Podophyllotoxin analogues active versus Trypanosoma brucei

Article abstract of DOI:10.1016/j.bmcl.2010.01.009

In an effort to discover novel anti-trypanosomal compounds, a series of podophyllotoxin analogues coupled to non-steroidal anti-inflammatory drugs (NSAIDs) has been synthesized and evaluated for activity versus Trypanosoma brucei and a panel of human cell lines, revealing compounds with low nano-molar potencies. It was discovered that coupling of NSAIDs to podophyllotoxin increased the potencies of both compounds over 1300-fold. The compounds were shown to be cytostatic in nature and seem to act via de-polymerization of tubulin in a manner consistent with the known activities of podophyllotoxin. The potencies against T. brucei correlated directly with Log P values of the compounds, suggesting that the conjugates are acting as hydrophobic tags allowing podophyllotoxin to enter the cell.

Full text of DOI:10.1016/j.bmcl.2010.01.009

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1787–1791
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Podophyllotoxin analogues active versus Trypanosoma brucei
Md. Jashim Uddin a, , David C. Smithson b,c, , Kristin M. Brown b, Brenda C. Crews a, Michele Connelly b,
Fangyi Zhu b, Lawrence J. Marnett a, R. Kiplin Guy b,
*
a Department of Biochemistry, Chemistry and Pharmacology, Vanderbilt Institute of Chemical Biology, Vanderbilt University School of Medicine, Nashville, TN 37232, USA
b St. Jude Children’s Hospital, Department of Chemical Biology and Therapeutics, 262 Danny Thomas Place, Memphis, TN 38105-3678, USA
c University of California San Francisco, Chemistry and Chemical Biology Graduate Program, San Francisco, CA 94158-2517, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
In an effort to discover novel anti-trypanosomal compounds, a series of podophyllotoxin analogues cou-
pled to non-steroidal anti-inflammatory drugs (NSAIDs) has been synthesized and evaluated for activity
versus Trypanosoma brucei and a panel of human cell lines, revealing compounds with low nano-molar
potencies. It was discovered that coupling of NSAIDs to podophyllotoxin increased the potencies of both
compounds over 1300-fold. The compounds were shown to be cytostatic in nature and seem to act via
de-polymerization of tubulin in a manner consistent with the known activities of podophyllotoxin. The
potencies against T. brucei correlated directly with Log P values of the compounds, suggesting that the
conjugates are acting as hydrophobic tags allowing podophyllotoxin to enter the cell.
Received 10 November 2009
Revised 29 December 2009
Accepted 4 January 2010
Available online 11 January 2010
Keywords:
Sleeping sickness
Trypanosoma brucei
Microtubule
Ó 2010 Elsevier Ltd. All rights reserved.
The eukaryotic parasite Trypanasoma brucei is the causative
agent of Human African Trypanosomiasis, a major health concern
in much of sub-Saharan Africa.1 Current treatments for this disease
are highly toxic or suffer from poor pharmacokinetic properties
which necessitate lengthy dosing regimens.2 The need for novel
therapeutics is further exacerbated by developing resistance to
clinically used drugs.3
Previous work in developing anti-trypanosomal compounds
targeted at the related parasite Trypanosoma cruzi showed that
indomethacin amides were potent inhibitors of T. cruzi sterol
Following this discovery, the effects of podophyllotoxin (Fig. 1B)
and indomethacin (C) as both single agents and in various fixed ra-
tio combinations (1:1, 1:10, and 1:100 in both directions) were
tested. The compounds were inactive (IC50 > 50 lM) under all
tested conditions, indicating that this was not a simple case of syn-
ergy. While podophyllotoxin was inactive in T. brucei, it was active
in transformed mammalian cell lines, with similar potencies to
compound 1, but not in primary human cell lines. (Table 1)
14a-demethylase (TCCYP51), an important enzyme in the sterol
biosynthetic pathway.4–6 These compounds were also active in
in vitro whole cell assays. Since T. brucei shares this pathway and
also possesses an active sterol 14a-demethylase, it was decided
to test this compound series for anti-trypanosomal activity in this
parasite as well.7
While most of the analogues were relatively inactive
(IC50 > 10
lM) versus T. brucei, a small set of compounds (1–3)
linked to podophyllotoxin showed very good potencies (IC50 values
from 51 6 nM to 5.3 0.8
activity relationship.
lM) as well as a preliminary structure–
The observed activities are unlikely to be a result of CYP51 inhi-
bition, since the T. cruzi and T. brucei enzymes are highly homolo-
gous, and the observed SAR for these compounds is significantly
different from that observed for trypanosomal CYP51 inhibitors.4
* Corresponding author. Tel.: +1 901 595 5714; fax: +1 901 595 5715.
E-mail address: kip.guy@stjude.org (R.K. Guy).
These authors contributed equally to this Letter.
Figure 1. Strucutres of compounds A–D.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.009

Table 1
Primary hits (1–3) NSAID COX inhibitor analogues (4–7), non-NSAID based analogues (8–10), and linker modifications (11,12) tested for anti-trypanosomal activities
a
a
a
c
c
No. Compounds
T. brucei IC50
M)
HEPG2 IC50
M)
BJ IC50
M)
% Tubulin
COX-1 IC50
M)
COX-2 IC50
M)
T1/2 plasmad
(h)
A Log P
(
l
(
l
(
l
inhibitionb
(
l
(
l
1
0.051 (0.006)
0.20 (0.06)
>50
>50
60 (8)
>4
>4
0.295
0.254
2.7 (0.1)
6.2
2
3.4 (0.2)
1.0 (0.2)
NT
NT
6.2
3
4
5.3 (0.8)
0.32 (0.09)
>50
>50
NT
>4
>4
>4
>4
NT
6.8
5.6
0.46 (0.08)
0.22 (0.07)
64 (8)
4.1 (0.3)
5
6
0.17 (0.01)
0.27 (0.1)
>50
>50
NT
NT
>4
>4
>4
>4
9.7 (0.6)
5.6
4.8
4.5 (0.4)
0.54 (0.36)
NT
7
>50
>50
>50
89 (9)
>4
>4
6.2 (0.4)
5.3

8
9
32 (12.4)
31 (12.7)
0.44 (0.22)
>50
>50
74 (6)
>4
>4
>4
>4
NT
NT
4.3
3.8
0.090 (0.061)
NT
10
11
27 (15)
0.028 (0.015)
>50
>50
58 (5)
>4
>4
>4
2.6 (0.1)
5.1
6.6
0.22 (0.02)
0.50 (0.23)
NT
0.147
2.9 (0.1)
12
0.99 (0.08)
0.71 (0.33)
>50
NT
>4
0.602
2.9 (0.2)
5.8
A
B
C
D
Succinylpodophyllotoxin
Podophyllotoxin
Indomethacin
>50
>50
>50
0.15 (0.2)
0.17 (0.019)
0.024 (0.005)
>50
>50
>50
>50
>50
NT
103 (0.8)
NT
>4
>4
0.05
NT
>4
>4
0.75
NT
NT
9.8 (0.4)
NT
2.3
2.1
4.2
5.5
Rhizoxin
0.40 (0.19)
102 (2)
NT
(NT = not tested).
a
Values are expressed as the mean of three experiments with standard deviation given in parentheses. HEPG2 is a liver carcinoma cell line while BJ are normal human foreskin fibroblasts. The IC50 values for pentamidine and
eflornithine versus T. brucei under the tested conditions were 25 10 nM and 243 32 M, respectively.
Values are expressed as the mean of four experiments with standard deviation given in parentheses. Compounds were tested at a single concentration of 10
l
b
l
M for their ability to inhibit tubulin polymerization.
IC50 values were determined by incubating several concentrations of inhibitors in DMSO with purified murine COX-2 (63 nM) and ovine COX-1 (22.5 nM) for 20 min followed by treatment with 1-14C-AA (50
M) at 37 °C for
30 s. Assays were run in duplicates and IC50 values are reported as the mean of both experiments. IC50 values for control COX inhibitors are as follows: Celecoxib (COX-1 IC50 = >4 M, COX-2 IC50 = 0.21 M) Rofecoxib (COX-1
IC50 > 4 M, COX-2 IC50 = 0.15 M).
Values for half-life in mouse plasma are expressed as the mean of three experiments with standard deviation given in parentheses.
c
l
l
l
l
l
d

1790
M. J. Uddin et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1787–1791
Indomethacin did not show toxicity in any of the cell lines tested.
Following these observations, further experiments aimed at de-
convoluting the mechanism of action of compound
performed.
1 were
Compound 1 was shown to be reversibly cytostatic in T. brucei
and mammalian cell lines with similar potencies in both. (Fig 2, Ta-
ble 1) However, when tested against primary human mammary
epithelial or foreskin fibroblast cell lines, 1 appeared completely
non-toxic. Since podophyllotoxin, a tubulin destabilizing agent, is
known to be a reversible cytostatic compound the effects of 1 on
T. brucei tubulin was evaluated. Staining of T. brucei in the presence
of 1 lM compound 1 with Tubulin Tracker (Molecular Probes, Eu-
gene, OR), a fluorescently labeled taxol analogue which binds to
polymerized tubulin, showed that the test compound caused a de-
crease in the amount of polymerized tubulin present in the cell.
Rhizoxin (D), a microtubule toxin known to be active in T. brucei
displayed similar potencies (IC50 = 150 20 nM) and phenotypes
in these assays. After 24 h of exposure to either 1 or D, the cells be-
came rounded and began displaying aberrant nuclei counts. (Fig 3)
This phenotype has been reported for 100 nM rhizoxin and has
been interpreted as an indication of inhibition of microtubule
dependent cellular processes.8 These results suggest that the com-
pound is acting via the classical podophyllotoxin mechanism in
which polymerization of the microtubules is inhibited.9
This caused us to hypothesize that indomethacin was increasing
the import of podophyllotoxin into the cell. Indomethacin is a non-
steroidal anti-inflammatory drug (NSAID) which works by inhibit-
ing the cyclooxygenase enzymes (COX-1 and COX-2), which are
not present in T. brucei.10,11 Therefore, it is unlikely that the NSAID
was acting in COX dependent manner. A second possible mecha-
nism for the increased uptake is the sterol pathway. T. brucei is
known to possess high affinity transporters for low density
lipoprotein (LDL) and it is possible that the linked NSAID could
be acting as a cholesterol mimetic and entering the cells via this
pathway.12
To determine if the observed increase in potencies was NSAID
dependent, a set of podophyllotoxin analogues conjugated to vari-
ous NSAIDs as well as non-NSAID groups were synthesized. The ef-
fects of varying the linker were also evaluated in this effort.
Synthesis of the test library of podophyllotoxin conjugates was
performed using liquid phase parallel synthetic methodologies,
where succinyl-podophyllotoxin was reacted with tethered NSA-
IDs or other moieties containing terminal amino-groups. The
succinylpodophyllotoxin (A) was synthesized from a reaction of
succinic anhydride and podophyllotoxin. Tethered NSAIDs contain-
ing a terminal amino-group were synthesized from the reaction of
carboxylic acid containing NSAIDs with BOC-protected alkyldimine
or pyperzinyl, or aminomethylaniline in the presence of ethyl-1-
{3-(dimethylamino)propyl}-3-ethylcarbodiamide (EDCI) followed
Figure 3. Phase contrast and fluorescence images of T. brucei showing the effect of 1
(1 M, 24 h exposure) and D (100 nM, 24 h exposure) on tubulin polymerization
and DNA content. Green fluorescence indicates polymerized tubulin as determined
l
by Tubulin Tracker™ while blue is Hoescht stained DNA.
by treatment with HCl gas. Compounds were purified via LC/MS
and Prep-HPLC. Structures were assigned by NMR and mass spec.
analysis. All compounds were tested in cytotoxicity assays versus
T. brucei, HEPG2, HEK293, BJ, and Raji cell lines. In addition, com-
pounds were tested for their ability to inhibit COX-I and COX-II en-
zymes as well as tubulin polymerization.
After testing, it was clear that compounds containing NSAID
groups (4–7) were more potent against T. brucei than those with-
out NSAIDs (8–10). However, with the exception of some indo-
methacin containing compounds (1, 2, 11, 12), most of the
compounds were unable to directly inhibit purified human COX
enzymes. This suggests that if a NSAID binding target in T. brucei
is responsible for the observed activities, the pharmacophore nec-
essary for activity is substantially different than that for mamma-
lian COX enzymes. All of the tested compounds displayed
significant selectivity (>100-fold) for T. brucei and transformed hu-
man cell lines versus primary fibroblast and epithelial human cells.
The modest set of linker modifications explored (1, 2, 11, and
12) show that some modifications are tolerated. The lengthening
by a single methylene group (11), only lowers potency fourfold
against T. brucei while the piperizine linker (12) lowers potency
20-fold. Shortening the linker and substituting an ester linkage
for the amide (2), lowers potency by 60-fold. Compounds 4 and
8, also follow this pattern.
Since the linkers used in this study contained ester and amide
moieties, a subset of the compounds were tested for stability in
simulated gastric fluid (SGF, pH 1) as well as stability in mouse
plasma. All tested compounds displayed T1/2 > 24 h in SGF. Half-
lives in plasma ranged from 2.6 to 9.8 h. Plasma T1/2 was affected
more by the NSAID moiety than linker type, indicating that stable
compounds based on this scaffold may be achievable. (see com-
pounds 1, 5, 10–12).
The ability of the compounds to inhibit tubulin polymerization
did not correlate with anti-trypanosomal potencies, suggesting
that cellular uptake was driving the observed differential activities.
In support of this, it was observed that the potencies against T. bru-
cei directly correlated with ALog P values. (P <0.001) T. brucei pos-
sesses an extremely high rate of membrane turnover, recycling the
cellular surface once every 12 min.13 If the hydrophobic conjugates
are interacting directly with the cellular membrane, this could re-
sult in the rapid uptake and concentration of the compounds.
In summary, we have reported the ability of NSAID motifs con-
jugated to podophyllotoxin to increase the potency of both com-
pounds against T. brucei by over 1300-fold. This phenomenon is
not a result of synergy of the two compounds and is hypothesized
Figure 2. T. brucei growth curve of 1
counted manually by haemocytometer. Values are a mean of three experiments.
lM 1 (j) and DMSO control (N). Cells were

M. J. Uddin et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1787–1791
1791
to be the result of increased cellular uptake, either via passive dif-
References and notes
fusion or via active uptake through the LDL import pathway. How-
ever, it remains a possibility that these compounds are acting
through other mechanisms, such as inhibition of topoisomerase
II, a known target of other podophyllotoxin analogues. Though
the use of drug conjugates has been previously reported in other
cases, it is extremely rare for the combination of two inactive com-
pounds to result in such dramatic increases in potency. Efforts to
further characterize this phenomenon are currently underway in
our laboratory.
1. WHO. Fact Sheet, Vol. 2007, World Health Organization Media Centre, 2006.
2. Barrett, M. P.; Boykin, D. W.; Brun, R.; Tidwell, R. R. Br. J. Pharmacol. 2007, 152,
1155.
3. Brun, R.; Schumacher, R.; Schmid, C.; Kunz, C.; Burri, C. Trop. Med. Int. Health
2001, 6, 906.
4. Konkle, M. E.; Hargrove, T. Y.; Kleshchenko, Y. Y.; von Kries, J. P.; Ridenour, W.;
Uddin, M. J.; Caprioli, R. M.; Marnett, L. J.; Nes, W. D.; Villalta, F.; Waterman, M.
R.; Lepesheva, G. I. J. Med. Chem. 2009, 52, 2846.
5. Lepesheva, G. I.; Zaitseva, N. G.; Nes, W. D.; Zhou, W.; Arase, M.; Liu, J.; Hill, G.
C.; Waterman, M. R. J. Biol. Chem. 2006, 281, 3577.
6. Lepesheva, G. I.; Ott, R. D.; Hargrove, T. Y.; Kleshchenko, Y. Y.; Schuster, I.; Nes,
W. D.; Hill, G. C.; Villalta, F.; Waterman, M. R. Chem. Biol. 2007, 14, 1283.
7. Lepesheva, G. I.; Hargrove, T. Y.; Ott, R. D.; Nes, W. D.; Waterman, M. R. Biochem.
Soc. Trans. 2006, 34, 1161.
Acknowledgments
8. Ploubidou, A.; Robinson, D. R.; Docherty, R. C.; Ogbadoyi, E. O.; Gull, K. J. Cell Sci.
1999, 112, 4641.
9. Wilson, L.; Meza, I. J. Cell Biol. 1973, 58, 709.
10. Murakami, M.; Kudo, I. Prog. Lipid Res. 2004, 43, 3.
11. Kubata, B. K.; Duszenko, M.; Kabututu, Z.; Rawer, M.; Szallies, A.; Fujimori, K.;
Inui, T.; Nozaki, T.; Yamashita, K.; Horii, T.; Urade, Y.; Hayaishi, O. J. Exp. Med.
2000, 192, 1327.
We acknowledge support by the American Lebanese Syrian
Associated Charities (ALSAC), St. Jude Children’s Research Hospital,
and NIH Grant IU54CA105296. We thank Dr. J. Pietenpol for the
primary human mammary epithelial cells.
12. Coppens, I.; Baudhuin, P.; Opperdoes, F. R.; Courtoy, P. J. Proc. Natl. Acad. Sci.
U.S.A. 1988, 85, 6753.
Supplementary data
13. Grunfelder, C. G.; Engstler, M.; Weise, F.; Schwarz, H.; Stierhof, Y. D.; Morgan,
G. W.; Field, M. C.; Overath, P. Mol. Biol. Cell 2003, 14, 2029.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.01.009.