Antitumor agents. 207. Design, synthesis, and biological testing of 4β-anilino-2-fluoro-4′-demethylpodophyllotoxin analogues as cytotoxic and antiviral agents

Article abstract of DOI:10.1021/jm000377f

2-Fluoropodophyllotoxin (11) and several 4β-anilino-2-fluoro-4′-O-demethyl analogues were synthesized and evaluated in both antineoplastic and antiviral assays. These compounds were moderately active against some cancer cell lines, but they were less acti

Full text of DOI:10.1021/jm000377f

1422
J . Med. Chem. 2001, 44, 1422-1428
An titu m or Agen ts. 207.1 Design , Syn th esis, a n d Biologica l Testin g of
4â-An ilin o-2-flu or o-4′-d em eth ylp od op h yllotoxin An a logu es a s Cytotoxic a n d
An tivir a l Agen ts
David S. VanVliet,† Yoko Tachibana,† Kenneth F. Bastow,† Eng-Shang Huang,‡ and Kuo-Hsiung Lee*,†
Natural Products Laboratory, School of Pharmacy, University of North Carolina, Chapel Hill, North Carolina 27599-7360,
and Lineberger Comprehensive Cancer Center, School of Medicine, University of North Carolina,
Chapel Hill, North Carolina 27599-7295
Received August 31, 2000
2-Fluoropodophyllotoxin (11) and several 4â-anilino-2-fluoro-4′-O-demethyl analogues were
synthesized and evaluated in both antineoplastic and antiviral assays. These compounds were
moderately active against some cancer cell lines, but they were less active than the
corresponding nonfluorinated analogues. Compound 11 exhibited the best activity against KB
carcinoma with a GI50 of approximately 30 nM. Most compounds exhibited moderate activity
against HCMV with ID50 and ID90 values in the range of 1 µM and 4 µM, respectively. Both 9
and 11 showed an unusual 10-fold selectivity for HSV-2 compared to HSV-1.
Ch a r t 1. Structures of Podophyllotoxin (1), Etoposide
(2), Teniposide (3), NK-611 (4), and GL331 (5)
In tr od u ction
Podophyllum pelatum L. (and P. emodi L., found in
China and India), also known as the American Man-
drake or Mayapple, is the source of the natural product
podophyllotoxin (1). Its medicinal properties have been
well-recognized for centuries. The first written account
of its medical value occurred in 1731 in The Natural
History of the Carolina’s, where this plant is described
as an emetic and a purgative. From 1820 to 1942, P.
pelatum L. was listed in the United States Pharma-
copeia as a cathartic and a cholagogue, but in 1942, it
was removed from the 11th edition due to its undesir-
able toxicity. However, it is still listed in the Pharma-
copeias of several foreign countries, including England,
France, Germany, Norway, and Russia. Podophyllotoxin
itself was first isolated in either 1880 or 1891, while its
structure was elucidated in 1932. In 1952, the originally
assigned stereochemistry was corrected to give the
currently accepted structure shown in Chart 1.
The biological activity of 1 has led to extensive
structural modification resulting in several clinically
useful compounds. Etoposide (2)2 and teniposide (3),3
developed in the late 1960s and early 1970s by Sandoz,
are two epipodophyllotoxins4 in clinical use as antine-
oplastic agents. They are used against a variety of
cancers, including germ-cell malignancies, small-cell
lung cancer, non-Hodgkin’s lymphoma, leukemia, Ka-
posi’s sarcoma, neuroblastoma, and soft tissue sarcoma.
In another podophyllum glycoside derivative, NK611 (4),
the 2′′-hydroxyl group in the glucose ring of 2 has been
replaced with an N,N-dimethylamino moiety. The main
advantage of this compound is its superior water
solubility compared to that of 2. It is currently undergo-
ing phase I clinical evaluation, and further development
in phase II is expected.5 GL331 (5), which was developed
in our laboratory in the early 1990s and licensed to
Genelabs Technologies, Inc., contains a p-nitroanilino
moiety at the 4â-position instead of a glycoside. It is
currently in phase II clinical trials, focusing on gastric
carcinoma, colon cancer, and non-small-cell lung cancer.
It has also shown positive results against malignancies
resistant to 2.6 Etopophos, an etoposide prodrug in
which a disodium phosphate salt was prepared at the
4′-phenolic oxygen, is also in phase II clinical trials.7
Several different mechanisms of action have been
connected to this compound family. Podophyllotoxin (1)
inhibits tubulin polymerization through interaction at
the colchicine binding site on the tubulin monomer.
While the binding of 1 to tubulin is slow (1000/M‚s at
37 °C) and has a high activation energy (14.7 kcal/mol),
the affinity is relatively low (1.8 µM at 37 °C).8 Dem-
ethylation of the 4′-methoxy moiety eliminates the
* To whom correspondence should be addressed. Tel: 919-962-0066.
Fax: 919-966-3893. E-mail: khlee@unc.edu.
† Natural Products Laboratory.
‡ Lineberger Comprehensive Cancer Center.
10.1021/jm000377f CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/24/2001

Antitumor Agents
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 9 1423
Ch a r t 2. Oxidative Metabolic Pathway for
Ch a r t 3. Hydrolytic Metabolism for Podophyllium
Podophyllotoxin
Compounds
metabolites are inactive as antitumor agents.15 In rat
livers isolated after tritium-labeled etoposide adminis-
tration, 90% of radiolabel was found in the bile within
3 h of administration.16 Studies in cancer patients
following intravenous administration of radiolabeled 2
detected up to 19% of the label in the bile and 58% in
the urine.17 In both studies, the compounds were
confirmed either as 6, 7, or 8 by HPLC analysis. The
C-2 hydrogen can be removed chemically at pH 9.3 to
give the enolate. Subsequent reprotonation gives the
thermodynamically more stable picro form.18 At pH 12,
the lactone ring opens to give the hydroxy acids, and
the equilibrium shifts back to the lactone when the pH
drops below 4.11
Several different approaches have been tried to
ameliorate the problem of hydrolytic inactivation. One
method involves replacing the C-2 hydrogen atom in a
diastereospecific manner with a functionality that is not
susceptible to enolization. Glinski and co-workers pre-
pared a series of C-2 substituted derivatives where the
hydrogen had been replaced with a series of small
functional groups.19 The enolate was formed using LDA
in dry THF and quenched in a diastereospecific manner
with various electrophilic groups. Using this procedure,
the 2-methyl, 2-chloro, 2-bromo, 2-(methylthio), and
2-hydroxyl compounds were prepared. Only the chloro
derivative showed significant activity in the P388 and
L1210 assays, giving a T/C of 156 against P388 leuke-
mia at 40 mg/kg dosing.
A second method attempted to inhibit metabolic
inactivation was replacement of the C-2 carbon with a
nitrogen atom.20 Several derivatives of the 2-aza com-
pounds were prepared and tested for biological activity.
They were all active against KB carcinoma; however,
they were less active than 2.21 Many other D-ring
modified derivatives of the epipodophyllotoxins have
been prepared, including D-ring cyclopentanone, cyclo-
pentane, ether, sulfide, sulfone, and sulfoxide analogues.
All of these derivatives were found to be less active than
their parent compounds in mitotic inhibition and cyto-
toxicity assays.22,23 Clinical evaluation was begun but
discontinued on two other derivatives due to a poor
activity profile and unacceptable cytotoxicity.24 The
activity found with 2-chloroetoposide initiated an at-
tempt to prepare the corresponding 2-fluoro derivative.
Unfortunately, at that time, the current electrophilic
fluorinating reagents were precarious. One reported
antimitotic activity of these compounds. Etoposide and
other related compounds are potent inhibitors of DNA
topoisomerase II (topo II). Topo II is an essential enzyme
that functions in the segregation of newly replicated
chromosome pairs, in chromosome condensation, and in
alteration of DNA supercoiling.9 GL331, which is also
an inhibitor of topo II, has recently been connected to
apoptotic cell death via activation of CDC2 kinase.10
Another enzyme, CDC25, is a dual phosphatase respon-
sible for the dephosphorylation of cyclin B-associated
CDC2 in the G2 stage of mitosis. This dephosphorylation
activates CDC2, which has been previously implicated
in apoptotic cell death.11 One study indicates that
GL331 facilitates the association between CDC25 and
Raf-1. This interaction stimulates the downstream
signaling pathways of Raf-1, which includes the CDC2
dephosphorylation, causing apoptosis in treated cells.
Due to the rich variety of biological activity found in
the podophyllum family of compounds, numerous stud-
ies have been performed to determine the routes of
metabolism in vivo. Two main pathways have been
found: an oxidative pathway, which is also important
for the mechanism of action, and a hydrolytic pathway,
which leads directly to inactive compounds.
The oxidative pathway for 2 and 3 is shown in Chart
2.12 Part of or this entire route will likely apply to other
related compounds. Etoposide-induced DNA strand
cleavage requires the presence of metal ions (specifically
Cu2+ and Fe3+).13 This cleavage can also be induced by
irradiation with UV light and inhibited with known
radical scavengers. Using ESR spin-trapping tech-
niques, Sakurai et al. were able to conclusively detect
the formation of a hydroxy (not phenoxy) radical. This
result suggests that the active species for DNA cleavage
is a hydroxy radical formed by metal/drug-induced
activation of a water molecule. SAR studies have
indicated that the 4′-phenol also is important for this
activity.14
Three major hydrolytic metabolites have been found
for 1-3: the picro form (6), the cis-hydroxy acid (7), and
the trans-hydroxy acid (8) (see Chart 3). All of these

1424 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 9
VanVliet et al.
Ta ble 1. Electrostatic and Mulliken Charge Comparisons for
the Lactone Carbonyl Carbon (C-13)
Sch em e 1. Preparation of 2-Fluoropodophyllotoxin (11)
2-F derivatives
2-H derivatives
electrostatic Mulliken electrostatic Mulliken
R group
charge
charge
charge
charge
F
0.88
0.87
0.87
0.88
0.89
1.09
0.35
0.35
0.35
0.35
0.35
1.00
0.86
0.89
0.86
0.90
0.87
1.12
0.38
0.38
0.38
0.38
0.38
0.99
CN
NO2
3,4-OCH2CH2O
NH2
DMEPa
a
This compound was optimized with the 3-21G* basis set. All
others were optimized using PM3.
attempt using perchloryl fluoride (FClO4, also known
as fluoroperchlorate) in LDA resulted in a “violent
explosion, causing serious injury”.19
We have reported already on the diastereospecific
synthesis of 2-fluoropodophyllotoxin.25 In this report, we
present the continued synthesis and examination of
related fluorinated compounds. We have prepared sev-
eral 4â-anilino-2-fluoro-4′-O-demethylpodophyllotoxins
related to 5 and examined their biological activity in
both antiviral and antineoplastic assays.
withdrawing inductive effect; however, its several lone
pairs of electrons can be donated via a mesomeric effect.
In these compounds, these two effects nearly balance
out, leading to a negligible difference for the partial
positive charge on the carbonyl carbon between the
fluorinated and nonfluorinated compounds at both
theory levels. Although unexpected and difficult to
explain, this result did not immediately counter-indicate
the preparation of the fluorinated series of compounds.
Ca lcu la tion s
Before attempting the synthesis of the 2-fluoropodo-
phyllotoxin, we considered the possibility that, due to
its electronegativity, fluorination might increase the
partial positive charge at the carbonyl carbon and
greatly increase the likelihood of hydrolysis of the D-ring
lactone. This question was addressed computationally
using both semiempirical and ab initio calculations. The
geometries of the test compounds were optimized using
both SYBYL and MacSpartan Plus software packages.
A conformational search of all rotatable bonds was done
using the systematic search option within SYBYL. The
rotatable bonds were examined in 5° increments with
a minimum energy cutoff of 999 kcal/mol, and all
conformations were saved into a separate database. The
structures then were optimized using the SYBYL force
field with Gasteiger-Huckel charges. The five lowest-
energy conformations were exported into MacSpartan
Plus for optimization at either the PM3 or 3-21G* level.
Compound 12 and its corresponding nonfluorinated
analogue were optimized using 3-21G* while, due to a
limit on the number of atoms permitted for the ab initio
basis sets in MacSpartan Plus, the other compounds
were optimized at the PM3 semiempirical level. The
lowest-energy conformation was kept for each compound
and all others were discarded. The electrostatic and
Mulliken charges of the lactone carbonyl carbons were
compared for the fluorinated and nonfluorinated deriva-
tives. These results are shown in Table 1. Because of
its high electronegativity, fluorine has a strong electron-
Ch em istr y
The synthesis of 2-fluoropodophyllotoxin (11) is shown
in Scheme 1.25 Briefly, 1 was protected as the TBDMS
ether (9) using standard conditions.26 The electrophilic
fluorination was performed using a combination of the
methods of Davis27 and Differding to give 10.28 The
deprotection was carried out with tetrabutylammonium
fluoride in THF to give the final product (11). The
overall yield for the three steps was 85%.
Demethylation of the 4′-methoxy group was extremely
difficult. Attempts to prepare 2-fluoro-4′-O-demethyl-
epipodophyllotoxin (12) using standard conditions [HBr-
(g) in 1,2-dichloroethane followed by water, acetone, and
BaCO3] gave a complex mixture. Varying the reaction
temperature from -40 to +20 °C did not change the
composition. TLC analysis showed five or six spots in
all cases. BBr3 at varying temperatures (-78 to 0 °C)
also gave a complex mixture. Although several spots
were positive for a phenol,29 a cleaner method was
desired. Literature examination revealed that trimeth-
ylsilyl iodide (TMS-I) had been used successfully for the
selective demethylation of podophyllotoxin.30 Accord-
ingly, 2-fluoropodophyllotoxin was selectively and cleanly
demethylated using the procedure of Daley et al.

Antitumor Agents
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 9 1425
Ta ble 3. HCMV Screening Data for
4â-Anilino-2-fluoropodophyllotoxins
Sch em e 2. Preparation of
4â-Anilino-2-fluoro-4′-demethylpodophyllotoxin
Analogues
2-F compd 2-H compd HEL cells
ID50 ID90
(µM) (µM)
ID50
(µM)
ED50
(µM)
R groupa
13: 4-F
14: 4-CN
15: 4-NO2
SIb
6.2
11.1
4.2
5.2
3.0
<1.0
<1.6
8.4
0.97 4.47
0.76 4.29
1.45 4.60
0.65
0.62
0.35
6.10
8.43
6.09
6.70
7.50
5.0
16: 3,4-OCH2CH2CO 1.30 2.31
17: 4-NH2
11
12
2.54 5.02
1.25
>5
>5
<0.10c
>5
>5
8.20
8.64
2: etoposide
not made
1.03
a
b
Numbers are for the fluorinated compounds. SI ) selectivity
index, ED50/ID50 of 2-F compounds or of 2-H compound for 2
(etoposide). c ID90 ) 0.86 µM.
Ta ble 4. Data for Antineoplastic Screening of
4â-Anilino-2-fluoro Derivatives
percent inhib (%)
concn
(µg/mL)
KB
cells
CAKI-1
cells
A549
cells
MCF-7
cells
compd
13
4
70.9
10.1
7.5
30.5
4.2
5
41.6
3.8
0
44.3
15.3
10.3
28.4
9.5
7.9
21.9
2.6
61.4
26.2
16.1
52.5
11.9
11.1
51.8
6.5
61.2
20
13.2
45.8
8.2
7.7
42.1
7.3
0.4
0.04
4
0.4
0.04
4
0.4
0.04
4
0.4
0.04
4
0.4
0.04
4
0.4
0.04
4
14
15
16
17
11
12
0.8
26.3
1.3
0
0.9
29.4
1.4
30.5
0
60.1
19.6
0
51
7.7
Ta ble 2. Herpesvirus Activity for 2-Fluoropodophyllotoxin and
Its Intermediatesa
cell growth inhib
HSV-1 inhib
HSV-2 inhib
61.5
7.5
0
87.4
72
53.3
64
25.6
0
38.0
2.5
0
40.5
39.7
18.1
39
48.1
5.8
0
51
48
34.2
53.8
21.8
0
ED25
(nM)
MIC
(nM)
ED100
(nM)
MIC
(nM)
ED100
(nM)
MIC
(nM)
0
compd
73.1
69.6
59.4
78.1
34.3
0
1
9
10
11
20
6000
25000
1000
8
3000
6000
800
50
10000
NA
10
1000
NA
100
1000
NA
10
100
NA
100
10000
1000
1000
0.4
0.04
10.6
0
a
NA ) not active; MIC ) minimum inhibitory concentration.
Ta ble 5. Comparison of Fluorinated and Nonfluorinated
Compounds as Antineoplastic Agents
(Scheme 2). The only observed product was 12. The
conversion was only 25% (based on recovered starting
material); however, the yield was 80%. This procedure
was selected because the starting material was readily
recovered and recycled. Optimal purity and yields were
obtained at 0 °C using 3.1 equiv of TMS-I. Selected side
chains were added at position-4 based upon previous
studies in our laboratory.31 HBr(g) in 9:1 1,2-dichloro-
ethane:diethyl ether was added to 12 at 0 °C to generate
the secondary bromide. A substituted aniline and BaCO3
were then added, and the reaction mixture stirred for
12 h. Flash chromatography gave the desired 4â-anilino
compounds 13-16 in moderate to good yields. Com-
pound 17 was prepared by catalytic hydrogenation of
15. NMR analysis of the final products confirmed both
the â-conformation of the C-4 anilino moiety and the
trans C/D-ring fusion.
2-F compd: log GI50
2-H compd: log GI50
compd
KB
CAKI A549 compd
KB
CAKI A549
11
13
15
16
17
2-F
-7.5 >-5
-5.7 >-5
>-5 >-5
>-5 >-5
-5.4 >-5
-5.4 11-H
-5.7 13-H
-5.4 15-H
-5.4 16-H
-5.4 17-H
2
-7.0
-6.5
-6.8
-6.9
-6.5
-5.2
-8.1
-6.6
-6.8
-7.0
-6.5
-5.2
-6.6
-6.3
-6.2
-6.1
-6.7
not made
HSV-2. 2-Fluoropodophyllotoxin (11) and its precursors
(9, 10) were less cytotoxic than the podophyllotoxin
control. Compounds 9 and 11 were less active against
both HSV-1 and HSV-2 and 10 was inactive. On
comparing 1 and 11, fluorination resulted in a 50-fold
drop in cytotoxicity, but also a 200-fold drop in the ED100
against HSV-1 and a 10-fold drop against HSV-2. The
minimum inhibitory concentration (MIC) also dropped
similarly. Both 9 and 11 showed an unusual 10-fold
selectivity for HSV-2 compared to HSV-1. The 4â-
anilino-2-fluoro-4′-O-demethylpodophyllotoxins 12-17
were not examined in this assay, because 2-fluoropodo-
phyllotoxin displayed greatly decreased activity and 4â-
anilino compounds normally lack HSV activity.33 In
addition, etoposide is a nonselective inhibitor of HSV,34
Resu lts a n d Discu ssion
The in vitro antiviral and antineoplastic data are
presented in Tables 2-5. The herpesvirus simplex 1 and
2 (HSV-1 and HSV-2) screening was done according to
the procedure of Bastow et al.32 The cell growth inhibi-
tion (column 1) measures the death of the healthy cells;
the other columns indicate the inhibition of HSV-1 and

1426 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 9
VanVliet et al.
and etoposide and 4â-anilino compounds both function
as DNA topo II inhibitors.
pared in Table 5. As was seen in the antiviral screens,
these compounds were moderately active against some
of the cancer cell lines, but they were less active than
the corresponding nonfluorinated analogues. Compound
11 exhibited the best activity against KB carcinoma
with a GI50 of approximately 30 nM, while 13 and 17
had GI50 values of 2 and 4 µM, respectively. Although
not as potent as the nonfluorinated compounds, the
decreased cytotoxicity may portend a greatly increased
therapeutic index. Further study is underway to deter-
mine exact therapeutic indices.
In conclusion, fluorination of the 2-position appears
not to be a positive modification for the podophyllotox-
ins. The dramatic loss of activity in many of the screens
is unusual and difficult to explain. The steric param-
eters of hydrogen and fluorine are extremely similar
(van der Waals radii of 1.2 Å for hydrogen vs 1.35 Å for
fluorine). Although the calculated charges do not indi-
cate an increased likelihood for hydrolysis of the lactone,
the explanation may nevertheless be electrostatic in
nature. Previous pharmacophoric models of these com-
pounds have suggested that they may interact with
DNA in a nonintercalative manner.40 Thus, because the
phosphate backbone of DNA is strongly negatively
charged and the fluorine atom has an increased partial
negative charge (due to a through-space effect), the
DNA-compound interaction may be inhibited with the
fluorinated compounds. Accordingly, if such interaction
with DNA is essential to the primary mechanism of
action for these compounds, any decrease could result
in a dramatic loss of activity. Further studies, including
DNA binding and topo II inhibition assays, are under-
way to determine the extent and importance of the
interaction with DNA and its relation to antiviral
mechanism of action.
Human cytomegalovirus (HCMV) is another type of
herpesvirus that has come into recent prominence with
the emergence of the AIDS pandemic. It is one of the
numerous devastating opportunistic infections con-
tracted by AIDS patients. HCMV retinitis, infection in
the eyes of a patient, frequently results in blindness.35
In non-AIDS patients, the worst effects of HCMV are
in infants. Between 1:200 and 1:100 births in the United
States involves congenital or maternal transmission,
respectively, of HCMV to the newborn; 63% of congeni-
tally infected infants have some form of psychomotor
damage, and 59% suffer hearing loss. Also, a significant
portion (47%) will meet the clinical definition of men-
tally retarded, and 36% of these children will also meet
the definition of severely mentally retarded.36 HCMV
infection activates cellular DNA topo II synthesis, and
this activation is essential for viral DNA replication.
Accordingly, topo II inhibitors can exhibit good activity
against HCMV. Thus, we decided to test the fluorinated
series of compounds against HCMV.37,38
The results of the HCMV screening are shown in
Table 3. Most compounds exhibited moderate activity
against HCMV in the virus yield (as measured by
plaque assay) and in DNA dot-dot hybridization assays.
For 13-16, the ID50 and ID90 values were in the range
of 1 µM and 4 µM, respectively. Compounds 15 and 16
also exhibited a narrow therapeutic range. Where the
data was available, comparison was made to the non-
fluorinated analogues.39 For most compounds, activity
decreased slightly (1.4- to 4-fold) upon fluorination. One
unusual result from this study was the complete inac-
tivity of 11 and 12. While 2-fluoropodophyllotoxin (11)
showed no activity, podophyllotoxin (1) had an ID50
below 100 nM. This dramatic loss of activity is unex-
plained. The ED50 values for HEL cell growth are also
shown in Table 3 along with a selectivity index. For all
tested compounds, the ED50 was calculated against the
growing HEL cell. The fluorinated compound 11 was
extremely toxic. Compound 14 exhibited a selectivity
index of 11, while most of the remaining derivatives
showed values in the range of 3-8. Compounds 11 and
12 did not demonstrate any selectivity.
Exp er im en ta l Section
All reagents were purchased from Aldrich Chemical Co.
unless otherwise specified. Water used in all reactions was
deionized. “Dry” solvents were distilled from an appropriate
drying agent and stored over 4 Å molecular sieves until used.
Oven-dried glassware was stored in a 145 °C drying oven for
a minimum of 12 h prior to use. Molecular sieves were dried
at 350 °C for 6 h then stored at 145 °C until used. All NMR
assignments (δ, ppm) were made using a combination of 1D
and 2D methods, including COSY, NOSY, HETCOR, and LR-
HETCOR.
P r ep a r a tion of 4′-O-Dem eth yl-2-flu or op od op h yllotox-
in (12). 2-Fluoropodophyllotoxin (11) (8.1 mmol) was dissolved
in 100 mL of dry CH2Cl2 and cooled to 0 °C under an Ar(g)
atmosphere. Trimethylsilyl iodide (3.1 equiv), as a solution in
10 mL of dry CH2Cl2, was added slowly over 30 min. The
reaction was stirred at 0 °C for 6 h and poured into 100 mL of
1:1 water:acetone (precooled to 0 °C). BaCO3 (10 mmol) was
added, and the reaction mixture was heated to 40 °C for 60
min. After cooling, it was poured into 500 mL of 10% (w/v)
Na2SsO3(aq) and stirred for 60 min. The layers were separated,
and the water later was extracted with 2 × 100 mL of CH2-
Cl2. The combined organic layers were dried over MgSO4, and
the solvent was removed under reduced pressure. The crude
solid was dissolved in a minimum of hot acetone, and the
product was precipitated with hexanes. The supernatant was
then reduced by rotary evaporation and the starting material
was recovered: yield ) 80% (25% conversion based on recov-
ered starting material); mp ) 237 °C dec; TLC (6:4 hexanes:
acetone) Rf ) 0.50; 1H NMR (400 MHz, acetone-d6) 4.72 (d,
1H, H1), 3.13 (dq, 1H, H3), 5.07 (d, 1H, H4), 2.86 (br, 1H,
4-OH), 7.06 (s, 1H, H5), 6.59 (s, 1H, H8), 4.66 (dd, 1H, H11R),
4.55 (dd, 1H, H11â), 6.00 (d, 2H, H14), 6.38 (s, 2H, H2′), 3.67
The antiviral mechanisms of action can only be
speculative at this time. The HSV inhibition by podo-
phyllotoxin and the related 9 and 11 could be related
to antimicrotubule activity and consequent inhibition
of viral transport, and the anti-HCMV activity of the
â-anilino-4′-demethyl compounds could be related to
their topo II inhibition. However, although the topo II
inhibitor etoposide causes irreversible inhibition of
HCMV DNA synthesis, topo II involvement has not been
definitely associated.38
Compounds 2-5 all are either in clinical use or in
clinical trials against various cancers; accordingly, the
fluorinated compounds were examined versus several
different cancer cell lines, including KB cells (nasopha-
ryngeal carcinoma), CAKI-1 cells (renal carcinoma),
A549 cells (lung cancer), and MCF-7 cells (breast
cancer). Table 4 presents the screening data as the
percent inhibition of growth compared to the control
(untreated) cells. The GI50 values for the fluorinated
compounds and the nonfluorinated analogues are com-

Antitumor Agents
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 9 1427
(s, 6H, 3′-OCH3); 13C NMR (100 MHz, acetone-d6) 49.88 (C1),
97.65 (C2), 41.51 (C3), 68.51 (C4), 102.31 (C5), 148.16 (C6),
149.14 (C7), 109.85 (C8), 128.25 (C9), 132.87 (C10), 68.51
(C11), 170.42 (C13), 102.32 (C14), 132.87 (C1′), 110.22 (C2′),
136.42 (C3′), 129.40 (C4′), 56.63 (C3′-OCH3); 19F NMR (376.5
MHz, acetone-d6) -162.50 (dd, 3J 1H/2F ) 11.1 Hz, 3J 2F/3H ) 40.8
Hz). Anal. (C21H19FO8) C, H.
Gen er a l P r ep a r a tion of 4â-An ilin o-4′-O-d em eth yl-2-
flu or op od op h yllotoxin s 13-16. In dry glassware, 12 (0.10
mmol) was dissolved in 10 mL of dry 1,2-dichloroethane and
1 mL of dry diethyl ether and cooled to 0 °C. HBr(g) was
bubbled into the reaction for 20 min. Excess gas was scrubbed
with a NaHCO3(aq,satd) bubbler. After stirring for 2 h at 0
°C, the solvent was removed under reduced pressure to give a
slightly colored oil. The oil was dissolved in 10 mL of dry THF.
The substituted aniline (0.30 mmol) and BaCO3 (0.40 mmol)
were then added and stirred at ambient temperature for
several hours. The reaction was then filtered through a pad
of Celite, and the solvent was removed under reduced pressure.
The crude solid was chromatographed on the FlashElute
chromatography system using a 25M silica cartridge.
Her p esvir u s. 1. Cell Gr ow th In h ibition Assa y.32 Vero
cells (African green monkey kidney cells) were cultured in
RPMI-1640 with 100 µg/mL of kanamycin and 5% (v/v) of calf
serum in a humidified atmosphere containing 5% CO2. Cul-
tures were seeded at 50 000 cells/mL in 96-microtiter-well
format with various concentrations of the test compounds.
Drug exposure was for 2 days and cell numbers were deter-
mined by the sulforhodamine B (SRB) staining method. The
ED25 value, the drug concentration that reduced absorbency
by 25%, was interpolated from dose-response data. The MIC
concentration is the minimum value from the dilution series
that gave detectable cell growth inhibition. Each test was
performed in triplicate, and absorbency readings varied by no
more that 5%. ED25 values determined in independent tests
varied by no more than 30%.
2. P la qu e Elim in a tion Assa y.32 Antiviral activity was
evaluated in pre-formed Vero monolayer cultures using KOS
(HSV-1) and 186 (HSV-2) viral strains. Confluent Vero cultures
were inoculated with approximately 100 PFU (plaque forming
units) of virus, and after 30 min, the inoculum was replaced
with 1 mL of medium supplemented with 0.5% (v/v) calf serum,
1% (v/v) carboxymethylcellulose and drugs at various concen-
trations. The cells were stained and fixed after 2 days with
0.8% (w/v) crystal violet in 50% ethanol and macroscopic
plaques were scored. Determinations of drug susceptibility
were repeated at least twice on separate occasions and
reproducible results were obtained. The ED100 concentration
is the dilution tested that completely eliminated macroscopic
plaque formation without over toxicity to cell monolayers based
on staining intensity and microscopic examination prior to
staining. The MIC concentration value is the minimum value
from the dilution series that gave antiviral effect (i.e. reduction
in plaque size and/or number relative to infected controls).
4â-(p-F lu or oa n ilin o)-4′-O-d em eth yl-2-flu or op od op h yl-
lotoxin (13): yield ) 45%; mp ) 188-9 °C; 1H NMR (400 MHz,
acetone-d6) 4.75 (d, 1H, H1), 3.39 (dq, 1H, H3), 5.24 (d, 1H,
H4), 6.87 (s, 1H, H5), 6.60 (s, 1H, H8), 4.54 (dd, 1H, H11R),
4.34 (dd, 1H, H11â), 5.98 (d, 2H, H14), 6.43 (s, 2H, H2′), 3.70
(s, 6H, 3′-OCH3), 6.94 (m, 2H, H2′′), 7.49 (m, 2H, H3′′); 19F
NMR (376.5 MHz, acetone-d6) -128.87 (t, 1F, F4′′), -162.95
3
3
(dd, 1F, F2, J 1H/2F ) 11.2 Hz, J 2F/3H ) 41.4 Hz); chromatog-
raphy solvent: 1:1 EtOAc:hexanes. Anal. (C27H23F2NO7) C, H,
N.
4â-(p-Cya n oa n ilin o)-4′-O-d em eth yl-2-flu or op od op h yl-
1
lotoxin (14): yield ) 68%; mp ) 199-201 °C; H NMR (400
Cytom ega lovir u s. 1. Titer Red u ction Assa y.37 Extra-
cellular virus yield was measured by plaque formation assay
in human embryonic lung (HEL) fibroblasts. In brief, HEL cells
were grown to confluency in 24-well culture plates. Cells were
rinsed once with MEM, then infected with Towne strain CMV
at a multiplicity of 1-2 PFU/cell in the presence or absence
of various concentrations of drug. Media with appropriate
concentration of drug were changed on the day 3 post-infection.
On day 6 post-infection, the culture supernatant was removed
and used to perform a standard CMV plaque assay in HEL
cells. 10-fold serial dilution and 1% methylcellulose (in 4% fetal
calf-MEM medium) overlayer were employed. The plaque
number was counted 2 weeks after infection under an inverted
microscope.
2. DNA-DNA Dot Hybr id iza tion Assa y.38 Viral DNA
synthesis was monitored by DNA-DNA dot hybridization
using nick-translated 32P-labeled CMV DNA as the probe.
Confluent HEL fibroblasts (human embryonic lung, HEL 229,
ATCC 137-CLL), on 24-well culture plates, were infected with
Towne strain CMV at the multiplicity of 1-2 FPU/cell. After
2 h of absorption, MEM medium (supplemented with 4% fetal
calf serum, 100 U/mL penicillin and 100 µg/mL streptomycin)
and drugs at various concentrations were added to the CMV-
infected cultures. At various times after infection, CMV-
infected and mock-infected HEL cells, with or without drug
treatment, were lysed with 0.2 mL of lysis buffer solution [0.01
M Tris-HCl (pH 8.0), 0.01 M EDTA, 1% sodium dodecyl sulfate
(SDS), and 0.1 M CaCl2], and digested with proteinase K at
100 µg/mL for 1 h at 37 °C. The lysate was extracted with
phenol-chloroform (1:1) once, chloroform once, and then
precipitated with 2.5 vol of alcohol at -20 °C overnight. The
nucleic acid precipitate was dissolved in 0.2 mL of TE buffer
(0.01 M Tris-HCl, pH 7.4, and 0.001 M EDTA). For dot
hybridization, 0.05 mL of the infected or mock-infected DNA
solution from each experimental sample was first denatured
in 0.5 M NaOH with 1.5 M NaCl for 1 h at room temperature
and then neutralized on ice with 1.1 N HCl in 0.2 M Tris to a
final pH around 7.4. The mixture was then adjusted to 6 ×
SSC (SSC ) 0.15 M NaCl and 0.015 M sodium citrate), and
single-stranded DNA was immobilized on nitrocellulose mem-
brane using Minifold apparatus (Schleicher and Schuell,
MHz, acetone-d6) 4.77 (d, 1H, H1), 3.45 (dq, 1H, H3), 5.42 (d,
1H, H4), 6.90 (s, 1H, H5), 6.62 (s, 1H, H8), 4.59 (dd, 1H, H11R),
4.25 (dd, 1H, H11â), 6.00 (d, 2H, H14), 6.43 (s, 2H, H2′), 3.70
(s, 6H, 3′-OCH3), 6.99 (m, 2H, H2′′), 7.50 (m, 2H, H3′′); 19F
3
NMR (376.5 MHz, acetone-d6) -162.87 (dd, J 1H/2F ) 14.3 Hz,
3J 2F/3H ) 39.9 Hz); chromatography solvent: 3:2 EtOAc:
hexanes. Anal. (C28H23FN2O7) C, H, N.
4â-(p-Nitr oa n ilin o)-4′-O-d em eth yl-2-flu or op od op h yllo-
toxin (15): yield ) 70%; mp ) 192-3 °C; 1H NMR (400 MHz,
acetone-d6) 4.77 (d, 1H, H1), 3.50 (dq, 1H, H3), 5.54 (d, 1H,
H4), 6.92 (s, 1H, H5), 6.64 (s, 1H, H8), 4.66 (dd, 1H, H11R),
4.25 (dd, 1H, H11â), 6.01 (d, 2H, H14), 6.44 (s, 2H, H2′), 3.69
(s, 6H, 3′-OCH3), 7.90 (m, 2H, H2′′), 8.44 (m, 2H, H3′′); 19F
3
NMR (376.5 MHz, acetone-d6) -162.68 (dd, J 1H/2F ) 9.4 Hz,
3J 2F/3H ) 39.5 Hz); chromatography solvent: 1:1 EtOAc:
hexanes. Anal. (C27H23FN2O9) C, H, N.
4â-(3′′,4′′-(Eth ylen ed ioxy)a n ilin o)-4′-O-d em eth yl-2-flu -
or op od op h yllotoxin (16): yield ) 85%; mp ) 225-7 °C; 1H
NMR (400 MHz, acetone-d6) 4.77 (d, 1H, H1), 3.44 (dq, 1H,
H3), 5.70 (d, 1H, H4), 6.92 (s, 1H, H5), 6.62 (s, 1H, H8), 4.59
(dd, 1H, H11R), 4.27 (dd, 1H, H11â), 6.00 (d, 2H, H14), 6.44
(s, 2H, H2′), 3.70 (s, 6H, 3′-OCH3), 6.90 (d, 1H, H2′′), 7.82 (d,
1H, H2′′), 7.82 (d, 1H, H3′′), 4.24 (m, 2H, H-OCCO); 19F NMR
(376.5 MHz, acetone-d6) -128.87 (t, 1F, F4′′), -162.80 (dd, 1F,
3
3
F2, J 1H/2F ) 13.2 Hz, J 2F/3H ) 43.3 Hz); chromatography
solvent: 1:1 EtOAc:hexanes. Anal. (C29H26FNO9) C, H, N.
P r ep a r a tion of 4â-(p-Am in oa n ilin o)-4′-O-d em eth yl-2-
flu or op od op h yllotoxin (17). In a Parr bottle, 15 (0.10 mmol)
was dissolved in 25 mL of EtOAc and 50 mg of 10% Pd/C was
added. The reaction was placed on a hydrogenator with 50 psi
of H2(g) and shaken for 12 h. The reaction was removed and
filtered through Celite to give a clear (nonyellow) solution. TLC
gave only one spot and no starting material. The solvent was
removed under reduced pressure to give an off white solid:
1
yield ) 95%; mp ) 180 °C dec; H NMR (400 MHz, acetone-
d6) 4.81 (d, 1H, H1), 3.52 (dq, 1H, H3), 5.50 (d, 1H, H4), 6.96
(s, 1H, H5), 6.71 (s, 1H, H8), 4.64 (dd, 1H, H11R), 4.27 (dd,
1H, H11â), 5.98 (d, 2H, H14), 6.42 (s, 2H, H2′), 3.68 (s, 6H,
3′-OCH3), 7.70 (m, 2H, H2′′), 8.45 (m, 2H, H3′′); 19F NMR (376.5
MHz, acetone-d6) -162.97 (dd, 3J 1H/2F ) 11.6 Hz, 3J 2F/3H ) 40.2
Hz). Anal. (C27H25FN2O7) C, H, N.

1428 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 9
VanVliet et al.
Keene, NH 03431). The 32P-labeled nick-translated Towne
strain of HCMV DNA (specific activity: 5 × 107 to 1 × 108
cpm/µg) at a level of 1 × 106/mL of hybridization solution was
used. Hybridization was carried out at 66 °C for 18 h. Kodak
RP/R2 X-ray film was used for radioautography to trace the
radiolabeled 32P DNA hybridized. After autoradiography,
individual dots from samples on the hybridized membrane
were cut for measuring the radioactivity by liquid scintillation
counter. All assays were performed in duplicate with repro-
ducible results generated for each compound.
(13) Sakurai, H.; Miki, T.; Imakura, Y.; Shibuya, M.; Lee, K. H. Metal
and Photoinduced Cleavage of DNA by Podophyllotoxin, Etopo-
side, and Their Related Compounds. Mol. Pharmacol. 1991, 40,
965-973.
(14) Saulnier, M. G.; Vyas, D. M.; Langley, D. R.; Doyle, T. W.; Rose,
W. C.; Crosswell, A. R.; Long, B. H. E-Ring Desoxy Analogues
of Etoposide. J . Med. Chem. 1989, 32, 1418-1420.
(15) Cragg, G.; Suffness, M. Metabolism of Plant-Derived Anticancer
Agents. Pharmacol. Ther. 1988, 37, 425-461 and references
therein.
(16) Hande, K. R.; Combs, G. L.; Cotham, R.; Guengerich, F. P.;
Branch, R. A. Biliary Clearance and Hepatic Metabolism of
Etoposide (VP-16). Proc. Am. Soc. Clin. Oncol. 1985, 4, 34-50.
(17) Ho, D. H.; Kanellopoulow, K. A.; Yap, Y. H.; Casimir, M.; Savaraj,
N.; Issell, S. B.; Benjamin, R. S.; Bodey, E. P. Clinical Pharma-
cology of Etoposide by Radioimmunoassay. Proc. Am. Assoc.
Cancer Res. 1983, 24, 519-526.
(18) Gensler, W. J .; Gatsonis, C. D. The Podophyllotoxin - Picropodo-
phyllin Equilibrium. J . Org. Chem. 1966, 31, 3224-3227.
(19) Glinski, M. B.; Freed, J . C.; Durst, T. Preparation of 2-Substi-
tuted Podophyllotoxin Derivatives. J . Org. Chem. 1987, 52,
2749-2753.
3. HEL Cytotoxicity Assa y.38 HEL cells were cultured in
MEM with 10% (v/v) FBS, 100 units and 100 µg/mL each of
penicillin and streptomycin, respectively, in a humidified
atmosphere containing 6% CO2. Cultures were seeded at
50 000 cells/mL in 24-well plates (1 mL/well) with various
concentrations of the test compounds. Drug exposure was
followed for 15 days, and cell viability and cell numbers were
determined by the trypan blue staining method. The ED50
value, the drug concentration that reduced cell growth as
compared to nontreated cells by 50%, was interpolated from
dose-response data. Each test was performed in triplicate, and
readings varied by no more that 5%. ED50 values determined
in independent tests varied by no more than 30%.
(20) Pearce, H. L.; Bach, N. J .; Cramer, T. L. Synthesis of 2-Aza-
podophyllotoxin. Tetrahedron Lett. 1989, 30, 907-910.
(21) Tomioka, K.; Kubota, Y.; Koga, K. Synthesis and Antitumor
Activity of Podophyllotoxin Aza-Analogues. Tetrahedron Lett.
1989, 30, 2953-2954.
(22) Gensler, W. J .; Murthy, C. D.; Trammell, M. H. Nonenolizable
Podophyllotoxin Derivatives. J . Med. Chem. 1977, 20, 635-644.
(23) J ardine, I.; Strife, R. J .; Kozlowski, J . Synthesis, 470-MHz 1H
NMR Spectra, and Activity of Delactonized Derivatives of the
Anticancer Drug Etoposide. J . Med. Chem. 1982, 25, 1077-1081.
(24) Vaitkevicius, V. K.; Reed, M. L. Clinical Studies with Podophyl-
lum Compounds SPI-77 (NSC-72274) and SPG-827 (NSC-42076).
Cancer Chemother. Rep. 1966, 50, 565-571.
(25) VanVliet, D. S.; Lee, K. H. A High Yield Preparation of
2-Fluoropodophyllotoxin. Tetrahedron Lett. 1999, 40, 2259-
2262..
(26) Corey, E. J .; Venkateswarlu, A. A New Silyl Protecting Group.
J . Am. Chem. Soc. 1972, 94, 6190-6191.
(27) Davis, F.; Han, W. N-Fluoro-O-Benzenedisulfonimide: A Useful
New Fluorinating Reagent. Tetrahedron Lett. 1991, 32, 1631-
1634.
(28) Differding, E.; Ofner, H. N-Fluorobenzenesulfonimide: A Practi-
cal Reagent for Electrophilic Fluorinations. Synlett 1991, 187-
189.
Ack n ow led gm en t. This research was supported by
a grant from the Elsa U. Pardee Foundation (UNC No.
49849) awarded to K.-H. Lee.
Refer en ces
(1) For part 206, see: Lee, K. H. Antitumor Agents. 206. Anti-AIDS
Agents. 39. Molecular Modifications of Lead Compounds to
Develop Potent Antitumor and Anti-HIV Agents. Abstracts of
UNESCO Regional Symposium on Chemistry of Medicinal
Plants, Kumming, China, 2000; pp 14-26.
(2) Sta¨helin, H. Activity of a New Glycosidic Lignan Derivative (VP-
16-213) Related to Podophyllotoxin in Experimental Tumors.
Eur. J . Cancer 1973, 9, 215-221.
(3) Sta¨helin, H. 4′-Demethylepipodophyllotoxin Thenylidene Glu-
coside (VM 26), a Podophyllum Compound with a New Mecha-
nism of Action. Eur. J . Cancer 1970, 6, 303-311.
(4) The ‘epi’ prefix attached to the podophyllotoxin name indicates
that the stereochemistry in the C-4 position is â (up relative to
the drawn structure and trans relative to the E-ring). Without
this or other notation, the structure is R (down at the C-4
position relative to the drawn structure and cis to the E-ring).
(5) Rassmann, I.; Schrodel, H.; Schilling, T.; Zucchetti, M.; Kaeser-
Frohlich, A.; Rastetter, J .; Hanauske, A. R. Clinical and Phar-
(29) This test was done using TLC developed with FeCl3 (2.7% w/v
in 2 N HCl) purchased from Sigma.
(30) Daley, L.; Meresse, P.; Bertounesque, E.; Monneret, C. A One-
Pot, Efficient Synthesis of the Potent Cytotoxic Podophyllotoxin
Derivative NPF. Tetrahedron Lett. 1997, 38, 2673-2676.
(31) Zhang, Y.; Lee, K. H. Recent Progress in the Development of
Novel Antitumor Etoposide Analogues. Chinese Pharm. J . 1994,
46, 319-369 and references therein.
macokinetic Phase
I Trial of Oral Dimethylaminoetoposide
(NK611) Administered for 21 Days Every 35 Days. Invest. New
Drugs 1996, 14, 379-86.
(32) Bastow, K. F.; Derse, D. D.; Cheng, Y. C. Susceptibility of
Phosphonoformic Acid-Resistant Herpes Simplex Virus Variants
to Arabinosylnucleosides and Aphidicolin. Antimicrob. Agents
Chemother. 1983, 23, 914-917.
(6) Zhu, X. K.; Guan, J .; Tachibana, Y.; Bastow, K. F.; Cho, S. J .;
Cheng, H. W.; Cheng, Y. C.; Gurwith, M.; Lee, K. H. Antitumor
Agents 194. Synthesis and Biological Evaluations of 4â-Mono-,
Di-, and Trisubstituted Aniline-4′-O-demethyl-podophyllotoxin
and Related Compounds with Improved Pharmacological Pro-
files. J . Med. Chem. 1999, 42, 2441-2446 and literature cited
therein.
(33) Lee, K. H. Unpublished results.
(34) Akanitapichat, P.; Lowden, C. T.; Bastow, K. F. 1,3-Dihydroxy-
acridone Derivatives as Inhibitors of Herpes Virus Replication.
Antiviral Res. 2000, 45, 123-134.
(35) Hoover, D. R.; Saah, A. J .; Bacellar, H.; Phair, J .; Detels, R.;
Anderson, R. Clinical Manifestations of AIDS in the Era of
Pneumocystis Prophylaxis. Multicenter AIDS Cohort Study. N.
Engl. J . Med. 1993, 329, 1922-1926.
(36) Boppana, S. B.; Fowler, K. B.; Vaid, Y.; Hedlund, G.; Satgno,
S.; Britt, W. J .; Pass, R. F. Neuroradiographic Findings in the
Newborn Period and Long-Term Outcome in Children with
Symptomatic Congenital Cytomegalovirus Infection. Pediatrics
1997, 99, 409-414.
(37) J ohnson, R. A.; Huong, S. M.; Huang, E. S. Inhibitory Effect of
4-(4-Fluorophenyl-2-(4-hydroxyphenyl)-5-(4-pyridyl)1H-imida-
zole on HCMV DNA Replication. Antiviral Res. 1999, 41, 101-
111.
(38) Huang, E. S.; Benson, J . D.; Huong, S. M.; Wilson, B.; van der
Horst, C. Irreversible Inhibition of Human Cytomegalovirus
Replication by Topoisomerase II Inhibitor, Etoposide: a New
Strategy for the Treatment of Human Cytomegalovirus Infection.
Antiviral Res. 1992, 17, 17-32.
(7) Budman, D. R.; Igwemezie, L. N.; Kaul, S.; Behr, J .; Lichtman,
S.; Schulman, P.; Vinciguerra, V.; Allen, S. L.; Kolitz, J .; Hock,
K.; O’Neil, K.; Schacter, L.; Barbhaiya, R. H. Phase I Evaluation
of a Water-Soluble Etoposide Prodrug, Etoposide Phosphate,
Given as a 5-Minute Infusion on Days 1, 3, and 5 in Patients
with Solid Tumors. J . Clin. Oncol. 1994, 12, 1902-1909.
(8) Engelborghs, Y.; Fitzgerald, T. J . A Fluorescence Stopped Flow
Study of the Competition and Displacement Kinetics of Podo-
phyllotoxin and the Colchicine Analogue 2-Methoxy-5-(2′,3′,4′-
trimethoxyphenyl)tropone on Tubulin. J . Biol. Chem. 1987, 262,
5204-5209.
(9) Yanagida, M.; Sternglanz, R. In DNA Topology and its Biological
Implications; Cozzarelli, N. R., Wang, J . C., Eds.; Cold Spring
Harbor Laboratory Press: Cold Spring Harbor, NY, 1990; pp
243-263.
(10) Huang, T. S.; Shu, C. H.; Yang, W. K.; Whang-Peng, J . Activation
of CDC25 Phosphatase and CDC2 Kinase Involved in GL331-
Induced Apoptosis. Cancer Res. 1997, 57, 2974-2978.
(11) Shimizu, T.; O’Connor, P. M.; Kohn, K. W.; Pommier, Y.
Unscheduled Activation of Cyclin B1/CDC2 Kinase in Human
Promyelocytic Leukemia Cell Line HL60 Cells Undergoing
Apoptosis Induced by DNA Damage. Cancer Res. 1995, 55, 228-
231.
(39) Huang, E. S.; Lee, K. H. Unpublished results.
(40) Cho, S. J .; Tropsha, A.; Suffness, M.; Cheng, Y. C.; Lee, K. H.
Antitumor Agents. 163. Three-Dimensional Quantitative Struc-
ture-Activity Relationship Study of 4′-O-Demethylepipodophyl-
lotoxin Analogues Using the Modified CoMFA/q2-GRS Approach.
J . Med. Chem. 1996, 39, 1383-1395.
(12) Van Maanen, J . M.; Retel, J .; de Vries, J .; Pinedo, H. M.
Mechanism of Action of Antitumor Drug Etoposide: a Review.
J . Natl. Cancer Inst. 1988, 80, 1526-1533.
J M000377F