubation times were required to achieve a

similar level of metabolism. Control experiments showed

that there was no reaction in the absence of cofactor or if the

S9 was inactivated by prior heat treatment.

Under oxic conditions no aminoCBI was produced, and

the significant new metabolite (marked with an asterisk in

Figure 8C,D), which shared the same UV-visible spectrum

as the parent nitroCBI but with a mass spectrum 16 m/z units

higher (Supporting Information), was tentatively identified

as the corresponding N-oxide of the dimethylamino side

chain. Similar levels of this metabolite were observed for

each of the two nitroCBIs.

Cytotoxicity of 28 in Dissociated and Intact Spheroids. The

cytotoxicity of 28 was further evaluated in HT29 cells using a

clonogenic assay. In single cell suspensions (circles in Figure

9) there was a clear dependence of cytotoxicity on the nature

of the gas phase; under oxic conditions the surviving fraction

remained >10% even at the solubility limit of 30 μM, while

under hypoxic conditions a C₁₀(concentration for 1 log of

cell kill) of 0.40 μM was calculated from the fitted curve

shown in Figure 9. Extrapolation of the oxic data suggests

7266 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22

Tercel et al.

This is a direct consequence of the presence of the EWG,

since it directs nitration to the desired 5-position of inter-

mediates 57 and 58, allowing preparation of the indoline ring

by a well-precedented and reliable radical route. As an

illustration, our previous synthesis of 6 proceeded in 14 steps

and 4% overall yield,²⁵whereas 14 can be obtained from

6-(methoxycarbonyl)-2-naphthoic acid in six steps and 12%

overall yield. While the other appropriately substituted

2-naphthoic acids are not commercially available, they can

be obtained in a few steps on a multigram scale (as previously

described or as reported in Scheme 2), as can the alternative

nitration substrates illustrated in Scheme 3. The key nitration

step is not completely selective, but in all cases the desired

5-nitro isomer could be readily purified by recrystallization or

column chromatography.

Measurement of one-electron reduction potentials for the

more water-soluble DEI analogues established that the un-

substituted nitroCBI 19 has E(1) = -512 mV. This is outside

the window of -400 to -200 mV predicted to be suitable for

the observation of hypoxic selectivity with nitroaromatic

compounds, in that it may be too negative to allow efficient

enzymatic reduction.^39,40As a comparison, the nitrogen

mustard N,N-bis(2-chloroethyl)-4-nitroaniline, proposed as

a potential HAP, was calculated to have E(1) = -508 mV,

and it was not until analogues bearing extra EWGs were

prepared that significant HCRs were achieved for these

compounds in vitro.^41,42The addition of any EWG to the

A-ring of the nitroCBI raises the reduction potential (Table 2

and Figure 3), but the magnitude is sensitive to the substituent

location. A variety of EWGs in the 6-position only raise E(1)

by a maximum of 45 mV, even for the strongly electron-

deficient NO₂substituent. It is possible that steric interaction

with the 5-NO₂group twists the 6-substituent out of plane

with the aromatic ring, diminishing the effective conjugation.

A similar effect may also explain the weak impact that a

9-NO₂substituent has on E(1). When the EWG is in the 8- or

especially 7-position, a larger effect on reduction potential is

observed and found to correlate with σ_pof the substituents

examined. Four examples of 7-substituents (NO₂, SO₂Me,

CN, and SO₂NH₂) were found (or calculated) to provide

nitroCBIs with one-electron reduction potentials above

-400 mV.

The A-ring substituents also have a substantial effect on the

cytotoxicity of the nitroCBIs under both oxic and hypoxic

conditions. It seems reasonable to expect that such substitu-

ents will influence the binding of the compounds to the minor

groove of DNA and that bulky substituents positioned at the

base of the groove (for example, 9-NO₂or 8-SO₂Me) may

hinder binding, while smaller or planar substituents that can

enhance van der Waals contacts could have the opposite

effect. Even with these considerations it is not obvious why

6-substituents, directed toward the mouth of the minor

groove, have such a large and disparate effect on IC₅₀; a

6-COMe substituent confers greater toxicity than the unsub-

stituted parent(compare 8 with6 and compare23with19) and

a 6-CONH₂substituent the opposite (compare 24 with 19), to

the point where 23 is approximately 100 times more cytotoxic

than 24 under oxic conditions. At this point the mechanism of

oxic toxicity for the nitroCBI compounds is not known;

possibilities include noncovalent DNA binding, direct DNA

alkylation, redox cycling, and oxygen-insensitive reductive

metabolism (or a combination of these). However, it is

interesting to note that neither oxic nor hypoxic cytotoxicity

correlate with reduction potential (Figure S1).

Figure 10. Antitumor activity of 28 in SiHa human tumor xeno-

grafts assayed by tumor excision and clonogenic assay 18 h after iv

dosing with nitroCBI (42 μmol/kg) only or nitroCBI plus γ irradia-

tion (15 Gy). Values are the mean ( SEM for groups of three mice

(control or 28 alone) or five mice (radiation alone or radiation þ 28).

that the HCR for 28 by clonogenic assay is approximately

125. The large improvement compared to 19 (HCR = 2.5 in

this cell line³¹) appears to be due to the increased sensitivity

of HT29 cells to 28 under hypoxic conditions, as also

observed in the IC₅₀assay (Table 2).

The single cell suspensions described above were gener-

ated by dissociation of HT29 spheroids. When the intact

spheroids themselves were incubated with 28 under the same

conditions, there was again a hypoxia-selective response

(squares in Figure 9). Under oxic conditions no significant

cell kill was achieved even up to the solubility limit, suggest-

ing that the multicellular structure provides some resistance

to nitroCBI-induced aerobic toxicity. In contrast, under

hypoxic conditions, the survival curves for spheroids and

single cell suspensions were superimposable. One possible

explanation for these results is that a reduction product is

formed under hypoxic conditions that is capable of diffusing

from and killing cells distinct from those in which it is

generated.

In Vivo Activity. To examine whether 28 has therapeutic

activity in vivo, the compound was formulated in a

DMSO-lactate buffer-PEG vehicle and administered as a

single dose to mice. By ip administration, no toxicity was

evident up to 100 μmol/kg, but at this dose there was clear

precipitation in the peritoneum at the site of injection. As an

alternative, an iv dose of 42 μmol/kg was administered to

SiHa tumor-bearing mice either alone or 5 min after a single

15 Gy dose of ionizing radiation. Eighteen hours later the

tumors were excised and surviving clonogens assessed

(Figure 10). An irradiation dose of 15 Gy is sufficient to

sterilize aerobic tumor cells and provided 1.79 logs of cell kill

compared to control in this experiment. When 28 was

combined with radiation, a greater cell kill was achieved

(2.70 log), indicating elimination of radioresistant hypoxic

cells within the tumor by the nitroCBI prodrug. As a single

agent, 28 exhibited weak but not statistically significant

activity.

Discussion and Conclusions

This study was undertaken to test the idea that the addition

of an EWG to the A-ring of 6 or 19 would raise the reduction

potential and provide HAPs with enhanced reductive meta-

bolism and improved hypoxia-selective cytotoxicity. While

the target compounds are structurally more complex than the

unsubstituted parents, their syntheses are generally easier.

Article

The nature of the minor groove binding side chain (TMI

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22 7267

Experimental Section

versus DEI) also has an effect on cytotoxicity. Although poor

aqueous solubility precluded reduction potential measure-

ment for the TMI compounds, it is considered highly unlikely

that the side chain has any significant impact on E(1). DEI

nitroCBIs may be more toxic than the TMI compounds

because of enhanced DNA binding by the positively charged

basic side chain. The difference is smaller when comparing

aminoCBI compounds (here a DEI side chain generates only

marginally more toxicity than TMI), and it is possible that the

basic side chain affects nitroCBI toxicity by influencing cell

uptake or subcellular distribution relative to the enzyme(s)

responsible for reductive activation.

General Chemistry Methods. All final products were analyzed

by reverse-phase HPLC (Alltima C18 5 μm column, 150 mm ꢀ

3.2 mm; Alltech Associated, Inc., Deerfield, IL) using an Agilent

HP1100 equipped with a diode array detector. Mobile phases

were gradients of 80% acetonitrile/20% H₂O (v/v) in 45 mM

ammonium formate at pH 3.5 and 0.5 mL/min. Purity was

determined by monitoring at 330 ( 50 nm and was >95% in all

cases. Final product purity was also assessed by combustion

analysis carried out in the Campbell Microanalytical Labora-

tory, University of Otago, Dunedin, New Zealand. Melting

points were determined on an Electrothermal 2300 melting

point apparatus. NMR spectra were obtained on a Bruker

1

Avance 400 spectrometer at 400 MHz for H and 100 MHz

for ¹³C spectra.

Whatever the explanation for the cytotoxicity of individual

compounds under particular conditions, it is clear that a

subset of the nitroCBIs investigated have substantial hypoxic

selectivity. These compounds (with the exception of 9) are

those with a DEI side chain and a primary sulfonamide or

carboxamide in one of the 6-, 7-, or 8-positions on the A-ring.

For this set hypoxic selectivity increases with increasing

reduction potential (Figure 5), although significant HCRs

(>10 in both cell lines) are observed for several compounds

with E(1) well below -400 mV (24, 31, 34, 37). Higher

reduction potential alone is not sufficient to observe hypoxic

selectivity; the 7-substituted analogues 27 and 28 have the

same E(1) within experimental error, but the former (7-CN)

gives HCRs of 1.7 and 4.2 in the two cell lines tested, while the

latter (7-SO₂NH₂), the most selective of all the compounds

surveyed, has HCRs of 275 and 330. It is possible that the

H-bond donor capacity of the sulfonamide and carboxamide

substituents makes them better substrates for the reductase(s)

responsible for hypoxic activation, and when H-bond donor

capacity is combined with higher reduction potential, then

highly selective compounds are produced.

Illustrative examples are provided for the synthesis of all

6-substituted nitroCBI final products of Tables 1 and 2, and an

example of the general method for the synthesis of an aminoCBI

of Table 3 is provided. The preparation of all other compounds

is fully documented in the Supporting Information.

Preparation of 1-(Chloromethyl)-5,6-dinitro-3-(5,6,7-trime-

thoxyindol-2-carbonyl)-1,2-dihydro-3H-benzo[e]indole (7). 1-(Chloro-

methyl)-5-nitro-1,2-dihydro-3H-benzo[e]indole (104). A solu-

tion of tert-butyl 1-(chloromethyl)-5-nitro-1,2-dihydro-3H-

benzo[e]indole-3-carboxylate²⁵(103) (600 mg, 1.65 mmol) in

dioxane (15 mL) was saturated with dry HCl, stirred at 20 °C for

1 h, and then evaporated under reduced pressure below

30 °C. The residue was partitioned between CH₂Cl₂and dilute

aqueous KHCO₃and the organic phase was washed with water,

dried, and then filtered through a column of silica gel to give 104

(372 mg, 86%) as a red solid: mp (CH₂Cl₂/petroleum ether)

100-101 °C; ¹H NMR [(CD₃)₂SO] δ 8.11 (d, J = 8.7 Hz, 1 H),

7.87 (d, J = 8.4 Hz, 1 H), 7.65 (s, 1 H), 7.55 (ddd, J = 8.2, 7.0, 1.2

Hz, 1 H), 7.40 (ddd, J = 8.7, 6.8, 1.0 Hz, 1 H), 6.27 (br s, 1 H),

4.23-4.15 (m, 1 H), 3.89 (dd, J = 11.0, 3.7 Hz, 1 H), 3.81 (t,

J = 9.7 Hz, 1 H), 3.78-3.66 (m, 2 H). Anal. (C₁₃H₁₁ClN₂O₂)

C, H, N.

Further investigation of 28 shows that it has a number

of attractive features as a HAP and that the improve-

ments compared to the previously reported 19 are consistent

with enhanced metabolism under hypoxic but not oxic con-

ditions. Metabolism studies using S9 fractions from liver or

tumor (Figure 8) show conversion to the corresponding

aminoCBI as the sole product under hypoxic incubation, a

reaction that is completely suppressed in the presence of

oxygen. There is clearly more extensive reductive metabolism

of 28 than 19, so much so that after only 4 h of hypoxic

incubation 28 becomes as cytotoxic to HT29 (Figure 6)

or Skov3 cells (Tables 2 and 3) as directly administered

aminoCBI 48, which is clearly not the case for 19. Although

the enzyme(s) responsible for the reductive activation of 28

have not been identified, beyond the implication of NADPH/

cytochrome P450 oxidoreductase, they appear to be widely

expressed, since 28 displays substantial hypoxic selectivity in

every cell line so far investigated (Figure 7). Hypoxic selectiv-

ity is also demonstrated using clonogenicity as the end point

(Figure 9).

The major limitation remaining for the application of

nitroCBI 28 as a HAP is its poor water solubility. Although

28 could be formulated for in vivo analysis, ip administration

to mice led to precipitation in the peritoneum. When dosed iv,

28 produced a highly significant increase in tumor cell kill in

combination with radiation. This encouraging result suggests

that analogues of nitroCBI28thatretainthe hypoxia-selective

activity in combination with improved water solubility would

be very interesting candidates as hypoxia-activated prodrugs

for antitumor therapy.

General Method of Nitration Using H₂SO₄and KNO₃. Pre-

paration of 1-(Chloromethyl)-5,6-dinitro-1,2-dihydro-3H-benzo-

[e]indole (105). A stirred solution of 104 (500 mg, 1.90 mmol) in

concentrated H₂SO₄(5 mL) was cooled to -5 °C and treated

with powdered KNO₃(288 mg, 2.85 mmol). The mixture was

stirred at 0 °C for a further 15 min, then poured into ice-water,

and the solid was collected and dissolved in CH₂Cl₂. The

solution was filtered through a column of silica gel and the

product was recrystallized from EtOAc/ⁱPr₂O to give 105 (446

mg, 76%) as a red solid: mp 206-207 °C; ¹H NMR [(CD₃)₂SO] δ

8.23 (dd, J = 8.7, 1.0 Hz, 1 H), 8.00 (dd, J = 7.7, 0.9 Hz, 1 H),

7.76 (s, 1 H), 7.67 (dd, J = 8.4, 7.6 Hz, 1 H), 6.72 (s, 1 H),

4.32-4.22 (m, 1 H), 3.94-3.83 (m, 2 H), 3.83-3.75 (m, 2 H).

Anal. (C₁₃H₁₀ClN₃O₄) C, H, N.

General Method of Amide Formation Using 5,6,7-Trimethoxy-

indole-2-carbonyl Chloride. Preparation of 7. A suspension of

5,6,7-trimethoxyindole-2-carboxylic acid (122 mg, 0.49 mmol)

in dry CH₂Cl₂(10 mL) was treated with oxalyl chloride (0.13

mL, 1.49 mmol) followed by DMF (10 μL). The mixture was

stirred at room temperature for 15 min, then evaporated under

reduced pressure and azeotroped dry with benzene. The result-

ing acid chloride was cooled to -5 °C and treated with an ice-

cold solution of amine 105 (100 mg, 0.33 mmol) in dry pyridine

(2 mL) containing DMAP (5 mg). The stirred mixture was

warmed to room temperature for 30 min, then poured into

dilute aqueous KHCO₃. The solid was collected, purified by

chromatography on silica gel, eluting with CH₂Cl₂/EtOAc

(19:1), then crystallized from CH₂Cl₂/EtOAc to give 7 (84 mg,

48%) as a yellow solid: mp 278-279 °C; ¹H NMR [(CD₃)₂SO] δ

11.67 (s, 1 H), 9.16 (s, 1 H), 8.61 (d, J = 8.0 Hz, 1 H), 8.38 (d, J =

7.4 Hz, 1 H), 7.92 (t, J = 8.0 Hz, 1 H), 7.21 (d, J = 1.9 Hz, 1 H),

6.99 (s, 1 H), 4.94 (t, J = 10.6 Hz, 1 H), 4.73-4.60 (m, 2 H),

7268 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22

Tercel et al.

4.19-4.05 (m, 2 H), 3.94 (s, 3 H), 3.83 (s, 3 H), 3,81 (s, 3 H).

HRMS (FAB) calcd for C₂₅H₂₁³⁵ClN₄O₈(M^þ) m/z 540.1048,

found 540.1051. Anal. (C₂₅H₂₁ClN₄O₈) C, H, N.

dissolved in CH₂Cl₂at room temperature, dried, and concen-

trated under reduced pressure below 30 °C. The residue was

triturated with EtOAc to give crude 20. Treatment of a solution

of the free base in CH₂Cl₂with HCl_(g)/EtOAc/hexane, followed

Preparation of 6-Acetyl-1-(chloromethyl)-5-nitro-3-(5,6,7-

trimethoxyindol-2-carbonyl)-1,2-dihydro-3H-benzo[e]indole (8).

7-Acetyl-1-(chloromethyl)-3-(trifluoroacetyl)-1,2-dihydro-3H-

benzo[e]indole (92) and 6-Acetyl-1-(chloromethyl)-3-(trifluoro-

acetyl)-1,2-dihydro-3H-benzo[e]indole (93). Solid 89 (4.7 g,

15 mmol) was added to a mixture of AlCl₃(7.0 g, 52 mmol)

and AcCl (2.5 mL, 35 mmol) in CS₂(60 mL) at 0 °C, and the

stirred mixture was heated at 70 °C for 3 h. Solvent was boiled

off at 60 °C, and the black residue was cooled and treated with

ice and concentrated HCl. The mixture was extracted with

CH₂Cl₂(3 ꢀ 100 mL). The extracts were dried and concen-

trated under reduced pressure, and the residue was chroma-

tographed on silica gel. Elution with EtOAc/petroleum ether

(1:4) gave a product (3.9 g, 73%) that was shown by NMR to

be a mixture of 64% 93 and 23% 92, with the remainder of the

material being a mixture of other acetylated products. Pure 93

was obtained by crystallization from EtOAc/petroleum ether

as a white solid: mp 121-123 °C; ¹H NMR [(CD₃)₂SO] δ 8.72

(d, J = 1.6 Hz, 1 H), 8.60 (d, J = 9.4 Hz, 1 H), 8.38 (d, J = 9.3

Hz, 1 H), 8.25 (d, J = 8.4 Hz, 1 H), 8.00 (dd, J = 7.1, 0.9 Hz,

1 H), 7.70 (dd, J = 8.3, 7.3 Hz, 1 H), 4.60-4.40 (m, 3 H),

4.18-4.10 (m, 1 H), 4.07-3.99 (m, 1 H), 2.76 (s, 3 H);

¹³C NMR δ 201.8, 153.2 (q, J_C-F= 36.9 Hz), 139.9, 136.1,

129.6, 128.1, 127.8, 127.5, 127.2, 126.8, 126.3, 124.9, 116.1 (q,

J_C-F= 288 Hz), 52.5, 47.6, 41.1, 30.0. Anal. (C₁₇H₁₃ClF₃-

NO₂) C, H, N.

6-Acetyl-1-(chloromethyl)-5-nitro-3-(trifluoroacetyl)-1,2-di-

hydro-3H-benzo[e]indole (59, EWG = 6-COMe). 93 was ni-

trated using fuming HNO₃as described in the general method

below to give 59 (EWG = 6-COMe) (57%) as a brown solid: mp

182-184 °C (EtOAc/petroleum ether); ¹H NMR (CDCl₃) δ 9.18

(s, 1 H), 9.06 (d, J = 1.4 Hz, 1 H), 8.28 (dd, J = 8.8, 1.6 Hz,

1 H), 7.95 (d, J = 8.8 Hz, 1 H), 4.68-4.63 (m, 1 H), 4.57-4.49

(m, 1 H), 4.48-4.30 (m, 1 H), 3.93-3.87 (m, 1 H), 3.65-3.58 (m,

1 H), 2.70 (s, 3 H); ¹³C NMR δ 200.4, 154.7 (q, J_C-F= 39.2 Hz),

148.7, 139.2, 138.2, 130.9, 130.4, 127.7, 127.5, 125.9, 119.6,

115.7, 115.6 (q, J_C-F= 288 Hz), 52.7, 45.4, 42.7, 28.5. Anal.

(C₁₇H₁₂ClF₃N₂O₄) C, H, N.

by crystallization from MeOH/Me₂CO/EtOAc, gave 20 HCl

3

1

(129 mg, 69%) as a yellow solid: mp 225-226 °C; H NMR

[(CD₃)₂SO] δ 11.88 (d, J = 1.6 Hz, 1 H), 10.12 (br s, 1 H), 9.22 (s,

1 H), 8.63 (d, J = 7.9 Hz, 1 H), 8.40 (dd, J = 7.6, 0.6 Hz, 1 H),

7.93 (t, J = 8.0 Hz, 1 H), 7.47 (d, J = 8.9 Hz, 1 H), 7.27 (d, J =

2.3 Hz, 1 H), 7.26 (d, J = 1.6 Hz, 1 H), 7.04 (dd, J = 8.9, 2.4 Hz,

1 H), 4.99 (t, J = 10.2 Hz, 1 H), 4.79-4.66 (m, 2 H), 4.36 (t, J =

4.4 Hz, 2 H), 4.20-4.07 (m, 2 H), 3.53 (t, J = 5.0 Hz, 2 H), 2.87

(s, 6 H). Anal. (C₂₆H₂₄ClN₅O₆HCl 1¹/₂H₂O) C, H, N.

3

Preparation of 1-(Chloromethyl)-3-{5-[2-(dimethylamino)-

ethoxy]indol-2-carbonyl}-5-nitro-1,2-dihydro-3H-benzo[e]indole-

6-carbonitrile (21). tert-Butyl 1-Bromo-2-naphthylcarbamate

(86). A stirred solution of tert-butyl 2-naphthylcarbamate

(85)^34,35(20.3 g, 83 mmol) in MeCN (150 mL) was treated

portionwise at 0 °C with NBS (17.82 g, 100 mmol), then stirred

for a further 2 h at 0 °C. The mixture was concentrated under

reduced pressure, and the residue was dissolved in CH₂Cl₂. The

solution was filtered through a short column of silica gel, and the

product was recrystallized from MeOH to give 86³⁵(24.09 g,

1

90%) as a white solid: mp 90-91 °C; H NMR [(CD₃)₂SO] δ

8.82 (s, 1 H), 8.15 (d, J = 8.5 Hz, 1 H), 7.96 (d, J = 9.6 Hz, 1 H),

7.93 (d, J = 9.3 Hz, 1 H), 7.71 (d, J = 8.8 Hz, 1 H), 7.66 (t, J =

7.7 Hz, 1 H), 7.56 (t, J = 7.4 Hz, 1 H), 1.49 (s, 9 H). Anal.

(C₁₅H₁₆BrNO₂) C, H, N, Br.

tert-Butyl 1-Bromo-2-naphthyl-(3-chloro-2-propen-1-yl)car-

bamate (87). A stirred solution of 86 (800 mg, 2.48 mmol) in

DMF (6 mL) was treated portionwise at 0 °C with NaH (119 mg,

60% in oil, 2.98 mmol). The mixture was warmed to room

temperature for 30 min, then cooled to 0 °C and treated with 1,3-

dichloropropene (0.72 mL, 7.8 mmol, mixed isomers). The

mixture was stirred at room temperature for a further 4 h, then

diluted with 10% aqueous NaCl and extracted with EtOAc

(ꢀ2). The combined organic extracts were washed with water

(ꢀ3), dried, and concentrated under reduced pressure at 100 °C.

The residue was chromatographed on silica gel, eluting with

CH₂Cl₂/petroleum ether (7:3), to give 87 (958 mg, 97%) as an

1

oil: H NMR [(CD₃)₂SO] (mixture of rotamers and E and Z

Preparation of 8. A solution of 59 (EWG = 6-COMe) (53 mg,

0.13 mmol) in CH₂Cl₂/MeOH (1:1, 20 mL) was treated with

Cs₂CO₃(100 mg, 0.31 mmol), and the mixture was stirred at

room temperature for 15 min, then poured into water (100 mL)

and extracted with CH₂Cl₂(3 ꢀ 50 mL). The extracts were dried

and concentrated under reduced pressure. The crude product

was reacted with 5,6,7-trimethoxyindole-2-carbonyl chloride as

described in the general method above [chromatography with

EtOAc/petroleum ether (1:1), followed by crystallization from

CH₂Cl₂/petroleum ether] to give 8 (40 mg, 57%) as a yellow

solid: mp 180-183 °C; ¹H NMR (CDCl₃) δ 9.41 (s, 1 H), 9.11 (s,

1 H), 7.89 (dd, J = 8.4, 1.0 Hz, 1 H), 7.72 (dd, J = 6.9, 0.9 Hz, 1

H), 7.61 (dd, J = 8.3, 7.3 Hz, 1 H), 7.00 (d, J = 2.4 Hz, 1 H), 6.86

(s, 1 H), 4.85-4.80 (m, 1 H), 4.74-4.67 (m, 1 H), 4.33-4.25 (m, 1

H), 4.08 (s, 3 H), 3.94 (s, 3 H), 3.91 (s, 3 H), 3.93-3.87 (m, 1 H),

3.59-3.51 (m, 1 H), 2.70 (s, 3 H); ¹³C NMR δ 200.5, 160.5,

150.5, 148.6, 141.4, 140.9, 138.9, 138.3, 130.6, 129.7, 128.7,

127.2, 126.6, 126.0, 125.7, 123.5, 118.8, 116.6, 107.1, 97.7,

61.5, 61.1, 56.3, 54.7, 45.6, 43.4, 28.5. Anal. (C₂₇H₂₄-

forms) δ 8.23 (d, J = 8.4 Hz, 1 H), 8.07-7.94 (m, 2 H), 7.71 (t,

J = 7.5 Hz, 1 H), 7.65 (t, J = 7.4 Hz, 1 H), 7.51, 7.45 (2 d, J = 8.6

Hz, 1 H), 6.44-6.26 (m, 1 H), 6.21-5.99 (m, 1 H), 4.58-4.46,

4.44-4.17, 4.14-3.96 (3 m, 2 H), 1.50, 1.26 (2 s, 9 H). HRMS

(EI) calcd for C₁₈H₁₉⁷⁹Br³⁵ClNO₂(M^þ) m/z 395.0288, found

395.0261.

tert-Butyl 1-(Chloromethyl)-1,2-dihydro-3H-benzo[e]indole-3-

carboxylate (88). A mixture of 87 (23.0 g, 58 mmol), Bu₃SnH

(16.4 mL, 61 mmol), and AIBN (1.2 g, 7.3 mmol) in dry benzene

(200 mL) was stirred at reflux under N₂for 2 h, then concen-

trated under reduced pressure. The residue was chromato-

graphed on silica gel, eluting with CH₂Cl₂/petroleum ether, to

provide an oil. This was dissolved in MeOH, and following

prolonged refrigeration the precipitate was collected and recrys-

tallized from petroleum ether to give 88³⁵(13.6 g, 74%) as a

white solid: mp 107-108 °C; ¹H NMR [(CD₃)₂SO] δ 8.07 (v br, 1

H), 7.94-7.80 (m, 3 H), 7.52 (t, J = 7.4 Hz, 1 H), 7.39 (t, J = 7.5

Hz, 1 H), 4.29-4.11 (m, 2 H), 4.08 (dd, J = 11.1, 2.3 Hz, 1 H),

4.03 (dd, J = 11.1, 2.9 Hz, 1 H), 3.88 (dd, J = 11.0, 7.1 Hz, 1 H),

1.55 (s, 9 H). Anal. (C₁₈H₂₀ClNO₂) C, H, N.

ClN₃O₇¹/₂H₂O) C, H, N.

3

Preparation of 1-(Chloromethyl)-3-{5-[2-(dimethylamino)-

ethoxy]indol-2-carbonyl}-5,6-dinitro-1,2-dihydro-3H-benzo[e]-

indole (20). General Method of Amide Formation Using Indole-2-

carboxylic Acids and EDCI. A mixture of amine 105 (100 mg,

0.33 mmol), 5-[2-(dimethylamino)ethoxy]indole-2-carboxylic

acid hydrochloride (111 mg, 0.39 mmol), EDCI (249 mg, 1.30

mmol), and anhydrous TsOH (40 mg, 0.23 mmol) in dry DMA

(4 mL) was stirred at room temperature under N₂for 3 h, then

poured into dilute aqueous NH₃. The solid was collected,

1-(Chloromethyl)-3-(trifluoroacetyl)-1,2-dihydro-3H-benz[e]-

indole (89). A solution of 88 (400 mg, 1.26 mmol) in dioxane

(15 mL) was saturated with dry HCl, stirred at room tempera-

ture for 1 h, and then evaporated under reduced pressure below

30 °C. The residue was dissolved in pyridine (3 mL) and treated

dropwise at 0 °C with trifluoroacetic anhydride (0.21 mL,

1.49 mmol). The mixture was warmed to room temperature

for 5 min, then diluted with water, and the precipitated solid was

collected, dissolved in CH₂Cl₂, and filtered through a column of

Article

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22 7269

silica gel to give 89 (363 mg, 92%) as a white solid: mp (CH₂Cl₂/

petroleum ether) 157 °C; ¹H NMR [(CD₃)₂SO] δ 8.32 (d, J = 9.0

Hz, 1 H), 8.07-7.96 (m, 3 H), 7.62 (ddd, J = 8.2, 6.9, 1.2 Hz,

1 H), 7.53 (ddd, J = 8.1, 6.9, 1.1 Hz, 1 H), 4.61-4.52 (m, 1 H),

4.51-4.39 (m, 2 H), 4.15 (dd, J = 11.3, 3.0 Hz, 1 H), 4.04 (dd,

J = 11.3, 5.9 Hz, 1 H). Anal. (C₁₅H₁₁ClF₃NO) C, H, N.

Hz), 111.7, 52.7, 45.4, 42.6. Anal. (C₁₆H₁₀ClF₃N₂O ¹/₃H₂O) C,

H, N.

3

General Method of Nitration Using fHNO₃. Preparation of

1-(Chloromethyl)-5-nitro-3-(trifluoroacetyl)-1,2-dihydro-3H-ben-

zo[e]indole-6-carbonitrile (59, EWG = 6-CN). A solution of 97

(60 mg, 0.18 mmol) in CH₂Cl₂(10 mL) was treated with fuming

HNO₃(1.5 mL) and stirred for 30 min at room temperature. The

reaction was quenched with ice and extracted with CH₂Cl₂(3 ꢀ

50 mL). The extracts were dried and concentrated under reduced

pressure. The residue was chromatographed on silica gel, eluting

with EtOAc/petroleum ether (from 1:4 to 1:1), to give 59

(EWG = 6-CN) (28 mg, 41%) as a brown solid: mp (EtOAc/

petroleum ether) 201-205 °C (dec); ¹H NMR [(CD₃)₂SO] δ 8.87

(s, 1 H), 8.63 (dd, J = 8.5, 1.1 Hz, 1 H), 8.40 (dd, J = 7.3, 1.0 Hz,

1 H), 7.93 (dd, J = 8.5, 7.3 Hz, 1 H), 4.73-4.61 (m, 2 H),

1-(Chloromethyl)-3-(trifluoroacetyl)-1,2-dihydro-3H-benzo[e]-

indole-7-sulfonyl Chloride (94) and 1-(Chloromethyl)-3-(tri-

fluoroacetyl)-1,2-dihydro-3H-benzo[e]indole-6-sulfonyl Chloride

(95). Solid 89 (1.6 g, 5.1 mmol) was gradually added to chlor-

osulfonic acid (6.0 mL, 90 mmol) with ice bath cooling. The

mixture was then heated to 60 °C for 2 h, and the reaction was

quenched by pouring the mixture slowly, with stirring, into

ice-water. The precipitated solid was collected, washed with

water, dried, and chromatographed on silica gel. Elution with

EtOAc/petroleum ether (from 1:4 to 1:1) gave 94 (0.53 g, 25%)

as a pale-yellow solid: mp (EtOAc/petroleum ether) 189-192 °C;

¹H NMR (CDCl₃) δ 8.70-8.64 (m, 2 H), 8.13-8.08 (m, 2 H),

8.00 (d, J = 9.0 Hz, 1 H), 4.72-4.68 (m, 1 H), 4.55-4.48 (m,

1 H), 4.33-4.25 (m, 1 H), 3.98-3.93 (m, 1 H), 3.68-3.61 (m, 1 H);

¹³C NMR δ 154.9 (q, J_C-F= 38.4 Hz), 143.9, 140.7, 132.7,

4.55-4.49 (m, 1 H), 4.22-4.15 (m, 1H), 4.12-4.07 (m, 1 H); ¹³

C

NMR δ 153.8 (q, J_C-F= 37.6 Hz), 146.8, 139.8, 137.7, 132.6,

130.3, 129.4, 128.3, 118.7, 115.4 (q, J_C-F= 288 Hz), 115.1,

114.8, 105.4, 52.7, 47.4, 41.1. Anal. (C₁₆H₉ClF₃N₃O₃) C, H, N.

Preparation of 21. Cs₂CO₃(0.5 g, 1.5 mmol) was added to a

solution of 59 (EWG = 6-CN) (100 mg, 0.26 mmol) in CH₂Cl₂/

MeOH (1:1, 20 mL). The mixture was stirred at room tempera-

ture for 15 min, then poured into water (100 mL) and extracted

with CH₂Cl₂(3 ꢀ 50 mL). The extracts were dried, and the

solution was mixed with a solution of dry HCl in dioxane (10

mL). After 30 min the mixture was evaporated under reduced

pressure. To the residue was added 5-[2-(dimethylamino)-

ethoxy]indole-2-carboxylic acid hydrochloride (100 mg, 0.34

mmol), followed by EDCI (100 mg, 0.55 mmol), anhydrous

TsOH (20 mg, 0.12 mmol), and DMA (3 mL), and the mixture

was stirred at room temperature overnight. The mixture was

poured into dilute ice-cold aqueous NaHCO₃and extracted with

EtOAc (3 ꢀ 50 mL). The combined organic phases were washed

with water (3 ꢀ 30 mL) and then brine, dried, and evaporated to

give 21 (88 mg, 66%): mp (CH₂Cl₂/MeOH) >300 °C; ¹H NMR

(CDCl₃) δ 9.26 (br, 1 H), 9.10 (s, 1 H), 8.14 (dd, J = 8.5, 1.0 Hz,

1 H), 8.02 (dd, J = 8.2, 0.9 Hz, 1 H), 7.73 (dd, J = 8.5, 7.3 Hz,

1 H), 7.38 (d, J = 8.9 Hz, 1 H), 7.15-7.05 (m, 3 H), 4.95-4.90

(m, 1 H), 4.84-4.77 (m, 1 H), 4.37-4.29 (m, 1 H), 4.17-4.11

(m, 2 H), 3.96-3.90 (m, 1 H), 3.67-3.58 (m, 1 H), 2.81-2.76

(m, 2 H), 2.37 (s, 6 H); ¹³C NMR δ 160.7, 154.3, 148.3, 142.4,

135.9, 131.7, 130.2, 129.1, 128.7, 128.2, 127.9, 127.6, 119.2,

118.3, 116.3, 115.2, 112.8, 107.9, 106.8, 103.7, 66.8, 58.5, 54.8,

131.8, 130.1, 130.0, 125.8, 124.8, 123.2, 119.6, 115.9 (q, J_C-F=

288 Hz), 52.8, 45.4, 42.4. Anal. (C₁₅H₁₀Cl₂F₃NO₃S) C, H, N, Cl.

Later eluates gave 95 (1.54 g, 73%): mp (EtOAc/petroleum

1

ether) 181-183 °C; H NMR (CDCl₃) δ 8.87 (d, J = 9.5 Hz,

1 H), 8.73 (d, J = 9.5 Hz, 1 H), 8.36 (dd, J = 7.5, 1.0 Hz, 1 H), 8.19

(d, J = 8.4 Hz, 1 H), 7.72 (dd, J = 8.3, 7.5 Hz, 1 H), 4.72-4.66

(m, 1 H), 4.52-4.44 (m, 1 H), 4.31-4.23 (m, 1 H), 3.95-3.81 (m,

1 H), 3.63-3.56 (m, 1 H); ¹³C NMR δ 154.8 (q, J_C-F= 38.3 Hz),

141.5, 140.9, 131.0, 130.6, 128.6, 126.7, 126.5, 125.8, 125.5,

120.3, 115.9 (q, J_C-F= 288 Hz), 52.6, 45.4, 43.0. Anal.

(C₁₅H₁₀Cl₂F₃NO₃S) C, H, N, Cl.

1-(Chloromethyl)-6-iodo-3-(trifluoroacetyl)-1,2-dihydro-3H-

benzo[e]indole (96). Diglyme (10 mL) and Ti(OⁱPr)₄(200 mg,

0.7 mmol) were added to a mixture of 95 (660 mg, 1.6 mmol),

ZnI₂(653 mg, 2.4 mmol), LiCl (63 mg, 1.45 mmol), and

PdCl₂(PhCN)₂(16 mg, 0.04 mmol) under N₂, and the mixture

was stirred and heated at 155 °C for 30 min. The reaction

mixture was poured into aqueous HCl (0.05M, 50 mL) and

filtered through a wad of Celite. The filter cake was mixed with

CH₂Cl₂(50 mL, then 3 ꢀ 30 mL), and each time the mixture

was filtered. The filtrate was dried and concentrated, and the

residue was chromatographed on silica gel. Elution with

EtOAc/petroleum ether (from 1:5 to 1:2) gave 96 as a pale-

yellow solid (630 mg, 90%): mp (EtOAc/petroleum ether)

46.0, 45.5, 43.4. Anal. (C₂₇H₂₄ClN₅O₄¹/₄H₂O) C, H, N.

3

Preparation of 1-(Chloromethyl)-3-{5-[2-(dimethylamino)-

ethoxy]indol-2-carbonyl}-5-nitro-1,2-dihydro-3H-benzo[e]indole-

6-sulfonamide (22). 1-(Chloromethyl)-5-nitro-3-(trifluoroacetyl)-

1,2-dihydro-3H-benzo[e]indole-6-sulfonyl Chloride (59, EWG =

6-SO₂Cl). The 6-sulfonyl chloride 95 (750 mg, 1.9 mmol) was

dissolved in concentrated H₂SO₄(20 mL), the solution was

cooled in an ice bath, and a solution of KNO₃(195 mg, 1.95

mmol) in H₂SO₄(5 mL) was added slowly. The mixture was

stirred vigorously for 30 min, quenched with cold water, and

extracted with EtOAc (3 ꢀ 50 mL). The extracts were dried and

concentrated under reduced pressure, and the resulting solid was

separated by column chromatography on silica gel. Elution

with EtOAc/petroleum ether (from 1:4 to 1:1) gave 59 (EWG =

6-SO₂Cl) (202 mg, 59%, based on consumption of starting

material): mp (EtOAc/petroleum ether) 169 °C (dec); ¹H

NMR [(CD₃)₂SO] δ 8.63 (s, 1 H), 8.22 (dd, J = 7.6, 0.9 Hz,

1 H), 8.13 (dd, J = 8.4 Hz, 1 H), 7.71 (dd, J = 7.8, 7.8 Hz, 1 H),

1

174-177 °C; H NMR (CDCl₃) δ 8.48 (d, J = 9.3 Hz, 1 H),

8.14 (d, J = 9.3 Hz, 1 H), 8.08 (d, J = 7.3 Hz, 1 H), 7.76 (d, J =

8.4 Hz, 1 H), 7.28-7.22 (m, 1 H), 4.70-4.62 (m, 1 H),

4.48-4.39 (m, 1 H), 4.21-4.13 (m, 1 H), 3.96-3.89 (m, 1 H),

3.56-3.48 (m, 1 H); ¹³C NMR δ 154.6 (q, J_C-F= 38.3 Hz),

140.8, 137.3, 135.0, 132.8, 129.5, 128.4, 125.8, 123.4, 118.6,

116.0 (q, J_C-F= 288 Hz), 100.6, 52.8, 45.4, 42.8. Anal.

(C₁₅H₁₀ClF₃INO) C, H, N.

1-(Chloromethyl)-3-(trifluoroacetyl)-1,2-dihydro-3H-benzo[e]-

indole-6-carbonitrile (97). A mixture of 96 (148 mg, 0.34 mmol),

KCN (120 mg, 1.9 mmol), Pd(PPh₃)₄(10 mg, 0.01 mmol), and

CuI (50 mg, 0.26 mmol) in dry THF (30 mL) was heated to reflux

under N₂with vigorous stirring for 30 min. The mixture was

cooled to room temperature, diluted with EtOAc (50 mL), and

then filtered through Celite. The filtrate was washed with water

and brine, dried, and concentrated under reduced pressure.

Chromatography of the residue on silica gel, eluting with

EtOAc/petroleum ether (from 1:5 to 1:2), gave 97 (97 mg,

85%): mp (EtOAc/petroleum ether) 158-160 °C; ¹H NMR

(CDCl₃) δ 8.63 (d, J = 9.1 Hz, 1 H), 8.26 (d, J = 9.2 Hz,

1 H), 8.05 (d, J = 8.5 Hz, 1 H), 7.90 (dd, J = 7.2, 1.0 Hz, 1 H),

7.64 (dd, J = 8.2, 7.2 Hz, 1 H), 4.70-4.63 (m, 1 H), 4.51-4.43

(m, 1 H), 4.28-4.20 (m, 1 H), 3.95-3.89 (m, 1 H), 3.64-3.55 (m,

1 H); ¹³C NMR δ 154.7 (q, J_C-F= 38.3 Hz), 141.5, 132.0, 130.7,

129.0, 127.7, 127.5, 126.7, 119.7, 117.2, 115.9 (q, J_C-F= 288

4.66-4.56 (m, 2 H), 4.49-4.43 (m, 1 H), 4.17-4.02 (m, 2 H); ¹³

C

NMR δ 153.4 (q, J_C-F37.4 Hz), 149.2, 145.2, 137.7, 131.4,

130.5, 129.9, 127.7, 124.9, 118.6, 115.6 (q, J_C-F= 288 Hz) 114.3,

52.6, 47.5, 41.4. Anal. (C₁₅H₉Cl₂F₃N₂O₅S) C, H, N. 95 (457 mg,

61%) was also recovered.

Preparation of 22. A solution of 59 (EWG = 6-SO₂Cl) (300

mg, 0.66 mmol) in CH₂Cl₂/THF (1:1, 50 mL) was treated with

concentrated NH₃(0.5 mL) at room temperature for 30 min,

followed by Cs₂CO₃(0.5 g, 1.5 mmol) and stirring for another

7270 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22

Tercel et al.

15 min. The mixture was poured into water (100 mL) and

extracted with CH₂Cl₂(3 ꢀ 50 mL), and the extracts were dried.

To this solution was added a solution of dry methanolic HCl (10

mL). After 10 min the mixture was evaporated to dryness under

reduced pressure. To the residue was added 5-[2-(dimethyl-

amino)ethoxy]indole-2-carboxylic acid hydrochloride (160 mg,

0.55 mmol), EDCI (200 mg, 1.1 mmol), anhydrous TsOH (20

mg, 0.12 mmol), and DMA (5 mL). The mixture was stirred at

room temperature overnight, then poured into a dilute solution

of NaHCO₃in ice-water and extracted with EtOAc (3 ꢀ 50

mL). The combined organic phases were washed with water (3 ꢀ

30 mL) and then brine, dried, and concentrated under reduced

pressure, and the residue was crystallized from CH₂Cl₂/MeOH

to give 22 (200 mg, 53%): mp >320 °C; ¹H NMR [(CD₃)₂SO] δ

11.73 (s, 1 H), 9.03 (s, 1 H), 8.44 (d, J = 7.5 Hz, 1 H), 8.40 (d, J =

8.3 Hz, 1 H), 7.92 (dd, J = 8.0, 7.8 Hz, 1 H), 7.47 (s, 2 H), 7.41 (d,

J = 8.9 Hz, 1 H), 7.16-7.21 (m, 2 H), 6.95 (dd, J = 8.7, 2.4 Hz,

1 H), 4.98-4.89 (m, 1 H), 4.73-4.61 (m, 2 H), 4.17-4.02

(m, 4 H), 2.68-2.63 (m, 2 H), 2.24 (s, 6 H); ¹³C NMR δ 160.4,

153.0, 147.3, 141.1, 140.8, 131.9, 131.5, 130.6, 130.3, 129.7,

127.8, 127.5, 127.4, 116.8, 116.7, 116.4, 113.2, 106.1, 103.1,

66.0, 57.7, 54.7, 47.8, 45.4, 41.6. Anal. (C₂₆H₂₆ClN₅O₆S) C,

H, N, Cl.

Further elution gave the 9-nitro isomer 1-(chloromethyl)-

9-nitro-3-(trifluoroacetyl)-1,2-dihydro-3H-benzo[e]indole-6-car-

boxamide (100 mg, 30%), characterized only by NMR:

¹H NMR (CDCl₃) δ 8.71 (d, J = 9.2 Hz, 1 H), 8.49 (d, J =

9.2 Hz, 1 H), 7.83 (d, J = 7.6 Hz, 1 H), 7.67 (d, J = 7.6 Hz, 1 H),

6.30 (br, 1 H), 6.13 (br, 1 H), 4.57-4.50 (m, 1 H), 4.47-4.39 (m,

1 H), 4.02 (s, 1 H), 3.59-3.57 (m, 1 H), 3.33-3.25 (m, 1 H).

Preparation of 24. The trifluoroacetamide 59 (EWG =

6-CONH₂) was deprotected and the amide formed using EDCI

as described in the general method above to give 24 (48 mg, 72%):

1

mp (CH₂Cl₂/MeOH) >350 °C; H NMR [(CD₃)₂SO] δ 11.7 (s,

1 H), 8.95 (s, 1 H), 8.26 (dd, J = 8.4, 1.0 Hz, 1 H), 8.20 (s, 1 H),

7.84 (dd, J = 7.1, 1.0Hz, 1 H), 7.76(dd, J =8.3, 7.1Hz, 1 H), 7.52

(s, 1 H), 7.40 (d, J = 8.9 Hz, 1 H), 7.18 (d, J = 2.1 Hz, 2 H),

6.95 (dd, J = 8.9, 2.4 Hz, 1 H), 4.97-4.90 (m, 1 H), 4.72-4.58

(m, 2 H), 4.16-4.04 (m, 4 H), 2.68-2.63 (m, 2 H), 2.24 (s, 6 H);

¹³C NMR δ (one C not observed) 169.5, 160.4, 153.0, 147.9,

140.8, 133.4, 131.8, 131.0, 130.3, 129.9, 127.7, 127.4, 125.7, 118.3,

116.3, 115.2, 113.2, 106.0, 103.1, 66.2, 57.7, 54.7, 47.8, 45.5, 41.5.

Anal. (C₂₇H₂₆ClN₅O₅) C, H, N, Cl.

General Method for Hydrogenation of Nitro Compounds to

Anilines. Preparation of 5-Amino-1-(chloromethyl)-7-(methyl-

sulfonyl)-3-(5,6,7-trimethoxyindol-2-carbonyl)-1,2-dihydro-3H-

benzo[e]indole (40). A solution of 10 (63 mg, 0.11 mmol) in THF

(15 mL) with PtO₂(25 mg) was hydrogenated at 45 psi for

90 min. The mixture was filtered through Celite, the filtrate was

evaporated under reduced pressure, and the residue was tritu-

rated with EtOAc to give 40 (25 mg, 42%) as a yellow solid: mp

266-268 °C; ¹H NMR [(CD₃)₂SO] δ 11.42 (s, 1 H), 8.69 (d, J =

1.7 Hz, 1 H), 7.96 (d, J = 8.9 Hz, 1 H), 7.81 (dd, J = 8.8, 1.8 Hz,

1 H), 7.76 (s, 1 H), 7.08 (d, J = 2.0 Hz, 1 H), 6.97 (s, 1 H), 6.40 (s,

2 H), 4.71 (dd, J = 10.9, 9.0 Hz, 1 H), 4.44 (dd, J = 11.0, 1.8 Hz,

1 H), 4.18-4.11 (m, 1 H), 3.97 (dd, J = 11.0, 3.2 Hz, 1 H), 3.94

(s, 3 H), 3.82 (s, 3 H), 3.80 (s, 3 H), 3.78 (dd, J = 11.0, 7.6 Hz,

1 H), 3.25 (s, 3 H). Anal. (C₂₆H₂₆ClN₃O₆S) C, H, N.

One-Electron Reduction Potentials. Pulse radiolysis experi-

ments were performed using The University of Auckland’s

linear accelerator equipped with a radical detection system, as

previously described.⁴³One-electron reduction potentials, E(1),

were determined at pH 7 in deaerated solutions containing 2-

propanol (0.5 M) and phosphate buffer (2.5 mM) (compounds

19, 24, and 31) or imidazole buffer (3 mM) (all other

compounds). Deaeration by N₂gas exchange without bubbling

the solutions helped to prevent precipitation of the less soluble

compounds. The following reference compounds and reduction

potentials were used: triquat (7,8-dihydro-6H-dipyrido[1,2-

a:2⁰,1⁰-c][1,4]diazepinium dibromide), -548 ( 7 mV (for 19);

methylviologen (1,1⁰-dimethyl-4,4⁰-bipyridinium dichloride),

-447 ( 7 mV (for all other compounds). Regression analysis

provided the following equations:

Preparation of 6-Acetyl-1-(chloromethyl)-3-{5-[2-(dimethyl-

amino)ethoxy]indol-2-carbonyl}-5-nitro-1,2-dihydro-3H-benzo[e]-

indole (23). The trifluoroacetamide 59 (EWG = 6-COMe) was

deprotected and the amide formed using EDCI as described in

the general method above to give 23 (75%): mp (CH₂Cl₂/

1

MeOH) >300 °C; H NMR [(CD₃)₂SO] δ 11.71 (s, 1 H), 9.01

(s, 1 H), 8.35 (dd, J = 8.5, 0.9 Hz, 1 H), 8.13 (dd, J = 7.2, 0.8 Hz,

1 H), 7.80 (dd, J = 8.5, 7.2 Hz, 1H), 7.41 (d, J = 8.9 Hz, 1 H),

7.20-7.15 (m, 2 H), 6.94 (dd, J = 8.9, 2.4 Hz, 1 H), 4.98-4.89

(m, 1 H), 4.73-4.68 (m, 1 H), 4.67-4.60 (m, 1 H), 4.16-4.04 (m,

4 H), 2.69 (s, 3 H), 2.68-2.63 (m, 2 H), 2.24 (s, 6 H); ¹³C NMR δ

200.7, 200.6, 160.5, 153.0, 147.2, 141.3, 136.8, 131.9, 131.7,

130.3, 129.7, 127.6, 127.4, 127.0, 117.4, 116.3, 115.7, 113.2,

106.1, 103.1, 66.1, 57.7, 54.7, 47.7, 45.4, 41.5, 28.5. Anal.

(C₂₈H₂₇ClN₄O₅¹/₂H₂O) C, H, N, Cl.

3

Preparation of 1-(Chloromethyl)-3-{5-[2-(dimethylamino)-

ethoxy]indol-2-carbonyl}-5-nitro-1,2-dihydro-3H-benzo[e]indole-6-

carboxamide (24). 1-(Chloromethyl)-3-(trifluoroacetyl)-1,2-dihy-

dro-3H-benzo[e]indole-6-carboxamide (98). Solid 97 (500 mg,

1.48 mmol) was added to 90% H₂SO₄(5 mL) and heated to

70 °C for 1 h. After cooling to room temperature, the mixture

was poured into ice-water and extracted with EtOAc (3 ꢀ

50 mL). The extracts were dried and concentrated under reduced

pressure. Chromatography of the residue on silica gel, eluting

with EtOAc/petroleum ether (from 1:1 to 1:0), gave 98 (410 mg,

78%) as a white solid: mp (EtOAc/petroleum ether) 190-

193 °C; ¹H NMR (CDCl₃) δ 8.54-8.45 (m, 2 H), 7.89 (d, J = 5.8 Hz,

1 H), 7.69 (dd, J = 7.1, 1.1 Hz, 1 H), 7.59 (dd, J = 8.4, 7.1 Hz,

1 H), 5.91 (br, 2 H), 4.69-4.61 (m, 1 H), 4.48-4.40 (m, 1 H),

6-Substituted:

4.25-4.17 (m, 1 H), 3.97-3.89 (m, 1 H), 3.58-3.50 (m, 1 H); ¹³

C

Eð1Þ ðmVÞ ¼ -509 þ 42σ_p

ðn ¼ 5; r²¼ 0:57Þ

ðn ¼ 6; r²¼ 0:97Þ

ðn ¼ 5; r²¼ 0:88Þ

NMR δ 170.7, 153.8 (q, J_C-F= 38.0 Hz), 140.5, 134.6, 129.7,

128.6, 126.5, 125.6, 125.4, 124.8, 118.4, 115.4 (q, J_C-F= 288

Hz), 53.4, 45.4, 43.0. Anal. (C₁₆H₁₂ClF₃N₂O₂) C, H, N, Cl.

1-(Chloromethyl)-5-nitro-3-(trifluoroacetyl)-1,2-dihydro-3H-

benzo[e]indole-6-carboxamide (59, EWG = 6-CONH₂). 98 was

nitrated using fuming HNO₃as described in the general method

above [chromatography with EtOAc/petroleum ether/metha-

nol (from 5:1:0 to 9:0:1)] to give 59 (EWG = 6-CONH₂)

(150 mg, 45%) as yellow solid: mp 272-277 °C (EtOAc/

7-Substituted:

Eð1Þ ðmVÞ ¼ -508 þ 199σ_p

1

8-Substituted:

petroleum ether); H NMR [(CD₃)₂SO] δ 8.78 (s, 1 H), 8.30

(dd, J = 8.4, 1.1 Hz, 1 H), 8.24 (s, 1 H), 7.91 (dd, J=7.1, 1.0 Hz,

1 H), 7.80 (dd, J=8.4, 7.1 Hz, 1 H), 7.57 (s, 1 H), 4.69-4.61 (m,

2 H), 4.52-4.47 (m, 1 H), 4.21-4.15 (m, 1 H), 4.13-4.07 (m, 1 H);

¹³C NMR δ 169.3, 153.6 (q, J_C-F= 37.8 Hz), 148.1, 138.6,

133.4, 132.4, 130.1, 128.7, 127.9, 126.1, 119.3, 115.6 (q, J_C-F=

288 Hz), 114.4, 52.7, 47.5, 41.2. Anal. (C₁₆H₁₁ClF₃N₃O₄) C,

H, N, Cl.

Eð1Þ ðmVÞ ¼ -507 þ 122σ_p

In Vitro Cytotoxicity. For the cell lines in Tables 1-4 inhibi-

tion of proliferation of log phase monolayers was assessed in 96-

well plates as previously described.⁴⁴The drug exposure time

was 4 h under aerobic (20% O₂) or anoxic (<20 ppm O₂)

conditions followed by sulforhodamine B staining 5 days later.

Article

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22 7271

The IC₅₀was determined by interpolation as the drug concen-

tration required to inhibit cell density to 50% of that of the

controls on the same plate. Isolation of the A549 P450R^purocell

line has been previously described.²³

New Zealand, and the Australian Institute of Nuclear Science

and Engineering.

Supporting Information Available: Synthesis of all 7-, 8-, and

9-substituted nitroCBIs of Tables 1 and 2 and aminoCBIs of

Table 3 (compounds 9-18, 25-38, 41-52); combustion analy-

sis results for new compounds; 2D NMR spectra of 105, 162,

161, 97, 59 (EWG = 6-COMe), 92, 119, 144; X-ray crystal-

lographic data for 133 and 105; relationship between one-

electron reduction potential [E(1)] and cytotoxicity (IC₅₀) for

the nitroCBIs of Table 2; relationship between hypoxic cyto-

toxicity ratio (HCR) in HT29 and one-electron reduction po-

tential [E(1)] for the compounds of Table 2; correlation between

cytotoxicity in HT29 and Skov3; and MS and UV-visible

spectroscopy characterization of 28 and its major S9 metabolites

under oxic and hypoxic incubation conditions. This material is

available free of charge via the Internet at http://pubs.acs.org.

In Vitro Metabolism. Liver S9 was prepared by homogenizing

livers from female homozygous nude (CD1-Foxn1^nu/nu) mice

bred by The University of Auckland using breeding mice from

Charles River Laboratories (Wilmington, MA) in 3 volumes of

phosphate buffer (67 mM, pH 7.4, containing 1.15% KCl). The

homogenate was centrifuged at 4 °C (10000g for 20 min) and

stored at -80 °C. Tumor S9 was prepared by homogenizing

HT29 tumors grown in the same strain of mice in 5 volumes of

Tris buffer (100 mM, pH 7.4). The homogenate was centrifuged

at 4 °C (10000g for 5 min followed by 9000g for 15 min). The

protein concentrations of the liver and tumor S9 were 40.1 and

12.1 mg/mL respectively. Incubations were performed by adding

nitroCBI (50 μM in formate buffer, diluted from a 5 mM stock

solution in DMSO) to a mixture of S9, NADPH (10 mM in

phosphate buffer), and sufficient phosphate buffer (67 mM, pH

7.4) to give a total volume of 250 μL and final concentrations of 8

mg/mL protein, 1 mM NADPH, and 10 μM nitroCBI. Vials were

sealed and incubated at 37 °C; for incubations under hypoxia the

experiment was set up in an anaerobic chamber using deoxyge-

nated solvents. Reactions were quenched by the addition of an

equal volume of ice-cold acetonitrile and centrifuged at 4 °C

(12000 rpm for 10 min). A 50 μL aliquot of the supernatant was

diluted with 25 μL of ammonium formate buffer (45 mM, pH 4.5)

and analyzed by HPLC using an Altima 150 mm ꢀ 2.1 mm

column (5 μm C8 packing) and an aqueous ammonium formate

(45 mM, pH 4.5)/acetonitrile gradient. Gradients from 80% to

10% aqueous phase over 21 min and from 80% to 20% to 10%

aqueous phase over 16 min were used. Detection was by diode

array absorbance, or for MS analysis by electrospray (positive

mode and Auto MS(n); capillary voltage 4.5 kV) with an Agilent

LC/MSD Trap-SL equipped with Agilent capillary HPLC using

a drying gas flow rate of 4.4 L/min, nebulizer pressure of 12 psi,

and a drying gas temperature of 325 °C.

References

(1) Lunt, S. J.; Chaudary, N.; Hill, R. P. The tumor microenvironment

and metastatic disease. Clin. Exp. Metastasis 2009, 26, 19–34.

(2) Bertout, J. A.; Patel, S. A.; Simon, M. C. The impact of O₂

availability on human cancer. Nat. Rev. Cancer 2008, 8, 967–975.

(3) Brown, J. M.; Wilson, W. R. Exploiting tumor hypoxia in cancer

treatment. Nat. Rev. Cancer 2004, 4, 437–447.

(4) McKeown, S. R.; Cowen, R. L.; Williams, K. J. Bioreductive drugs:

from concept to clinic. J. Clin. Oncol. 2007, 19, 427–442.

(5) Chen, Y.; Hu, L. Design of anticancer prodrugs for reductive

activation. Med. Res. Rev. 2008, 29, 29–64.

(6) Stratford, I. J.; Workman, P. Bioreductive drugs into the next

millennium. Anti-Cancer Drug Des. 1998, 13, 519–528.

(7) Hicks, K. O.; Pruijn, F. B.; Secomb, T. W.; Hay, M. P.; Hsu, R.;

Brown, J. M.; Denny, W. A.; Dewhirst, M. W.; Wilson, W. R. Use

of three-dimensional tissue cultures to model extravascular trans-

port and predict in vivo activity of hypoxia-targeted anticancer

drugs. J. Natl. Cancer Inst. 2006, 98, 1118–1128.

(8) Wilson, W. R.; Hicks, K. O.; Pullen, S. M.; Ferry, D. M.; Helsby,

N. A.; Patterson, A. V. Bystander effects of bioreductive drugs:

potential for exploiting pathological tumor hypoxia with dinitro-

benzamide mustards. Radiat. Res. 2007, 167, 625–636.

Clonogenic Assay Using Single Cell Suspensions or Spheroids.

Clonogenic assay with HT29 cells comparing single-cell suspen-

sions with intact spheroids was performed as described pre-

viously.⁴⁵The mean cell density during the intact and

dissociated exposures (typically 1 ꢀ 10⁶cells/mL) and ambient

conditions (5% CO₂in O₂or N₂, 37 °C) were identical, with drug

exposure for 4 h. Colonies were grown for 14 days and stained

with methylene blue, and those with >50 cells were counted to

determine the plating efficiency (PE). Surviving fraction (SF)

was calculated as PE(treated)/PE(controls). Cell killing was

quantitated as the drug concentration required to lower the

surviving fraction to 10% (C₁₀).

Excision Assay. SiHa tumors were grown in female homo-

zygous nude (CD1-Foxn1^nu/nu) mice by subcutaneous inocula-

tion of 10⁷cells from tissue culture, and mice were randomized

to treatment when tumors reached a mean diameter of 8-10

mm. 28 was formulated in 5% DMSO, 20% PEG-400, 75%

lactate buffer, pH 4.0 (v/v/v), and administered via a tail vein.

Tumors were collected 18 h after treatment and dissociated

enzymatically to determine clonogenic survivors per gram of

tumor as previously described.⁴⁶Significance of treatment

effects was tested using ANOVA with Holm-Sidak post hoc

test on log-transformed data.

(9) Marcu, L.; Olver, I. Tirapazamine: from bench to clinical trials.

Curr. Clin. Pharmacol. 2006, 1, 71–79.

(10) Rischin, D.; Peters, L.; Fisher, R.; Macann, A.; Denham, J.;

Poulsen, M.; Jackson, M.; Kenny, L.; Penniment, M.; Corry, J.;

Lamb, D.; McClure, B. Tirapazamine, cisplatin, and radiation

versus fluorouracil, cisplatin, and radiation in patients with locally

advanced head and neck cancer: a randomized phase II trial of the

Trans-Tasman Radiation Oncology Group (TROG 98.02). J. Clin.

Oncol. 2005, 23, 79–87.

(11) Albertella, M. R.; Loadman, P. M.; Jones, P. H.; Phillips, R. M.;

Rampling, R.; Burnet, N.; Alcock, C.; Anthoney, A.; Vjaters, E.;

Dunk, C. R.; Harris, P. A.; Wong, A.; Lalani, A. S.; Twelves, C. J.

Hypoxia-selective targeting by the bioreductive prodrug AQ4N in

patients with solid tumors: results of a phase I study. Clin. Cancer

Res. 2008, 14, 1096–1104.

(12) Patterson, L. H. Bioreductively activated antitumor N-oxides: the

case of AQ4N, a unique approach to hypoxia-activated cancer

chemotherapy. Drug Metab. Rev. 2002, 34, 581–592.

(13) Patterson, A. V.; Ferry, D. M.; Edmunds, S. J.; Gu, Y.; Singleton,

R. S.; Patel, K.; Pullen, S. M.; Syddall, S. P.; Atwell, G. J.; Yang, S.;

Denny, W. A.; Wilson, W. R. Mechanism of action and preclinical

antitumor activity of the novel hypoxia-activated DNA crosslink-

ing agent PR-104. Clin. Cancer Res. 2007, 13, 3922–3932.

(14) Duan, J. X.; Jiao, H.; Kaizerman, J.; Stanton, T.; Evans, J. W.;

Lan, L.; Lorente, G.; Banica, M.; Jung, D.; Wang, J.; Ma, H.; Li,

X.; Yang, Z.; Hoffman, R. M.; Ammons, W. S.; Hart, C. P.;

Matteucci, M. Potent and highly selective hypoxia-activated achi-

ral phosphoramidate mustards as anticancer drugs. J. Med. Chem.

2008, 51, 2412–2420.

(15) Boger, D. L.; Johnson, D. S. CC-1065 and the duocarmycins:

understanding their biological function through mechanistic stu-

dies. Angew. Chem., Int. Ed. Engl. 1996, 35, 1438–1474.

(16) Tichenor, M. S.; Boger, D. L. Yatakemycin: total synthesis, DNA

alkylation, and biological properties. Nat. Prod. Rep. 2008, 25,

220–226.

Acknowledgment. The authors are grateful for the assis-

tance of Sisira Kumara and Karin Tan for HPLC studies,

Peter D. Boyd for X-ray crystallography, Susan Pullen and

Alison Hogg for in vitro cytotoxicity assays, and Sally Bai for

in vivo assessment. This study was funded by the Foundation

for Research, Science and Technology (New Economy

Research Fund), the Auckland Division of the Cancer

Society of New Zealand, the Health Research Council of

(17) Li, L. H.; DeKoning, T. F.; Kelly, R. C.; Krueger, W. C.;

McGovren, J. P.; Padbury, G. E.; Petzold, G. L.; Wallace, T. L.;

Ouding, R. J.; Prairie, M. D. Cytotoxicity and antitumor activity

7272 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 22

Tercel et al.

of carzelesin, a prodrug cyclopropylpyrroloindole analogue. Can-

cer Res. 1992, 52, 4904–4913.

(18) Kobayashi, E.; Okamoto, A.; Asada, M.; Okabe, M.; Nagamura,

S.; Asai, A.; Saito, H.; Gomi, K.; Hirata, T. Characteristics of

antitumor activity of KW-2189, a novel water-soluble derivative of

duocarmycin, against murine and human tumors. Cancer Res.

1994, 54, 2404–2410.

(19) Pavlidis, N.; Aamdal, S.; Awada, A.; Calvert, H.; Fumoleau, P.;

Sorio, R.; Punt, C.; Verweij, J.; van Oosterom, A.; Morant, R.;

Wanders, J.; Hanauske, A. R. Carzelesin phase II study in ad-

vanced breast, ovarian, colorectal, gastric, head and neck cancer,

non-Hodgkin’s lymphoma and malignant melanoma: a study of

the EORTC early clinical studies group (ECSG). Cancer Che-

mother. Pharmacol. 2000, 46, 167–171.

(20) Markovic, S. N.; Suman, V. J.; Vukov, A. M.; Fitch, T. R.;

Hillman, D. W.; Adjei, A. A.; Alberts, S. R.; Kaur, J. S.; Braich,

T. A.; Leitch, J. M.; Creagan, E. T. Phase II trial of KW2189 in

patients with advanced malignant melanoma. Am. J. Clin. Oncol.

2002, 25, 308–312.

(21) Jeffrey, S. C.; Torgov, M. Y.; Andreyka, J. B.; Boddington, L.;

Cerveny, C. G.; Denny, W. A.; Gordon, K. A.; Gustin, D.; Haugen,

J.; Kline, T.; Nguyen, M. T.; Senter, P. D. Design, synthesis, and in

vitro evaluation of dipeptide-based antibody minor groove binder

conjugates. J. Med. Chem. 2005, 48, 1344–1358.

(22) Tietze, L. F.; Major, F.; Schuberth, I.; Spiegl, D. A.; Krewer, B.;

Maksimenka, K.; Bringmann, G.; Magull, J. Selective treatment of

cancer: synthesis, biological evaluation and structural elucidation

of novel analogues of the antibiotic CC-1065 and the duocarmy-

cins. Chem.;Eur. J. 2007, 13, 4396–4409.

(31) Wilson, W. R.; Stribbling, S. M.; Pruijn, F. B.; Syddall, S. P.;

Patterson, A. V.; Liyanage, H. D. S.; Smith, E.; Botting, K. J.;

Tercel, M. Nitro-chloromethylbenzindolines: hypoxia-activated

prodrugs of potent adenine N3 DNA minor groove alkylators.

Mol. Cancer Ther. 2009, 8, 2903–2913.

(32) Patel, V. F.; Andis, S. L.; Enkema, J. K.; Johnson, D. A.; Kennedy,

J. H.; Mohamadi, F.; Schultz, R. M.; Soose, D. J.; Spees, M. M.

Total synthesis of seco (þ)- and ent-(-)-oxaduocarmycin SA:

construction of the (chloromethyl)indoline alkylating subunit by

a novel intramolecular aryl radical cyclization onto a vinyl chlor-

ide. J. Org. Chem. 1997, 62, 8868–8874.

(33) Note for substituent numbering that an n-substituted naphthoic

acid 53 becomes (n þ 1)-substituted as a benzindoline 57 or 58.

(34) Searcey, M.; Patterson, L. H. Preparation of Benzindole and

Benzoquinoline Derivatives as Prodrugs for Tumor Treatment.

Patent WO02067930A1, 2002.

(35) Jin, W.; Trzupek, J. D.; Rayl, T. J.; Broward, M. A.; Vielhauer,

G. A.; Weir, S. J.; Hwang, I.; Boger, D. L. A unique class of

duocarmycin and CC-1065 analogues subject to reductive activa-

tion. J. Am. Chem. Soc. 2007, 129, 15391–15397.

(36) For example, all isolated compounds 60 gave a singlet in the ¹H

NMR spectrum for H-4 at a highly consistent chemical shift of δ

7.75 ( 0.1 ppm in DMSO-d₆(13 examples).

(37) Hansch, C.; Leo, A.; Unger, S. H.; Kim, K. H.; Nikaitani, D.; Lien,

E. J. “Aromatic” substituent constants for structure-activity

correlations. J. Med. Chem. 1973, 16, 1207–1216.

(38) Similarly strong correlations were observed with σ_m, with the

estimated E(1) only varying by 0-14 mV compared to the values

given in Table 2.

(23) Ahn, G. O.; Botting, K. J.; Patterson, A. V.; Ware, D. C.; Tercel,

M.; Wilson, W. R. Radiolytic and cellular reduction of a novel

hypoxia-activated cobalt(III) prodrug of a chloromethylbenzindo-

line DNA minor groove alkylator. Biochem. Pharmacol. 2006, 71,

1683–1694.

(24) Tercel, M.; Denny, W. A. Synthesis of nitrogen and sulfur analo-

gues of the seco-CI alkylating agent. J. Chem. Soc., Perkin Trans. 1

1998, 509–519.

(25) Atwell, G. J.; Tercel, M.; Boyd, M.; Wilson, W. R.; Denny, W. A.

Synthesis and cytotoxicity of 5-amino-1-(chloromethyl)-3-[(5,6,7-

trimethoxyindol-2-yl)carbonyl]-1,2-dihydro-3H-benz[e]indole (amino-

seco-CBI-TMI) and related 5-alkylamino analogues: new DNA

minor groove alkylating agents. J. Org. Chem. 1998, 63, 9414–

9420.

(26) Tercel, M.; Gieseg, M. A.; Denny, W. A.; Wilson, W. R. Synthesis

and cytotoxicity of amino-seco-DSA: an amino analogue of the

DNA alkylating agent duocarmycin SA. J. Org. Chem. 1999, 64,

5946–5953.

(27) Tercel, M.; Gieseg, M. A.; Milbank, J. B.; Boyd, M.; Fan, J. Y.;

Tan, L. K.; Wilson, W. R.; Denny, W. A. Cytotoxicity and DNA

interaction of the enantiomers of 6-amino-3-(chloromethyl)-

1-[(5,6,7-trimethoxyindol-2-yl)carbonyl]indoline (amino-seco-CI-

TMI). Chem. Res. Toxicol. 1999, 12, 700–706.

(28) Gieseg, M. A.; Matejovic, J.; Denny, W. A. Comparison of the

patterns of DNA alkylation by phenol and amino seco-CBI-TMI

compounds: use of a PCR method for the facile preparation of

single end-labelled double-stranded DNA. Anti-Cancer Drug Des.

1999, 14, 77–84.

(29) Hay, M. P.; Anderson, R. F.; Ferry, D. M.; Wilson, W. R.; Denny,

W. A. Synthesis and evaluation of nitroheterocyclic carbamate

prodrugs for use with nitroreductase-mediated gene-directed en-

zyme prodrug therapy. J. Med. Chem. 2003, 46, 5533–5545.

(30) Tercel, M.; Denny, W. A.; Wilson, W. R. A novel nitro-substituted

seco-CI: application as a reductively activated ADEPT prodrug.

Bioorg. Med. Chem. Lett. 1996, 6, 2741–2744.

(39) Wardman, P. Electron transfer and oxidative stress as key factors

in the design of drugs selectively active in hypoxia. Curr. Med.

Chem. 2001, 8, 739–761.

(40) Denny, W. A.; Wilson, W. R. Considerations for the design of

nitrophenyl mustards as agents with selective toxicity for hypoxic

tumor cells. J. Med. Chem. 1986, 29, 879–887.

(41) Palmer, B. D.; Wilson, W. R.; Pullen, S. M.; Denny, W. A.

Hypoxia-selective antitumor agents. 3. Relationships between

structure and cytotoxicity against cultured tumor cells for substi-

tuted N,N-bis(2-chloroethyl)anilines. J. Med. Chem. 1990, 33, 112–

121.

(42) Palmer, B. D.; Wilson, W. R.; Cliffe, S.; Denny, W. A. Hypoxia-

selective antitumor agents. 5. Synthesis of water-soluble nitroani-

line mustards with selective cytotoxicity for hypoxic mammalian

cells. J. Med. Chem. 1992, 35, 3214–3222.

(43) Anderson, R. F.; Denny, W. A.; Li, W.; Packer, J. E.; Tercel, M.;

Wilson, W. R. Pulse radiolysis studies on the fragmentation

of arylmethyl quaternary nitrogen mustards by one-electron

reduction in aqueous solution. J. Phys. Chem. 1997, 101, 9704–

9709.

(44) Hay, M. P.; Gamage, S. A.; Kovacs, M. S.; Pruijn, F. B.; Anderson,

R. F.; Patterson, A. V.; Wilson, W. R.; Brown, J. M.; Denny, W. A.

Structure-activity relationships of 1,2,4-benzotriazine 1,4-diox-

ides as hypoxia-selective analogues of tirapazamine. J. Med. Chem.

2003, 46, 169–182.

(45) Lee, H. H.; Wilson, W. R.; Ferry, D. M.; van Zijl, P.; Pullen, S. M.;

Denny, W. A. Hypoxia-selective antitumor agents. 13. Effects of

acridine substitution on the hypoxia-selective cytotoxicity and

metabolic reduction of the bis-bioreductive agent nitracrine N-

oxide. J. Med. Chem. 1996, 39, 2508–2517.

(46) Wilson, W. R.; Pullen, S. M.; Hogg, A.; Hobbs, S. M.; Pruijn, F. B.;

Hicks, K. O. In Vitro and in Vivo Models for Evaluation of

GDEPT: Quantifying Bystander Killing in Cell Cultures and

Tumors. In Suicide Gene Therapy: Methods and Reviews; Springer,

C. J., Ed.; Humana Press: Totowa, NJ, 2003; pp 403-432.

Article Doi

Hypoxia-activated prodrugs: Substituent effects on the properties of nitro seco-1,2,9,9a-tetrahydrocyclopropa[c]benz[e]indol-4-one (nitroCBI) prodrugs of DNA minor groove alkylating agents

DOI: 10.1021/jm901202b

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm901202b

Full text of DOI:10.1021/jm901202b

Products guided by the article

R&D Labs maybe for 885228-03-5

Relevant to this article

Hot Product