An automated, polymer-assisted strategy for the preparation of urea and thiourea derivatives of 15-membered azalides as potential antimalarial chemotherapeutics

Article abstract of DOI:10.1016/j.bmc.2011.01.030

A series of 15-membered azalide urea and thiourea derivatives has been synthesized and evaluated for their in vitro antimalarial activity against chloroquine-sensitive (D6), chloroquine/pyremethamine resistant (W2) and multidrug resistant (TM91C235) strains of Plasmodium falciparum. We have developed an effective automated synthetic strategy for the rapid synthesis of urea/thiourea libraries of a macrolide scaffold. Compounds have been synthesized using a solution phase strategy with overall yields of 50-80%. Most of the synthesized compounds had inhibitory effects. The top 10 compounds were 30-65 times more potent than azithromycin, an azalide with antimalarial activity, against all three strains.

Full text of DOI:10.1016/j.bmc.2011.01.030

Bioorganic & Medicinal Chemistry 19 (2011) 1692–1701
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
An automated, polymer-assisted strategy for the preparation of urea
and thiourea derivatives of 15-membered azalides as potential
antimalarial chemotherapeutics
ˇ ´ a,
Antun Hutinec a, , , Renata Rupcic , Dinko Ziher , Kirsten S. Smith b, Wilbur Milhous b, William Ellis b,
⇑
a,
ˇ
b
a,à
´
Colin Ohrt , Zrinka Ivezic Schönfeld
a GlaxoSmithKline Research Centre Zagreb Ltd, Prilaz baruna Filipovi´ca 29, 10000 Zagreb, Croatia
b Division of Experimental Therapeutics, Walter Reed Army Institute of Research, Washington, DC 20910, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of 15-membered azalide urea and thiourea derivatives has been synthesized and evaluated for
their in vitro antimalarial activity against chloroquine-sensitive (D6), chloroquine/pyremethamine resis-
tant (W2) and multidrug resistant (TM91C235) strains of Plasmodium falciparum. We have developed an
effective automated synthetic strategy for the rapid synthesis of urea/thiourea libraries of a macrolide
scaffold. Compounds have been synthesized using a solution phase strategy with overall yields of
50–80%. Most of the synthesized compounds had inhibitory effects. The top 10 compounds were 30–
65 times more potent than azithromycin, an azalide with antimalarial activity, against all three strains.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 5 September 2010
Revised 13 January 2011
Accepted 15 January 2011
Available online 22 January 2011
Keywords:
Antimalarial
Azalide
Ureas
Thiourea
Automated parallel synthesis
Plasmodium falciparum
1. Introduction
With the recent introduction of artemisinin-based combination
therapy as the first-line therapy in many countries where malaria
Today approximately 40% of the world’s population is at risk of
malaria, mostly those living in the poorest countries.1 Malaria is
found throughout the tropical and subtropical regions of the world
and causes more than 300 million acute illnesses as well as 0.7–
2.7 million deaths annually.2,3
In humans, malaria is caused by four strains of single-celled
eukaryotic protozoa of the genus Plasmodium. These strains in-
clude Plasmodium malariae, Plasmodium vivax, Plasmodium ovale
and Plasmodium falciparum of which P. falciparum is by far the most
deadly.4,5P. falciparum causes malaria tropica, which without treat-
ment, is often lethal for the infected patient.
The worldwide increase in drug-resistance to most affordable
chemotherapies,6,7 such as chloroquine 1, has created a demand
for the development of new malaria treatments. Currently, arte-
misinin 2 and its derivatives are the drugs of choice to treat multi-
drug resistant malaria8 (Fig. 1).
is endemic, combination therapies have become the standard of
care for the treatment of P. falciparum malaria. However, the
spread of multidrug resistance from Southeast Asia to significant
parts of the world where malaria is endemic, calls for the develop-
ment of novel antimalarial compounds. The challenge today in ma-
laria chemotherapy is to find safe, effective and affordable new
drugs lacking cross-resistance with standard antimalarial drugs.9
Antibiotics with antimalarial activity may offer an interesting
alternative for the treatment of multidrug resistant falciparum ma-
laria. Azithromycin, a widely prescribed advanced-generation mac-
rolide antibiotic with a favorable toxicological profile, has shown
intrinsic activity against Plasmodium spp. both in vitro10,11 and
in vivo in prophylaxis and in treatment.12–18 Compared to other
antibiotics used for malaria treatment (e.g., tetracyclines), azithro-
mycin offers unique advantages due to its safety in children and
experience with use in pregnant subjects,19 the populations most
affected by malaria.
Through screening a set of in-house compounds from our com-
pound bank against a panel of parasites, various urea- and thiourea
derivatives 3 of azithromycin were found to have higher in vitro
potency than azithromycin (Fig. 2). Therefore, our effort has been
focused on the development of an automated synthetic strategy
to quickly and efficiently produce urea and thiourea libraries of a
⇑
Corresponding author. Tel.: +385 1 8886367; fax: +385 1 8886444.
E-mail address: antun.hutinec@glpg.com (A. Hutinec).
Present address: Galapagos Research Centre Zagreb Ltd, Prilaz baruna Filipovic´a
29, 10000 Zagreb, Croatia.
à
Present address: Nabriva Therapeutics AG, Leberstrasse 20, 1110 Vienna, Austria.
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.01.030

A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
1693
Work-Up
High-Throughput
Evaporation
MiniBlock
H
Library Synthesis
O
N
HN
N
O
H
O
Sample
Preparation
Analysis
Purification
O
Cl
O
1
2
MiniMapper
FractionLynx (LC/MS)
Sample Processing
Station
Figure 1. Chloroquine and artemisinin.
H
N
R
N
R1
Balance Automator
biological screening
X
N
N
N
N
HO
O
HO
O
HO
OH
O
HO
OH
O
OH
OH
O
Figure 3. Automated 9a-aza-homoerythromycin-1-propyl urea and thiourea
O
O
O
library synthesis flowchart.
O
O
O
O
OH
OH
8{1–6}, derived from 9a-aza-homoerythromycin-1-propylamine
building block 4 and a set of isocyanate and isothiocyanate build-
ing blocks 5{1–6} and 6{1–6}, respectively (Chart 1), is shown in
Scheme 1.
O
O
Azithromycin
3
The synthesis of the amine 4 starts with the Michael addition of
acrylnitrile on 9-deoxo-9-dihydro-9a-aza-9a-homoerythromycin A
9. In the second step the cyanoethyl derivative 10 is hydrogenated
with PtO2 into the desired compound 4 (Scheme 2).23 In this sec-
ond step acetic acid was used as the solvent instead of ethanol. This
modification allowed the reaction to finish in 12 h at 5 bar instead
of 72 h at 20 bar, and additionally raised the yield of the product
from 40% to 60%.
Figure 2. Azithromycin and potentially active macrolide derivatives.
macrolide scaffold for the screening of potential antimalarial
compounds.
2. Results and discussion
2.1. Chemistry
Initially, the chemistry sequence was developed and exempli-
fied by the preparation of a representative 9a-aza-homoerythro-
mycin-9a-N-(
corresponding 9a-aza-homoerythromycin-9a-N-(
c
-aminopropyl) urea library 7{1 and 6} and the
The primary goal within any medicinal chemistry program is to
rapidly identify a lead compound in order to initiate the synthesis
of suitable derivatives and ultimately identify candidates with
good ADME properties, efficacy and safety. Solid-phase synthesis
techniques continue to be widely exploited for the preparation of
large compound libraries.20 This strategy has been further en-
hanced by the incorporation of automated compound synthesis
and processing protocols.21 However, in many cases, difficulty in
monitoring solid-phase chemistry, the inability to purify resin-
bound intermediates, and protracted chemistry development
times have focused attention on the need to develop complemen-
tary solution phase strategies for high-throughput chemistry.22
It is important when planning automated syntheses that the
synthetic procedures are chosen and optimized with regards to
the capabilities and characteristics (tube sizes, reaction block for-
mats, etc.) of the synthesizer to be used. Each step, such as solvent
addition, reagent addition, and evaporation, is considered in terms
of an automated synthesis because each step could represent a to-
tal success or failure in the preparation of libraries. For this reason,
it is helpful to represent the synthesis in the form of a flowchart in
which each successive synthesis operation is shown (Fig. 3). More-
over, it is preferable to rehearse chemistry on the robot from the
outset in order to develop robust protocols in the shortest time.
c-aminopropyl)
thiourea library 8{1 and 6}. The transformation was optimized on
an AdvancedChemtech PLS 4 Â 6—synthesizer in 4 mL vials. Hav-
ing developed the appropriate reaction conditions (Table 1), a set
of small trial arrays (‘rehearsals’, Chart 1) were prepared to estab-
lish that each of the chosen building blocks performed acceptably
under the reaction conditions chosen, before the full parallel li-
brary syntheses were initiated.
The preparation of the products 7{1–6} and 8{1–6} commences
with the addition between an excess of the isocyanate or isothiocy-
anate (1.3 equiv) and the amine 4 to afford the desired urea deriv-
atives 7{1–6}, and the corresponding thiourea derivatives 8{1–6}
and was observed to proceed efficiently at ambient temperature
in dichloromethane (CH2Cl2).
Following scavenging of excess 5{1–6} and 6{1–6} with a poly-
styrene-supported amino-resin (PS-trisamine, 4.11 mmol/g), the
products 7{1–6} and 8{1–6} were isolated in good yields (73–
90%) and high purity (90–96%; LC/MS). The use of a fivefold excess
of the resin ensured that complete scavenging was achieved within
6 h. The removal of excess 5{1–6} and 6{1–6} could also be per-
formed by the solid phase extraction methodology on a 1 g pre-
loaded SiO2 column.
However, we found that it is not possible to utilize a parallel
synthesis strategy using aromatic and aliphatic ureas in the same
library due to the big difference in the reactivity of the building
blocks. The reaction time for the reaction with aromatic isocya-
nates was about 60 min whereas with aliphatic isocyanates it took
2.1.1. Preparation oftrial library compounds 7{1–6} and 8{1–6}
Our approach to the 9a-aza-homoerythromycin-9a-N-(
nopropyl) urea derivatives 7{1–6}, and the corresponding 9a-aza-
homoerythromycin-9a-N-( -aminopropyl) thiourea derivatives
c-ami-
c

1694
A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
O
O
N
S
S
N
N
N
5(1)
6(1)
5(2)
5(4)
6(2)
O
S
N
Cl
N
S
N
O
N
O
O
Cl
5(3)
6(3)
6(4)
O
O
S
N
S
N
N
CF3
N
F
5(6)
6(5)
5(5)
6(6)
Chart 1. Set of isocyanate 5(x) and isothiocyanate 6(y) building blocks.
H
N
H
N
H
N
R
H
X
1,3 eq
X
N
N
N
N
N
R
HO
O
HO
O
HO
OH
O
HO
OH
O
OH
OH
CH2Cl2, RT
O
O
O
O
O
O
O
O
OH
X=O 5(x)
X=S 6(y)
OH
O
O
X=O 7(x)
X=S 8(y)
4
Scheme 1. Synthesis of urea and thiourea derivatives.
N
NH2
H
N
N
N
N
N
HO
O
N
N
HO
O
HO
O
HO
OH
O
OH
HO
OH
O
H2, PtO2
OH
HO
OH
O
OH
O
O
O
O
10h, refluks
acetic acid, 5 bar
O
O
O
O
O
O
O
O
OH
OH
O
OH
O
O
9
10
4
Scheme 2. Synthesis of 9a-N-(
c
-aminopropyl)-9a-aza-homoerythromycin A 4.
Table 1
the aliphatic ones in 24 h (Table 1). Therefore, two separate li-
braries had to be prepared: one where aromatic ureas were synthe-
sized and the second where aliphatic urea and thiourea derivatives
were formed.
Reaction conditions for urea- and thiourea libraries
X
R
Equivalents
Reaction time (h)
O
O
S
Aromatic
Aliphatic
Aromatic
Aliphatic
1.3
1.3
1.3
1.3
0.5–1.5
12–24
12
2.1.2. Building block rehearsal studies
S
12–24
Prior to committing to a full combinatorial library synthesis, it
is advantageous to ‘rehearse’ the building blocks chosen in order
to determine if they are compatible with and have sufficient reac-
tivity’s under the generic reaction conditions adopted. In this way,
the potential for failure, or the isolation of large numbers of com-
12–24 h. Such a big difference was not observed in the isothiocya-
nate series where the aromatic reagents reacted within 12 h and

A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
1695
threshold, but with >60% LC/MS purities, were further purified by
single-pass autopreparative RP-HPLC to afford additional samples
for biological screening. The yield of both libraries was 69–95%
overall, except for the compounds which had to be purified by
RP-HPLC. Importantly, the potential for the loss of material through
inefficient resin washing can be significantly reduced by the incor-
poration of thorough vortex mixing in the resin wash step. Even
with these yields, however, we were still able to synthesize suffi-
cient material for use in several biological assays.
Overall, full library analysis was accomplished with a randomly
selected sample of 38 of the 105 compounds. This sample group
was evaluated using 1H NMR, 13C NMR, LC/MS, and HRMS. Product
purity was determined by LC/MS screening (ESI+ detection)
2.2. Biological activity and SAR observations
The in vitro potency of 98 ureas and thioureas of 15-membered
azalides was evaluated against three P. falciparum strains:
TM91C235-multidrug resistant, D6-chloroquine sensitive and
W2-chloroquine/pyremethamine resistant and compared to azith-
romycin in order to select compounds suitable for further testing.
Parasites were exposed for 72 h. Data from in vitro potency tests
of azithromycin against the three parasite strains were averaged,
and these numbers were used to identify ‘active’ macrolide com-
pounds. These average IC50s against the TM91C235, D6, and W2
P. falciparum strains were 2065 nM, 1015 nM, and 2241 nM,
respectively. Only compounds of equal or better potency than
azithromycin against at least one of the parasites were selected
for further consideration. Active compounds were sorted by IC50
against the TM91C235 clone (Table 2). Almost all 9a-N substituted
15-member azalide (91/98) showed better in vitro activity than
azithromycin. In this assay, several compounds from the libraries
were found to be very active against both resistant strains of the
P. falciparum, with a notable decrease in parasitic growth. The most
interesting compounds from this library, particularly in inhibiting
the growth of all three strains including those that are highly drug
resistant, were 8{38}, 8{37}, and 8{25}.
The SAR observations commented here are primarily related to
the TM91C235 results, but they are also in line with the observa-
tions for the other two strains. The influence of various 9a-N sub-
stituents were investigated using these synthesized compounds.
Paired analogues differ only in the urea/thiourea moiety of the
9a-N substituent, and are presented in Figure 6. It seems that the
thiourea moiety increases in vitro antimalarial activity in compar-
ison to the urea moiety in most cases, with the exception of thio-
urea derivative that showed lower activity than the
corresponding urea analogue [pair 11: 8{55} < 7{35}].
Generally, the in vitro activities of the compounds increase by
increasing the rigidity of the 9a-N substituent making it energeti-
cally more favorable for possible intermolecular interaction. A
simple descriptor—the ratio of rotatable bonds in 9a-N substitu-
ent—was used to describe the rigidity of 9a-N substituent. In vitro
antimalarial activity increased as the ratio of rotatable bonds in the
9a-N substituent dropped (Fig. 7a). The average ratio dropped from
0.5 in inactive compounds (pIC50 6 6, N = 10) to 0.37 for highly ac-
tive compounds (pIC50 >7; N = 22). The ratio of rotatable bonds in
9a-N substituent was lowered mostly; (a) by increasing the num-
ber of aromatic atoms, and (b) by avoiding additional substitutions.
Lipophilic 9a-N substituents (Fig. 7b) increased the activity of non-
aromatic derivatives (e.g. in Fig. 8). However, these improvements
were still significantly lower in comparison to the highly active
aromatic 9a-N substituents.
Figure 4. (a) Building block set 5(x) rehearsal targeting product 7(x). (b) Building
block set 6(y) rehearsal targeting product 8(y).
pounds with low purities is reduced. The synthesis process of the
urea and thiourea derivatives is described with the corresponding
flowchart in Figure 3.
Using the previously determined conditions, the reaction be-
tween the amine 4 and the isocyanate and isothiocyanate building
blocks 5{1–6} and 6{1–6} was performed in a 48-position Mini-
block on the Mettler-Toledo MiniBlock synthesizer. A PS-trisamine
scavenge was then performed to remove the slight excess of 5{1–
6} and 6{1–6}, respectively that was used. In CH2Cl2 as the solvent,
no insolubility difficulties were encountered and all urea deriva-
tives 7{1–6}, and the corresponding thiourea derivatives 8{1–6}
were isolated in 90–96% purity (LC/MS) (Fig. 4a and b). Although
it is desirable for all products to have purities of >85%, a limited
number of products with lower purities (60–85%), that may subse-
quently be purified by reverse-phase HPLC (RP-HPLC), are
acceptable.
2.1.3. Preparation of a 33-member urea library 7{4–36}, and a
69-member urea and thiourea library 7{37–49} and 8{1–56}
The libraries were prepared using the reaction conditions devel-
oped for the synthesis of the trial compounds 7{1–6} and 8{1–6}.
The process is outlined for the 33-member urea library derived
from building block 4 and a set of aromatic isocyanate building
blocks 5{4–36} in Charts 1 and 2 and is the same as for the 69-
member urea and thiourea library derived from building blocks 4
and 5{37–49} and 6{1–56} in Charts 1–3. Details of the programs
used are given in Section 4.
Aliquots of the final product solutions were transferred and di-
luted to facilitate off-line LC/MS analysis of the library products.
For the 33-member urea library 7{4–36}, as shown in Figure 5a,
23 of the 33 library members (69.7%) were obtained in >85% purity
(LC/MS) with an average library purity of 81.3%. For the 69-mem-
ber urea and thiourea library 7{37–49} and 8{1–56}, 61 of the 69
desired compounds (88.4%) were obtained in >85% purity with
an average library purity of 86.7%, according to LC/MS ( Fig. 5a
and b). In both cases, compounds falling below the 86% purity
Generally, an aromatic moiety in the 9a-N substituent improved
the activity (Fig. 7c) compared to azithromycin—only one com-
pound 5(29) showed lower activity. Consequently, aromatic 9a-N

1696
A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
O
N
O
N
Cl
F
O
N
O
N
O
N
S
N
5(7)
5(8)
5(9)
5(10)
5(11)
O
N
O
N
F
O
N
O
N
O
N
N
F
F
F
O
F
F
5(12)
5(13)
5(14)
5(15)
5(16)
O
N
O
N
N
O
N
O
N
O
O
F
O
Cl
O
O
F
N
5(17)
5(18)
5(19)
5(20)
5(21)
F
O
N
O
O
N
O
O
O
O
O
N
O
N
O
O
O
O
5(22)
5(23)
5(24)
5(25)
5(26)
O
N
O
O
O
N
O
N
O
N
N
O
N
O
O
N
O
CF3
N
O
5(27)
5(28)
5(29)
5(30)
5(31)
O
N
S
Cl
S
N
O
N
N
O
N
S
O
N
O
N
N
Cl
O
5(32)
5(33)
5(34)
5(35)
5(36)
O
N
N
Cl
O
N
O
N
O
N
O
N
Cl
Cl
Cl
5(37)
5(38)
5(39)
5(40)
5(41)
O
N
O
O
N
O
N
O
N
Cl
O
5(42)
5(43)
5(44)
5(45)
5(46)
O
N
O
N
S
N
O
5(47)
5(48)
5(49)
O
N
O
Chart 2. Set of isocyanate 5(x) Building Blocks.
derivatives are preferred for further optimization in comparison to
nonaromatic derivatives.
erythromycin-9a-N-(c-aminopropyl) thiourea libraries. These
methods have been exemplified by the preparation of a 33-mem-
ber urea library 7{4–36} and a 69-member urea and thiourea li-
brary 7{37–49} and 8{1–56}, using a Mettler-Toledo MiniBlock
synthesizer. Building block rehearsals were performed in both
cases in order to confirm that the monomers selected were likely
to provide library compounds in acceptable yields and purities.
In this way, P48% of all library members were obtained with
3. Conclusion
In conclusion, we have described an automated polymer-
assisted solution phase protocol for the synthesis of 9a-aza-
homoerythromycin-9a-N-(c-aminopropyl) urea and 9a-aza-homo-

A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
1697
S
N
S
N
S
S
N
S
N
S
N
Cl
6(7)
6(8)
6(9)
6(10)
6(11)
S
S
N
S
N
S
N
S
S
N
S
S
N
F
N
F
F
6(12)
6(13)
N
6(14)
6(15)
6(16)
N
S
N
N
S
N
O
NO2
F
F
F
6(17)
6(18)
6(19)
6(20)
6(21)
S
N
S
N
O
S
N
S
N
O
F
S
N
N
Br
O
F
NO2
F
F
6(22)
6(23)
6(24)
6(25)
6(26)
S
N
S
N
S
N
N
O
O
S
S
N
N
O
O
6(27)
6(28)
6(29)
6(30)
6(34)
N
NO2
S
N
S
O
N
N
N
S
N
N
6(31)
6(32)
6(33)
Chart 3. Set of isothiocyanate 6(y) building blocks.
P90% LC/MS purities and these compounds required no additional
purification prior to primary biological evaluation. The average
purity of the raw products was 84.1%. In vitro potency data against
three P. falciparum strains shows that almost all of the new ana-
logues have greater potency than azithromycin, and the top 10
compounds are 30–65-fold more potent than azithromycin.
The present study confirmed the newly synthesized 9a-
evaporation system. Analytical high-pressure liquid chromatogra-
phy (HPLC) was carried out on a Waters FractionLynx LC/MS sys-
tem using a Micromass ZQ mass spectrometer and positive
electrospray ionization with a Waters Symmetry C8 4.6 mm Â
75 mm, 3 lm; column. Preparative high-pressure liquid chroma-
tography (HPLC) was carried out on a Waters FractionLynx LC/MS
system using a Micromass ZQ mass spectrometer and positive elec-
trospray ionization with a Waters Xterra RP18 19 mm  100 mm,
aza-homoerythromycin-9a-N-(c-aminopropyl) ureas and 9a-aza-
homoerythromycin-9a-N-( -aminopropyl) thioureas as new leads
c
5 lm; column.
in antimalarial chemotherapy. Further optimizations regarding the
functionality on the 9a-position as well as the in vivo profiling of
the most interesting compounds are being envisaged as next steps.
The outcome of these investigations will be shown in an upcoming
publication.
HRMS data were acquired using a Micromass Q-Tof 2 hybrid
quadrupole time-of-flight mass spectrometer, equipped with a Z-
spray interface, over a mass range of 100–1100 Da, with a scan time
of 0.9 s and an interscan delay of 0.1 s. Retention times (tR) are re-
ported in terms of minutes. NMR spectra were recorded in DMSO-
d6, using either a Bruker Avance III 600 (600 MHz) or DRX400 and
an internal TMS standard. 1H NMR spectra were recorded at
600 MHz or 400 MHz, 13C NMR spectra were recorded at
101 MHz. All shifts are reported as part per million relative to TMS.
4. Experimental section
4.1. General information
4.2. Generalautomated procedure for the 48-membered library.
Preparation of 9a-aza-homoerythromycin-9a-N-(c-
aminopropyl) urea and thiourea derivatives 7{4–36}, 7{37–49},
and 8{1–56}
All solvents and reagents were used as supplied, unless noted
otherwise. Isocyanates and isothiocyanates used in this work are
commercially available chemicals and were obtained from Aldrich,
Alfa Aesar, Lancaster and Maybridge. Polymer-supported reagents
and scavenger resins were obtained from Argonaut Technologies.
Reaction preparation was carried out on a Mettler Toledo
MiniMapper Sample Processing Station and the automated synthe-
sis was performed using a Mettler Toledo MiniBlock system. Evap-
Step 1. Weigh the building blocks and the amine needed for the
reaction and calculate the solvent volumes to get the
desired concentrations (0.1 mmol/ml—building blocks;
0.05 mmol/ml—amine).
oration of solvents was conducted using
a GeneVac HT-4X

1698
A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
S
N
F
S
N
F
F
N
S
N
S
N
O
Cl
6(35)
6(36)
6(37)
6(38)
S
N
O
S
N
O
S
N
S
N
S
N
Cl
Cl
Cl
6(39)
6(40)
6(45)
6(41)
6(42)
6(43)
S
N
S
N
S
S
N
N
S
N
O
O
N
N
N
N
S
N
S
O
O
O
O
6(44)
6(46)
6(47)
6(48)
O
S
N
S
N
S
N
S
N
S
O
N
N
O
S
N
N
O
O
6(49)
6(50)
6(51)
6(52)
6(53)
S
N
N
S
N
S
N
S
N
6(54)
6(55)
6(56)
Chart 3. (continued)
Step 2. Put the vials with the building blocks and amine on the
MiniMapper.
4.3. General procedure for the parallel synthesis of urea and
thioureas
Step 3. Paste calculated volumes into the MiniMapper.
Step 4. Add the solvent (CH2Cl2) to the vials containing the
building blocks.
Step 5. Add the solvent (CH2Cl2) to the 48-position reaction
tubes of the MiniBlock synthesis module.
Step 6. Add the solvent (CH2Cl2) to the vials containing the
amine.
Step 7. Transfer amine from the vials to the 48-position reaction
tubes of the MiniBlock synthesis module.
To a solution of 9a-N-(c-aminopropyl)-9a-aza-homoerythromy-
cin A 10 (30 mg, 0.0378 mmol) in CH2Cl2 (1 ml), isocyanate or isothi-
ocyanate (1.3 equiv) was added and stirred at room temperature for
2–48 h. At the end of the reaction PS-trisamine (5 equiv) was added
to the reaction mixture. After stirring for 12 h at room temperature
the resin was filtered off, and washed with CH2Cl2 (3 Â 1 ml). The
organic solvent was evaporated to yield the desired product.
Step 8. Transfer building blocks from the vials to the 48-position
reaction tubes of the MiniBlock synthesis module.
Step 9. Remove reaction block from MiniMapper and place it on
the MiniBlock synthesizer module and stir reactions vig-
orously at ambient temperature for 48 h.
Step 10. Add 5 equiv of PS-trisamine to every reaction tube and
stir reaction at ambient temperature for 12 h.
Step 11. Drain resins and wash with CH2Cl2 (2 Â 0.5 mL) into new
vials.
4.3.1. 9a-N-(b-Cyanoethyl)-9a-aza-homoerythromycin A (10)
9-Deoxo-9-dihydro-9a-aza-9a-homoerythromycin A 9 (100 g;
136 mmol) was dissolved in acrylonitrile (300 ml) and the reaction
mixture was heated at 80 °C for 24 h. The solvent was evaporated
under reduced pressure to afford the desired product 10 (106 g) in
99% yield as a white foam. This compound was used for the next
step without previous purification.
LC/MS (area %): 95%, MS(ES) m/z: [MH]+ 788.2 (calcd: 788.02).
Step 12. Prepare quality control (QC); dispense aliquots (5
into 48 vials and dilute each vial with acetonitrile
(100 L) and water (900 L).
Step 13. Remove solvent from bulk samples by parallel centrifu-
gal evaporation in vacuum to afford the urea and thio-
urea derivatives.
l
L)
4.3.2. 9a-N-(c-Aminopropyl)-9a-aza-homoerythromycin A (4)
To a solution of 9a-N-(b-cyanoethyl)-9a-aza-homoerythromy-
cin A 10 (5 g; 6.3 mmol) in glacial acetic acid (80 ml), PtO2 (0.5 g)
was added and the reaction mixture hydrogenated at a pressure
of 5 bar for 12 h. The catalyst was filtered off. Water was added
to the filtrate. The aqueous solution was washed with CH2Cl2
(3 Â 30 ml). The pH of the aqueous phase was adjusted to 6, 8,
10, 12, and 14. At each pH the aqueous layer was extracted with
CH2Cl2 (3 Â 30 ml). The organic layers at pH 10 and 12 were com-
l
l
Note: Compounds with crude purities <86% were purified by RP-
HPLC.

A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
1699
Figure 5. (a) Building block set 5(x) library targeting product 7(x). (b) Building block set 6(y) library targeting product 8(y).
bined, dried over Na2SO4 and the solvent evaporated under re-
duced pressure to obtain the desired product 4 (2.9 g) in 58% yield
as a white foam.
cides). Two Plasmodium falciparum clones from CDC/Indochina III
(W-2) and CDC/Sierra Leone I (D-6) were used for all assays.
TM91C235, a multiple drug resistant isolate from Thailand, was
used for the prescreening assay. W-2 is resistant to chloroquine,
quinine, and pyrimethamine and susceptible to mefloquine. D-6
tends to be more resistant to mefloquine and susceptible to
chloroquine, quinine, and pyrimethamine. The isolate was preex-
posed, in duplicate, at three concentrations (25,000 ng/ml,
2500 ng/ml, and 25 ng/ml) of the test compound for 72 h in a 96-
well microtiter plate (MTP) using the BIOMEKÒ 2000 automated
laboratory workstation. The assay relies on the incorporation of
radiolabeled hypoxanthine by the parasites, and inhibition of iso-
tope incorporation is attributed to activity of known or candidate
antimalarial drugs. After 96 h of total incubation time, the MTP
were frozen to lyse the erythrocytes and parasites. The parasite
DNA was recovered by harvesting the lysate onto glass-fiber filter
plates using a Packard FilterMate Cell Harvester. The radioactivity
was counted on a Packard TopCount microplate scintillation coun-
ter. The results were recorded as counts per minutes (CPM) per
well at each drug concentration divided by the arithmetic mean
of the CPM from the three untreated infection parasite control
wells.
1H NMR (600 MHz, DMSO-d6) d ppm: 0.77 (t, J = 7.33 Hz, 3H),
0.87 (d, J = 6.45 Hz, 3H), 0.94–0.98 (m, 6H), 0.99 (d, J = 7.33 Hz,
3H), 1.03–1.07 (m, 4H), 1.10 (d, J = 7.15 Hz, 3H), 1.12 (s, 3H), 1.14
(d, J = 6.11 Hz, 3H), 1.19 (s, 3H), 1.24–1.31 (m, 1H), 1.32–1.42 (m,
2H), 1.47–1.59 (m, 4H), 1.73–1.81 (m, 1H), 1.84–1.89 (m, 1H),
1.90–1.95 (m, 1H), 2.10 (t, J = 10.38 Hz, 1H), 2.21 (s, 6H), 2.27 (d,
J = 15.00 Hz, 1H), 2.30–2.36 (m, 1H), 2.38–2.44 (m, 1H), 2.47–
2.55 (m, 2H), 2.63–2.76 (m, 3H), 2.90 (d, J = 9.07 Hz, 1H), 2.95–
3.00 (m, 1H), 3.01–3.05 (m, 1H), 3.21 (s, 3H), 3.48 (s, 1H), 3.50
(d, J = 6.63 Hz, 1H), 3.64 (dd, J = 10.03, 5.50 Hz, 1H), 3.99 (d,
J = 5.93 Hz, 1H), 4.03–4.08 (m, 1H), 4.23 (br s, 1H), 4.40 (d,
J = 7.15 Hz, 1H), 4.80 (d, J = 4.54 Hz, 1H), 4.90 (d, J = 10.47 Hz, 1H).
13C NMR (101 MHz, DMSO-d6) d ppm: 7.24, 8.55, 9.97, 14.42,
16.83, 17.61, 20.01, 20.18, 20.55, 21.56, 26.77, 27.56, 29.38,
29.41, 34.00, 38.20, 39.56, 39.45, 39.50, 43.49, 47.90, 58.24,
63.07, 63.56, 63.97, 66.18, 69.82, 71.80, 72.67, 73.36, 74.13,
75.82, 76.48, 77.49, 82.71, 94.39, 101.16, 175.19.
LC/MS (area %): 89%. MS (ES) m/z: [MH]+ 792.3 (calcd: 792.05)
4.4. In vitro testing
4.5. SAR analysis
The compounds were tested for their intrinsic antimalarial
activity. The system is limited to the assessment of the intrinsic
activity against the erythrocytic asexual lifecycle (blood schizonto-
SAR analysis was performed using D-Score and Find Lookalike
procedures implemented as a part of the SAR Toolkit,24 a GSK

1700
A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
Table 2
Table 2 (continued)
In vitro activity of tested compounds against three P. falciparum strains with different
IC50 (nM)
Compound
susceptibility patterns
Multidrug
resistant
TM91C235
Chloroquine
sensitive
D6
Chloroquine/pyrimethamine
resistant
W2
IC50 (nM)
Compound
Multidrug
resistant
TM91C235
Chloroquine
Chloroquine/pyrimethamine
sensitive
D6
resistant
W2
398
1730
931.7
824.7
647.6
708.9
1043.1
229.3
119.8
184.7
1298.1
495.1
255.8
203.8
734.3
287
276.1
720.6
148.7
277.3
410.5
1198
370.2
1585.3
679.1
488.6
1957.6
13748.7
3297.9
2090.7
1273.1
3132.5
12804.4
12837.4
7{17}
7{47}
8{53}
7{22}
7{2}
405.1
407.1
410.8
411
42.4
2065
31.9
42.9
51.3
54
55
55.3
55.4
56.3
60
64.7
68
69.7
72.8
73.7
78.2
78.3
84.2
84.2
90.4
96.2
96.3
99.3
101.2
101.4
101.8
106.7
107.5
115.6
116.9
117.7
121.1
122.5
126.5
127.3
132.7
137.1
137.1
138.1
138.6
138.8
140.1
142.4
146
152.2
159.5
163.8
164.1
172.9
173.3
179.9
182.9
183
188.4
206.2
209.4
216.5
219.8
232.3
236.2
239.9
251.7
265.4
288.2
298.3
320.4
330.4
353.8
381.5
5.6
166.6
2241
37
48.3
37.1
30.3
37.7
113.5
42.5
35
40.8
69.2
53
50.1
134.9
164.7
50
70.9
74.6
229.4
99.3
73.6
47.9
73.8
109.4
88.2
46.5
70.9
53.4
199.8
53.2
187.2
63
399.6
273.8
89.6
109.2
73.9
202.1
201.4
72.8
88.7
85
Chloroquine
Azithromycin
8{38}
8{32}
8{21}
8{5}
8{39}
8{41}
8{4}
8{37}
7{25}
7{35}
8{46}
8{25}
7{13}
8{22}
8{14}
7{16}
7{14}
8{42}
7{20}
7{4}
8{31}
7{12}
8{45}
8{55}
8{12}
8{27}
8{43}
8{15}
7{5}
1015
124.8
53.4
102.8
119.8
218.3
58.1
115.7
54.3
28.1
1961.2
2605.1
688.9
1965.4
1620.2
936.2
5614.4
1141.1
964
1788.3
6022.4
5505.6
1211
5802.2
2051.4
3721.9
14025.9
2842.1
13748.7
9286.4
12372
7008.5
794.2
428.5
434.6
441.5
483.9
491.7
495.5
514.5
548.7
570.9
615.6
799.7
801
820.2
923.3
971.6
1175.5
1569.1
1980.7
3028.9
3319.1
5091.1
12371.4
12804.4
12837.4
8{8}
8{54}
7{15}
7{46}
7{31}
7{3}
8{13}
7{33}
7{21}
7{38}
7{42}
8{49}
7{1}
266.5
174.5
74.5
256.6
449.2
132.2
224.6
159.9
296.2
188.1
265.5
107.4
226.1
160
175.8
248.9
121.3
187.7
511.2
396
454.9
216.4
986.7
644.8
299.6
215.3
332.4
253.6
445.2
319.8
330.1
264.5
609.5
298.6
552.8
249.3
579.7
330
329.4
260.8
326
302.1
268.6
508.7
386.2
480.4
319
283.9
469.8
635.1
329.4
444.9
835.2
297.5
765.3
360.6
452.5
897.8
655.2
7{24}
8{2}
7{44}
8{50}
7{37}
8{35}
8{28}
8{33}
8{24}
7{29}
8{9}
12804.4
12837.4
7{9}
8{23}
8{16}
8{40}
7{19}
8{51}
7{7}
7{26}
7{27}
8{17}
7{6}
8{29}
7{49}
7{36}
8{10}
8{36}
7{10}
8{6}
8{44}
8{26}
8{47}
8{56}
7{45}
8{20}
7{8}
8{19}
8{30}
7{23}
7{11}
7{48}
8{48}
8{18}
7{18}
8{3}
8{34}
7{28}
8{52}
8{11}
8{1}
113.1
175
210
88.9
285
Figure 6. In vitro antimalarial activity against Plasmodium falciparum TM91C235
strain of paired analogues of 9a-N substituted 15-membered azalides differing only
in urea/thiourea moiety.
71.3
182.5
168.9
174.9
172.6
78.4
86.3
98.3
77
136.6
306.1
178.6
276.2
448.9
96.2
110.9
107
proprietary suite of applications for SAR data analysis, based on
Daylight Toolkit25 and embedded within Spotfire DecisionSite.26
D-Score is a method for the identification of molecular pairs differ-
ing by a single R-group change upon fragmentation of the analyzed
molecules. D-Score was calculated simply by averaging the differ-
ence in activity between two paired molecules over the number of
pairs. Pairs of molecules which differ by a single R-group change
were extracted and visualized using profile plots; it can be useful
to identify basic SAR trends as well as outliers, to investigate addi-
tivity in the data, and potentially predict novel combinations. Find
Lookalike is a method for the selection and visualization of mole-
cules that differ by a single localized change for the investigation
of basic SAR trends. This structural change may involve one, two
or up to a user-defined number of atoms as long as they are con-
nected. Any change in activity/property is clearly related to the sin-
gle change in structure.
683.9
309.5
374.7
202.3
187.8

A. Hutinec et al. / Bioorg. Med. Chem. 19 (2011) 1692–1701
1701
References and notes
1. (a) Breman, J. G. Am. J. Trop. Med. Hyg. 2001, 64, 1; (b) Guerin, P. J.; Olliaro, P.;
Nosten, F.; Druilhe, P.; Laxminarayan, R.; Binka, F.; Kilama, W. L.; Ford, N.;
White, N. J. Lancet Infect. Dis. 2002, 2, 564.
2. Snow, R. W.; Guerra, G. A.; Noor, A. M.; Myint, H. Y.; Hay, S. I. Nature 2005, 434,
214; Hay, S. I.; Guerra, G. A.; Tatem, A. J.; Noor, A. M.; Snow, R. W. Lancet Infect.
Dis. 2004, 4, 327.
3. (a) Africa malaria report. WHO/33 25 April 2003.; (b) Mount, A. M.; Mwapasa,
V.; Elliott, S. R.; Beeson, J. G.; Tadesse, E.; Lema, V. M.; Molyneux, M. E.;
Meshnick, S. R.; Rogerson, S. J. Lancet 2004, 363, 1860.
4. Haldar, K.; Milner, D. A.; Murphy, S. C., Jr.; Taylor, T. E. Annu. Rev. Pathol.: Mech.
Dis. 2007, 2, 217.
5. Frederich, M.; Dogne, J.-M.; Angenot, L.; Mol, P. D. Curr. Med. Chem. 2002, 9, 1435.
6. White, N. J. Br. Med. Bull. 1998, 54, 703.
7. (a) White, N. J. J. Clin. Invest. 2004, 113, 1084; (b) Biagini, G. A.; O’Neill, P. M.;
Nzila, A.; Ward, S. A.; Bray, P. G. Trends Parasitol. 2003, 19, 479.
8. (a) Klayman, D. L. Science 1985, 228, 1049; (b) Luo, X. D.; Shen, C. C. Med. Res.
Rev. 1987, 7, 29; (c) Zaman, S. S.; Sharma, R. P. Heterocycles 1991, 32, 1593; (d)
Cumming, J. N.; Ploypradith, P.; Posner, G. H. Adv. Pharmacol. 1997, 37, 253; (e)
Zhou, W. S.; Xu, X. X. Acc. Chem. Res. 1994, 27, 211; (f) Bhattacharya, A. K.;
Sharma, R. P. Heterocycles 1999, 51, 1681.
9. Murray, M. C.; Perkins, M. E. Ann. Rep. Med. Chem. 1996, 31, 141.
10. Noedl, H.; Wernsdorfer, W. H.; Krudsood, S.; Wilairatana, P.; Kollaritsch, H.;
Wiedermann, G.; Looareesuwan, S. Acta Trop. 2001, 80, 39.
11. Ohrt, C.; Willingmyre, G. D.; Lee, P.; Knirsch, C.; Milhous, W. Antimicrob. Agents
Chemother. 2002, 46, 2518.
12. Andersen, S. L.; Oloo, A. J.; Gordon, D. M.; Ragama, O. B.; Aleman, G. M.;
Berman, J. D.; Tang, D. B.; Dunne, M. W.; Shanks, G. D. Clin. Infect. Dis. 1998, 26,
146.
13. Dunne, M. W.; Singh, N.; Shukla, M.; Valecha, N.; Bhattacharyya, P. C.; Dev, V.;
Patel, K.; Mohapatra, M. K.; Lakhani, J.; Benner, R.; Lele, C.; Patki, K. J. Infect. Dis.
2005, 191, 1582.
14. Dunne, M. W.; Singh, N.; Shukla, M.; Valecha, N.; Bhattacharyya, P. C.; Patel, K.;
Mohapatra, M. K.; Lakhani, J.; Devi, C. U.; Adak, T.; Dev, V.; Yadav, R. S.; Lele, C.;
Patki, K. Am. J. Trop. Med. Hyg. 2005, 73, 1108.
15. Heppner, D. G., Jr.; Walsh, D. S.; Uthaimongkol, N.; Tang, D. B.; Tulyayon, S.;
Permpanich, B.; Wimonwattrawatee, T.; Chuanak, N.; Laoboonchai, A.; Sookto,
P.; Brewer, T. G.; McDaniel, P.; Eamsila, C.; Yongvanitchit, K.; Uhl, K.; Kyle, D. E.;
Keep, L. W.; Miller, R. E.; Wongsrichanalai, C. Am. J. Trop. Med. Hyg. 2005, 73, 842.
16. Miller, R. S.; Wongsrichanalai, C.; Buathong, N.; McDaniel, P.; Walsh, D. S.;
Knirsch, C.; Ohrt, C. Am. J. Trop. Med. Hyg. 2006, 74, 401.
17. Noedl, H.; Krudsood, S.; Chalermratana, K.; Silachamroon, U.; Leowattana, W.;
Tangpukdee, N.; Looareesuwan, S.; Miller, R. S.; Fukuda, M.; Jongsakul, K.;
Sriwichai, S.; Rowan, J.; Bhattacharyya, H.; Ohrt, C.; Knirsch, C. Clin. Infect. Dis.
2006, 43, 1264.
Figure 7. In vitro antimalarial activity against TM91C235 Plasmodium falciparum
strain for 15-member azalides increases with (a) decrease in the ratio of rotatable
bonds in 9a-N substituents, (b) increase in lipophilicity of 9a-N substituents; and (c)
with increase in the number of aromatic atoms in 9a-N substituent.
R
X
IC50 / nM
N
N
R
O
183.0
X
N
*
*
18. Taylor, W. R.; Richie, T. L.; Fryauff, D. J.; Picarima, H.; Ohrt, C.; Tang, D.;
Braitman, D.; Murphy, G. S.; Widjaja, H.; Tjitra, E.; Ganjar, A.; Jones, T. R.; Basri,
H.; Berman, J. Clin. Infect. Dis. 1999, 28, 74.
19. Gray, R. H.; Wabwire-Mangen, F.; Kigozi, G.; Sewankambo, N. K.; Serwadda, D.;
Moulton, L. H.; Quinn, T. C.; O’Brien, K. L.; Meehan, M.; Abramowsky, C.; Robb,
M.; Wawer, M. J. Am. J. Obstet. Gynecol. 2001, 185, 1209.
20. (a) Yoshida, J.; Itami, K. Chem. Rev. 2002, 102, 3693; (b) Ganesan, A. Drug
Discovery Today 2002, 7, 47; (c) Doerwald, F. Z. Organic Synthesis on Solid Phase;
Wiley-VCH: Weinheim, Germany, 2000; (d) Obrecht, D.; Villalgordo, J. M. Solid-
Supported Combinatorial and Parallel Synthesis of Small-Molecular-Weight
Compound Libraries; Pergamon Press: Oxford, U.K., 1998.
21. Gooding, O.; Hoeprich, P. D., Jr.; Labadie, J. W.; Porco, J. A., Jr.; van Eikeren, P.;
Wright, P. Mol. Divers. 1996, 2, 199.
22. For example: (a) Li, X.; Abell, C.; Congreve, M.; Warrington, B. H.; Ladlow, M.
Org. Biomol. Chem. 2004, 2, 989; (b) Villard, A.-L.; Warrington, B. H.; Ladlow, M.
J. Comb. Chem. 2004, 6, 611.
OH
O
HO
OH
O
S
182.9
152.2
O
N
*
OH
O
O
O
O
*
S
S
142.4
101.8
O
OH
*
Figure 8. Nonaromatic lipophilic 9a-N substituents increased in vitro antimalarial
activity against TM91C235 Plasmodium falciparum strain in 15-member azalides.
ˇ ´
ˇ ´
´
´
23. Bukvic´ Krajacic, M.; Kujundzic, N.; Dumic, M.; Cindric, M.; Brajša, K.; Metelko,
B.; Novak, P. J. Antibiot. 2005, 58, 380.
24. Atkinson, F.; Bravi, G. The SAR Toolkit. Abstract (http://www.iainm.demon.
co.uk/ukqsar/meetings/2007-10-11.html#abstract_e013ca26c3a3c590184b907a
3d2668c2) and presentation (http://www.documentarea.com/qsar/Francis_
Atkinson07.pdf). UK-QSAR and ChemoInformatics Group meeting, 2007.
25. Daylight v4.92 Chemical Information Systems Inc. Aliso Viejo, CA, http://
www.daylight.com.
Acknowledgements
The authors gratefully acknowledge the assistance of Lucia
Gerena and her group for the in vitro screening, Zeljko Osman for
LC/MS analysis and Biserka Metelko for NMR spectroscopy.
ˇ
26. Spotfire Decision Site 8.2.1, http://www.spotfire.com/.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2011.01.030.