Potent 4-aminopiperidine based antimalarial agents

Article abstract of DOI:10.1016/j.bmcl.2004.10.069

This paper describes potent antimalarial compounds effective against a multi-drug-resistant strain of P. falciparum (Dd2) with minimized toxicity, acceptable log D values, and stability to hepatocytic metabolism. A series of compounds with potent activity

Full text of DOI:10.1016/j.bmcl.2004.10.069

Bioorganic & Medicinal Chemistry Letters 15 (2005) 345–348
Potent 4-aminopiperidine based antimalarial agents
Kristin M. Brinner,a, Mary Ann Powles,b Dennis M. Schmatzb and Jonathan A. Ellmana,*
aCenter for New Directions in Organic Synthesis, Chemistry Department, University of California, Berkeley, CA 94720, USA
bMerck and Co., Rahway, NJ 07065, USA
Received 17 July 2004; revised 20 October 2004; accepted 23 October 2004
Available online 13 November 2004
Abstract—A series of compounds with potent activity against a multi-drug-resistant strain of Plasmodium falciparum, the causative
agent of the deadliest strain of malaria, is described. These compounds were also tested for cytotoxicity in human foreskin fibroblast
assays, evaluated to determine their logD, and assayed for metabolism by human and murine hepatocytes. This work resulted in the
development of compounds 9e and 10d, which showed good potency (IC50 = 75nM and <60nM, respectively, against Dd2), accept-
able logD values, and reasonable metabolic stability.
Ó 2004 Elsevier Ltd. All rights reserved.
Malaria is a disease of worldwide implications. It is
estimated that 40% of the worldÕs population resides
in malaria-endemic regions of the world. Over 300 mil-
lion cases are reported annually, resulting in 1–3 mil-
lion deaths.1 The clinical symptoms and the majority
of deaths caused by malaria occur during the intra-
erythrocytic phase of the parasiteÕs complex life cycle
within its human host. While the parasites develop
within human red blood cells, fever, chills, anemia,
and more severe clinical complications are all ob-
served.2 During this phase of the parasiteÕs life cycle,
hemoglobin is consumed as a primary source of amino
acids to fuel the explosive growth of the parasite. Upon
catabolism of hemoglobin, heme is released, which
would be cytotoxic to the parasite in its soluble form.
However, Plasmodium have evolved a unique heme
detoxification pathway, where free heme is dimerized
whereupon hydrogen bonding between heme dimers re-
sults in polymeric non-toxic hemozoin formation.3
There is considerable evidence that the aminoquinoline
based antimalarials, including chloroquine, quinine,
and mefloquine, act through inhibiting this heme
detoxification pathway.4 Unfortunately, increasing drug
resistance to these agents complicates the treatment of
malaria. Additionally, resistance to these common anti-
malarial medications cannot be overcome by increasing
their dosages, as they have extremely narrow therapeu-
tic windows.
We previously reported the synthesis and evaluation of a
novel series of amino-piperidine based compounds that
showed promising activity against both chloroquine-
sensitive (3D7) and resistant (W2) strains of P. falcipa-
rum, the deadliest strain of the malarial parasite.5 While
the mechanism of action of these compounds was not
known, there was evidence that they acted by inhibiting
heme polymerization. When comparing the structure of
the best compound from this series to that of chloroqu-
ine, we observed that the basic nitrogens are arrayed in a
similar fashion (Fig. 1).
By combining structural features of both compound 1
and chloroquine, we hoped to further improve the activ-
ity of the compounds. This would allow us not only to
capitalize on a drug intervention pathway already well
established by the aminoquinoline based drugs, but also
to bypass established mechanisms of resistance develop-
ment by avoiding the aminoquinoline scaffold itself.
Upon synthesis and evaluation of compound 2, we ob-
served that the compound maintained potent activity
against a chloroquine sensitive strain of P. falciparum,
but had significantly less activity against a multi-drug-
resistant strain of the parasite (Dd2). However, we were
encouraged to optimize the activity of additional com-
pounds of this structural type, because 2 had low levels
of toxicity.
Keywords: Antimalarial; Plasmodium falciparum; Multi-drug resistant;
Metabolism; LogD.
*
Corresponding author. Tel.: +1501 642 4488; fax: +1501 642
8369; e-mail: ellman@cchem.berkeley.edu
Present address: Chiron Corporation, Emeryville, CA 94608, USA.
Here we report the synthesis of analogs of compound 2,
their activity against a multi-drug-resistant strain of
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.10.069

346
K. M. Brinner et al. / Bioorg. Med. Chem. Lett. 15 (2005) 345–348
BnO
Boc
N
O
H
H
Cl
a, b
N
N
c, d
N
Ph
N
Ph
(66%)
1
N
N
quantitative
(2 steps)
IC50 30 nM (3D7)
6
H
N
Boc
N
O
N
N
e
, f
R
H
NH
3
Chloroquine
IC50 6 nM (3D7)
Cl
7a, 71%
7b, 62%
7c, 45%
7d, 63%
(2 steps)
H
N
N
H
N
N
N
R1
Ph
N
2
Ph
IC50 30 nM (3D7)
IC50 990 nM (Dd2)
Scheme 1. Reagents and conditions: (a) NaBH(OAc)3, AcOH, Et3N,
3-chloroaminopropane; (b) Boc2O, THF; (c) Et2NH, NaI, DMF,
70°C; (d) NH4HCO2, Pd/C, MeOH, EtOAc; (e) NaBH(OAc)3; (f) HCl,
dioxane.
Figure 1. Merging of structures of lead compounds and chloroquine.
P. falciparum, and their cytotoxicity. These compounds
showed promising activity as well as low levels of toxi-
city, and were further characterized for their drug-like
properties and metabolic lability.
kyl chloride 6 was then reacted with ethylamine, and the
benzyl group was subsequently removed under transfer
hydrogenation conditions to afford 3. The free piperi-
dine 3 was then reductively alkylated with a series of
four aldehydes, followed by acidic deprotection, to af-
ford compounds 7a–d.
Three regions of compound 2 could easily be modified
to rapidly assess structure–activity relationships (SAR)
(Fig. 2). The R1 substitution could be varied through
reductive alkylation of amino piperidine 3. Additionally,
R2 and R3 could be varied through alkylation of a vari-
ety of primary and secondary amines with alkyl chloride
4. Finally, R2 could be varied while holding R3 as an
ethyl substituent, by reductive alkylation of secondary
amine 5 with a series of aldehydes. As these synthetic se-
quences were efficient and generated sufficient material
for biological testing, they were not further optimized.
All compounds were purified by standard silica gel chro-
matography, and converted to the HCl salts for biolog-
ical testing.
Diversity at the R2 and R3 positions was investigated
through reaction of a variety of primary and secondary
amines with compound 4 (Scheme 2). 4-Piperidone
monohydrate hydrochloride was reductively alkylated
with 1,1-diphenylacetaldehyde to afford amino piper-
idone 8. This was then subjected to reductive amination
with 3-chloroaminopropane, and the resultant second-
ary amine was Boc protected to afford 4. Primary alkyl
chloride 4 was reacted with a series of five amines, fol-
lowed by cleavage of the Boc group under acidic condi-
tions to afford compounds 9a–e.
Finally, R3 was varied while holding R2 constant
(Scheme 3). Primary alkyl chloride 4 was reacted with
Diversity at the R1 position was investigated through
derivatization of 3 (Scheme 1). N-Benzyl-4-piperidone
was reductively alkylated with 3-chloroaminopropane,
and the resultant amine was Boc protected. Primary al-
HO
HO
O
a
b, c
Ph
N
NH2Cl
(77%)
R2
Ph
(78%, 2 steps)
H
N
N
R3
8
N
R1
Boc
Cl
N
d, e
Ph
R1CH2CHO
[H]
R2CHO
[H]
N
Ph
Boc
N
H
N
4
Boc
N
Ph
R2
N
N
9a, 22%
9b, 30%
9c, 45%
9d, 75%
9e, 51%
(2 steps)
N
H
5
R2R3NH
Ph
N
R3
NH
Ph
3
Boc
N
N
Ph
Cl
Ph
N
Scheme 2. Reagents and conditions: (a) 1,1-diphenylacetaldehyde,
NaBH(OAc)3, Et3N, AcOH, MeOH; (b) NaBH(OAc)3, Et3N, AcOH,
CH3CN, 3-chloroaminopropane; (c) Boc2O, THF; (d) NaI, DMF,
70°C; (e) HCl, dioxane.
Ph
4
Figure 2. Modification of 2 to assess SAR.

K. M. Brinner et al. / Bioorg. Med. Chem. Lett. 15 (2005) 345–348
347
Table 2. SAR with symmetrically substituted amines
Boc
N
Boc
N
Cl
R2
N
a
Ph
HN
H
Ph
N
N
R3
Ph
Ph
N
(73%)
Ph
N
Ph
5
4
Compd
R2
R3
IC50 (nM)a
Tox. (nM)b
O
10a, 25%
10b, 44%
10c, 38%
10d, 28%
(2 steps)
H
N
2
Ethyl
Methyl
Ethyl
Ethyl
Methyl
H
900
300
9900
1300
2600
2500
1200
4700
R3
N
, c
b
R3
H
Ph
9a
9b
9c
9d
9e
N
2500
200
Ph
–(CH2)4–
–(CH2)5–
n-Propyl
—
—
200
>60
Scheme 3. Reagents and conditions: (a) ethylamine, NaI, DMF, 70°C;
(b) NaBH(OAc)3, (ClCH2)2; (c) HCl, dioxane.
n-Propyl
a Dd2 P. Falciparum.
b Human foreskin fibroblast.
ethylamine to afford secondary amine 5, which was
reductively alkylated with a series of four aldehydes.
The Boc group was removed under acidic conditions
to afford compounds 10a–d.
amine at this position. Cyclic amines 9c and 9d showed
increases in activity and maintained an approximate
activity to toxicity ratio of 10:1. Compound 9e7 showed
a 15-fold increase in activity when compared to parent
compound 2, while improving the activity to toxicity ra-
tio from 11:1 to over 78:1.
The antimalarial activity of 7a–7d, 9a–9e, and 10a–d
were measured in red-blood-cell-based assays. Efficacy
3
was monitored by parasite H-hypoxanthine incorpora-
tion using parasite-infected human erythrocytes in late
ring stage.6 All compounds were assayed in duplicate
against the multi-drug-resistant parasite, Dd2. The cyto-
toxicities of these new compounds were measured in
quadruplicate against a human foreskin fibroblast assay.
Average values for both of these assays are reported.
The final set of data is shown in Table 3. All changes
from the diethylamine to unsymmetrically substituted
tertiary amines resulted in increases in activity, and most
compounds maintained similar levels of activity to toxi-
city ratios. However, compound 10d8 showed over a 10-
fold increase in activity when compared to 2, while also
improving its activity to toxicity ratio from only 11:1 to
over 100:1.
As shown in Table 1, compounds with aromatic func-
tionality at R1 showed improved antimalarial activity.
While compounds 7c and 7d had better activity than
parent compound 2, they had worse activity to toxicity
ratios. Consequently, further optimization of activity
was carried out through varying substituents at posi-
tions R2 and R3.
We next attempted to better understand the activity of
these compounds through correlation of their activity
with their lipophilicity profile, or LogD. A subset of
the compounds was tested and their LogD plotted ver-
sus their activity (Table 4). As the lipophilicity increases,
the activity of the compound increases. However, there
is no obvious trend when examining lipophilicity versus
toxicity.
The data in Table 2 demonstrates the striking change in
activity and toxicity observed through small permuta-
tions in the R2 and R3 substituents of compound 2.
Smaller substituents at R2 and R3 resulted in increased
activity, but with even greater increases in toxicity (com-
pound 9a). Secondary amine 9b has decreased activity
and a worse activity to toxicity ratio relative to tertiary
amine 2, underlying the importance of having a tertiary
Two of the most promising compounds, 9e and 10d,
were also tested to determine their metabolic vulnerabi-
lity. This assay uses LC/MS/MS to determine the degree
to which the compounds are metabolized upon exposure
Table 3. SAR with unsymmetrically substituted amines
Table 1. SAR in the R1 region
H
H
R3
N
N
N
N
Ph
N
N
R1
Ph
Compd
R1
IC50 (nM)a
Tox. (nM)b
Compd
R1
IC50 (nM)a
Tox. (nM)b
2
CH(C6H5)2
C6H11
900
4300
920
9900
6000
2900
2700
1200
2
CH(C6H5)2
C6H11
900
200
250
200
75
9900
2200
2100
2000
7600
7a
7b
7c
7d
10a
10b
10c
10d
CH2C6H11
CH2C6H4 (4-Cl)
CH2–C6H4 (4-C6H5)
CH2C6H11
CH2C6H4(4-Cl)
CH2–C6H4(4-C6H5)
850
400
a Dd2 P. Falciparum.
b Human foreskin fibroblast.
a Dd2 P. Falciparum.
b Human foreskin fibroblast.

348
K. M. Brinner et al. / Bioorg. Med. Chem. Lett. 15 (2005) 345–348
Table 4. Comparison of IC50 and toxicity with LogD
Acknowledgements
Compd
IC50 (lM)a
Tox. (lM)b
LogDc
We gratefully acknowledge the support of the Medicines
for Malaria Venture. The Center for New Directions in
Organic Synthesis is supported by Bristol-Myers Squibb
as a Sponsoring Member.
7a
9b
7b
7c
4.3
2.5
6.0
2.6
2.9
2.7
1.2
1.3
2.5
1.2
2.2
7.6
4.7
À0.25
À0.44
À0.59
À0.70
0.7
0.92
0.85
0.40
0.30
0.20
0.20
0.20
0.075
>0.060
7d
9a
9c
0
0.11
0.77
2.11
3.12
1.41
References and notes
9d
10a
10d
9e
1. Woster, P. M. Annu. Rep. Med. Chem. 2001, 36, 99.
2. Krogstad, D. J. In Mechanisms of Microbial Disease, 2nd
Edn.; Schaechter, M., Medoff, G., Eisenstein, B. I., Eds.;
Williams and Wilkins: Baltimore, MD, 1993; pp 597–615.
3. Pagola, S.; Stephens, P. W.; Bohle, D. S.; Kosar, A. D.;
Madsen, S. K. Science 2000, 404, 307.
a Dd2.
b Human Foreskin Fibroblast.
c Determined by Robertson Microlit Laboratories, Inc. (Madison, NJ).
4. Choi, C. Y. H.; Schneider, E. L.; Kim, J. M.; Gluzman, I.
Y.; Goldberg, D. E.; Ellman, J. A.; Marletta, M. A. Chem.
Biol. 2002, 9, 881; For recent advances in new heme-
polymerization inhibitors, see: OÕNeill, P. M.; Mukhtar, A.;
Stocks, P. A.; Randle, L. E.; Hindley, S.; Ward, S. A.;
Storr, R. C.; Bickley, J. F.; OÕNeil, I. A.; Maggs, J. L.;
Highes, R. H.; Winstanley, P. A.; Bray, P. G.; Park, B. K.
J. Med. Chem. 2003, 46, 4933, and references cited therein.
5. Brinner, K. M.; Kim, J. M.; Habashita, H.; Gluzman, I. Y.;
Goldberg, D. E.; Ellman, J. A. Bioorg. Med. Chem. 2002,
10, 3649.
Table 5. Metabolism by liver microsomes
Compd
Human S9% remaininga
Mouse S9% remaininga
9e
10d
56
45
118
113
Experiment performed by PPDTM Discovery (Morrisville, NC).
a Single time point at 1h.
6. Gluzman, I. Y.; Francis, S. E.; Oksman, A.; Smith, C. E.;
Duffin, K. L.; Goldberg, D. E. J. Clin. Invest. 1994, 93,
1602.
to murine and human liver S9-fractions for one hour. As
seen in Table 5, both compounds are only 50% degraded
by human S9 fractions, and are not degraded at all by
murine S9 fractions.
7. 1H NMR (400MHz, MeOD-d3) d 0.91(t, J = 7.2Hz, 6H),
1.51–1.79 (m, 6H), 1.80 (q, J = 6.8Hz, 2H), 1.90 (m, 2H),
2.11 (m, 2H), 2.60 (m, 4H), 2.76 (t, J = 6.8Hz, 2H), 2.80 (m,
1H), 2.95 (d, J = 6.8Hz, 2H), 3.02 (m, 4H), 4.21(t,
J = 7.6Hz, 1H), 7.12 (m, 2H), 7.23 (m, 8H). 13C NMR
(125MHz, MeOD-d3) d 11.9, 20.0, 24.2, 30.7, 45.8, 49.5,
50.3, 53.1, 54.0, 56.3, 64.1, 127.3, 129.1, 129.4, 145.2. ESI-
MS (LR) [M+H]+ calcd for C28H43N3, 422.7; found, 422.5.
8. 1H NMR (500MHz, MeOD-d3) d 1.07 (t, J = 7.5Hz, 3H),
1.24 (m, 2H), 1.66 (m, 4H), 1.91 (t, J = 12Hz, 2H), 2.39 (m,
1H), 2.62 (m, 6H), 2.77 (m, 4H), 2.83 (m, 2H), 2.93 (d,
J = 7.5Hz, 2H), 4.13 (t, J = 7.5Hz, 1H), 7.12 (m, 2H), 7.23
(m, 8H), 7.32 (m, 3H), 7.42 (m, 2H), 7.56 (m, 2H), 7.61(m,
2H). 13C NMR (125MHz, MeOD-d3) d 11.5, 25.9, 31.4,
33.2, 46.0, 48.1, 50.3, 52.7, 53.5, 55.5, 56.2, 64.3, 127.2,
127.8, 128.0, 128.3, 129.1, 129.4, 129.9, 130.03, 140.2, 140.7,
142.1, 145.3. ESI-MS (LR) [M+H]+ calcd for C38H47N3,
546.8; found, 546.5.
We have further improved the activity of our new class
of non-aminoquinoline antimalarial compounds. These
compounds are easily and rapidly synthesized from sim-
ple, commercially available starting materials. In addi-
tion to improving the activity of initial hit 2 with an
IC50 of 900nM to lead compound 9e with an IC50 of less
than 60nM against a multi-drug-resistant strain of P.
falciparum, we improved the activity to cytotoxicity ra-
tio from 11:1 to 78:1. The most promising compounds
had acceptable logD values and were reasonably meta-
bolically stable in both human and murine liver S-9
fractions.