Antibacterial agent discovery using thymidylate synthase biolibrary screening

Article abstract of DOI:10.1021/jm051187d

Thymidylate synthase (TS, Thy A) catalyzes the reductive methylation of 2′-deoxyuridine 5′-monophosphate to 2′-deoxythymidine 5′-monophosphate, an essential precursor for DNA synthesis. A specific inhibition of this enzyme induces bacterial cell death. As

Full text of DOI:10.1021/jm051187d

5958
J. Med. Chem. 2006, 49, 5958-5968
Antibacterial Agent Discovery Using Thymidylate Synthase Biolibrary Screening
M. Paola Costi,*,† Arianna Gelain,‡ Daniela Barlocco,‡ Stefano Ghelli,§ Fabrizia Soragni,† Fabiano Reniero,# Tiziana Rossi,|
Antonio Ruberto,| Claude Guillou,| Antonio Cavazzuti,† Chiara Casolari, and Stefania Ferrari†
Dipartimento di Scienze Farmaceutiche, UniVersita` degli Studi di Modena e Reggio Emilia (UNIMORE), Via Campi 183, 41100 Modena, Italy,
Istituto di Chimica Farmaceutica e Tossicologica, UniVersita` degli Studi di Milano, Viale Abruzzi 42, 20131 Milano, Italy, Spinlab S.r.l., Via
Tamagno 3, 42048 Rubiera (RE), Italy, Physical and Chemical Exposure Unit, Institute for Health and Consumer Protection, European
Commission Joint Research Centre, 21020 Ispra (VA), Italy, Dipartimento di Scienze Biomediche, Sezione di Farmacologia, UniVersita` degli
Studi di Modena e Reggio Emilia, Via Campi 287, 41100 Modena, Italy, and Dipartimento di Medicina di Laboratorio, UniVersita` degli Studi
di Modena e Reggio Emilia, Via del Pozzo 71, 41100 Modena, Italy
ReceiVed NoVember 23, 2005
Thymidylate synthase (TS, ThyA) catalyzes the reductive methylation of 2′-deoxyuridine 5′-monophosphate
to 2′-deoxythymidine 5′-monophosphate, an essential precursor for DNA synthesis. A specific inhibition of
this enzyme induces bacterial cell death. As a second round lead optimization design, new 1,2-naphthalein
derivatives have been synthesized and tested against a TS-based biolibrary, including human thymidylate
synthase (hTS). Docking studies have been performed to rationalize the experimentally observed affinity
profiles of 1,2-naphthalein compounds toward Lactobacillus casei TS and hTS. The best TS inhibitors have
been tested against a number of clinical isolates of Gram-positive-resistant bacterial strains. Compound
3,3-bis(3,5-dibromo-4-hydroxyphenyl)-1H,3H-naphtho[1,2-c]furan-1-one (5) showed significant antibacterial
activity, no in vitro toxicity, and dose-response effects against Staphylococcus epidermidis (MIC ) 0.5-
2.5 µg/mL) clinical isolate strains, which are resistant to at least 17 of the best known antibacterial agents,
including vancomycin. So far this compound can be regarded as a leading antibacterial agent.
Introduction
Thymidylate synthase (TS, ThyA) (E.C. 2.1.1.45) is a key
enzyme in DNA synthesis. It catalyzes the reductive methylation
of 2′-deoxyuridine 5′-monophosphate (dUMP) to 2′-deoxythy-
midine 5′-monophosphate (dTMP), which, as a triphosphate, is
a substrate of DNA polymerase and is incorporated into the
DNA. The catalyzed reaction is assisted by N5,N10-methylene-
tetrahydrofolate (mTHF), which acts as both a monocarbon unit
donor and reducing agent1 (Figure 1). TS inhibition leads to
thymineless death, which is a complex mechanism affecting the
replication process. Therefore, TS has been regarded as a target
for anticancer drugs.2 The enzyme has also been implicated in
the protein synthesis regulation pathway3,4 and apoptotic
process,4 as well as in the development of Alzheimer’s disease.5
It has also been regarded as a potential target for antibacterial
and antifungal chemotherapy.
TS is inhibited by substrate analogues, such as 5-fluoro-2′-
deoxyuridine 5′-monophosphate (FdUMP), or cofactor ana-
logues, such as ZD1694 and pemetrexed (Alimta) (Chart 1).
Recently, nonclassical anti-folate inhibitors (NCAI) that are able
to specifically inhibit the microbial enzyme with respect to
human thymidylate synthase (hTS)6-14 have been reported,
among them 3,3-bis(3-chloro-4-hydroxyphenyl)-1H,3H-naphtho-
[1,8-c,d]pyran-1-one (R156, compound 10 in this paper) and
didansyltyrosine (DDT)7,9,12-14 (Chart 1). In particular, com-
pound 10 proved to specifically bind Lactobacillus casei TS
Figure 1. TS catalytic reaction (dRbP ) 2′-dexyribose-5′-monophos-
phate).
(LcTS) with submicromolar inhibition constant (Ki ) 0.7 µM)
and was 30-fold more specific with respect to hTS.14 This
compound also showed antimicrobial activity against Gram-
positive bacteria, with a minimum inhibitory concentration
(MIC) of 1.8 µg/mL.13
Detailed studies on protein-ligand interaction, such as X-ray
three-dimensional structure determination, site-directed mu-
tagenesis experiments, and computational studies,14,15 have
demonstrated that these molecules show a multiple binding
mode, raising the important issue of the reliability of a structure-
based drug design approach. In a recent study,15 it was proposed
that the basis for species specificity of the naphthalein inhibitors
lies in the different flexibility of the human enzyme compared
to bacterial ones. The bacterial enzymes, such as LcTS, can
accommodate the naphthalein derivative (10) in its binding site
close to the small domain, a specific sequence (70 residues)
that enlarges the active site domain.14,16 By contrast, in human
enzyme the corresponding binding site is partially inaccessible.
As a consequence, inhibitor 10 cannot bind to its preferred site
in hTS and the affinity for this enzyme decreases, thus
ameliorating the specificity index (Ki hTS/ Ki LcTS).15
* To whom correspondence should be addressed. Phone: 0039-059-205-
5134. Fax: 0039-059-205-5131. E-mail: costimp@unimore.it.
† Dipartimento di Scienze Farmaceutiche, Universita` degli Studi di
Modena e Reggio Emilia.
In an attempt to enhance the affinity and specificity of the
naphthalein series, new 1,2-naphthalein derivatives (1-8, Chart
2) were designed and synthesized along with two new 1,8-
naphthalein derivatives (16, 18, Chart 2). All the compounds
and eight previously synthesized 1,8-naphthalein derivatives (9-
15, 17)9,13,14 (Chart 2), selected among the closest analogues
‡ Universita` degli Studi di Milano.
§ Spinlab S.r.l.
# Institute for Health and Consumer Protection.
| Dipartimento di Scienze Biomediche, Universita` degli Studi di Modena
e Reggio Emilia.
Dipartimento di Medicina di Laboratorio, Universita` degli Studi di
Modena e Reggio Emilia.
10.1021/jm051187d CCC: $33.50 © 2006 American Chemical Society
Published on Web 09/07/2006

Antibacterial Agent DiscoVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5959
Chart 2
Chart 1
shows that these compounds could be allocated to the same
binding site without visible impediment to good fitting. On these
bases, halogen substituents have also been introduced into this
new series to possibly enhance their inhibitory activity and
specificity, as seen in the previously reported 1,8- and 2,3-
naphthalein derivatives.9,13,14
1,2-Naphthalic anhydride (I) was synthesized according to a
previously reported method.18 Its condensation with phenol
under Friedel-Crafts conditions gave the 3,3-bis(4-hydroxy-
phenyl)-1H,3H-naphtho[1,2-c]furan-1-one (1) as the main reac-
tion product, together with its isomer 2 at a very low concentra-
tion. The iodo (3) and bromo (4, 5) derivatives were obtained
from 1 by treatment with iodine, chloride, or bromine, respec-
that were most potent in preliminary assays, were tested against
a TS-based biolibrary containing the following enzymes:
Lactobacillus casei (LcTS), Escherichia coli TS (EcTS), Cryp-
tococcus neoformans TS (CnTS), human TS (hTS), and human
dihydrofolate reductase (hDHFR). To rationalize the observed
affinity profiles of this new series, a modeling study was
performed and the interactions of compounds 1 and 2 with LcTS
and hTS were explored using the AutoDock 3.0.5 program.
Moreover, several molecules with significant affinity (5, 8, 10,
13-16, 18) were tested against wild-type bacterial strains. In
addition, the most active compounds were screened against
clinical isolate resistant bacterial strains and their cytotoxicity
was evaluated against VERO cells.
Design and Synthesis
The design of the new compounds was based on the known
X-ray crystal structure of the binary complex LcTS-phenol-
phthalein (pth) (Figure 2).17 Since the 1,2-naphthalein derivatives
may exist in two isomeric forms, namely, the R-isomers (1, 3-6)
and the â-isomers (2, 7, 8), both structures were considered.
Assuming that 1,2-naphthaleins have the same binding mode
as pth, the benzene ring added to the phthalein core in the
R-isomer derivatives would be directed toward the folate binding
site and would interact with I81, E84, and V314, whereas in
the â-isomer derivatives the benzene ring is directed toward
the small domain region and interacts with E84, W85, F104,
and V316. Although we hypothesized that increasing the
aromatic bulk with an additional benzene ring could be critical
for pth-derived molecules, the matching of 1 and 2 with pth
Figure 2. Detail of the X-ray crystal structure of the binary complex
LcTS-pth.

5960 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Costi et al.
Scheme 1a
a (a) phenol, SnCl4, ∆; (b) ICl, CH3COOH, room temp; (c) Br2, CH2Cl2, room temp; (d) Br2, CH3COOH, room temp.
Scheme 2a
a (a) thymol, AlCl3, ∆; (b) o-chlorophenol, H2SO4, ∆; (c) o-fluorophenol, H2SO4.
tively (Scheme 1). Compounds 6-8 were prepared from the
1,2-naphthalic anhydride (I) by heating with the appropriate
substituted 2-phenol in the presence of aluminum trichloride
(Scheme 2). It should be noted that while in the case of 6 only
the R-isomer was obtained, the opposite was true for 7 and 8,
which were collected in the â-form.
apparent inhibition constant (Ki) values of the tested compounds
were determined, and the specificity (SI) and selectivity indices
(SSI) were calculated. The results are shown in Table 1 and
Figure 3.
In the 1,2-naphthalein series, all the compounds, with the
exception of 6, inhibited the pathogen TS. Their Ki values were
1.4-11 µM against LcTS, 0.3-8.5 µM against EcTS, and 0.4-
13 µM against CnTS. In particular, 2 and 8 were the most active
compounds against LcTS with a Ki of 1.4 µM. Compound 2
also showed the lowest Ki (0.3 µM) against EcTS, while 1 was
the most potent against CnTS (Ki ) 0.4 µM). 1,2-Naphthalein
compounds generally inhibited hDHFR at a higher level than
hTS (Ki in the range 3.9-110 µM vs >100 µM), with the sole
exception of 2 (Ki of 35 µM vs 0.4 µM ). Compound 7 was
equipotent against the two enzymes (Ki of 35 and 34 µM,
respectively).
Compounds 16 and 18 were synthesized according to a
previously reported method.13
Profiling Enzyme Inhibition
A TS-based biolibrary was designed to determine a biological
affinity profile and to provide information on the species
specificity and selectivity of the synthesized compounds. In
particular, the issue of specificity was addressed, including a
number of TS from pathogenic organisms in the biolibrary, such
as LcTS, EcTS, CnTS, in addition to hTS. Selectivity was
verified toward the hDHFR enzyme, a typical TS metabolically
related enzyme that works in competition with hTS in the
binding of many classic folate analogues, such as CB3717 (10-
propargyl-5,8-dideazafolic acid) and methotrexate (MTX). The
In the 1,8-naphthalein series, Ki values of 0.7-31 µM vs
LcTS, 0.4 µM to 11 mM vs EcTS, and 3.3-35 µM vs CnTS
were observed. The most active compounds were 10 and 11
against LcTS (Ki ) 0.7 µM), 13 against EcTS (Ki ) 0.4 µM),

Antibacterial Agent DiscoVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5961
Table 1. Inhibition Constants (Ki), Specificity Index (SI ) Ki hTS/Ki enzyme) and Selectivity Index (SSI ) Ki hDHFR/Ki hTS) for Compounds 1-18 against
LcTS, EcTS, CnTS, hTS, and HDHFR
Ki (µM)
SI
SSI
LcTS
EcTS
4.4
CnTS
hTS
hDHFR
LcTS vs hTS
EcTS vs hTS
CnTS vs hTS
hTS vs hDHFR
1
2
3
4
5
6
7
8
11
0.4
0.6
4.5
2.7
13
.132a
0.4
110
35
3.9
33
69
13
35
109
94
.12
0.3
.123
.20
.144
.30
1
.163
.147
.29
.330
1
.54
.49
.19
,1
88
,0.01
,0.2
,0.3
1.4
2.0
6.6
1.7b
NDc
3.9
1.4
31
0.3
1.5
0.9
8.5b
NDc
1.2
1.4
.245a
.132a
.245a
NDc
NDc
8.1
3.5
16
34
9
28
.94
4
1
.132a
.132a
30b
.94
.4
43
.171
3
.132
1
4
.306
.251
.136
1
.38
.8
1
.13
2
.25
0.4
1
.60
.63
.38
,1
,1
0.2
9
10
11
12
13
14
15
16
17
18
CB3717
0.7b
0.7b
2
0.6b
13
5.3
0.4
0.9
35
5.2
50
.9
1
.330
5
.0.0003
.60
.365
.223
1
9.6
3.3
5.2
13
.120a
5.7
28
3.4
15
9.5
13
42
12
5
1
.132a
4.7
,0.02
3
3
,0.05
,0.1
,0.05
6.8
0.9
0.8c
1.6b
1.8b
0.06b
.11 mMa
4.1c
4
3.2
4.1
6.4
6.5
.245a
.402a
.245a
0.03b
1.1b
1.1b
0.06b
a Calculated supposing 2% inhibition at the maximum solubility concentration (24 µM for 11; 27 µM for 1, 4, 8, 9, 13; 50 µM for 3, 5, 16, 18; 82 µM
for 17; 2.2 mM for 15). b Already reported.9,13-15 c Not detectable because of the insolubility of the compound (solubility less than 10 µM).
Figure 3. Grid of affinity profiling. Ki ranges are described by color code.
and 12 against CnTS (Ki ) 3.3 µM). In addition, better activity
toward hDHFR versus hTS (Ki ) 3.4-94 µM) was found in
this series. Only three compounds were more active against hTS
than hDHFR, namely, 12 (Ki of 28 vs 5.7 µM), 14 (Ki of 25 vs
4.7 µM), and 15 (Ki of 9.5 vs 3.2 µM).
By contrast, 1,8-naphthalein derivatives were species-specific
against LcTS whereas 1,2-naphthalein compounds were species-
specific against CnTS with respect to hTS. Finally, the latter
compounds were less active against hDHFR than 1,8-naphtha-
lein derivatives.
In both series, all the compounds were competitive inhibitors
of the enzymes with respect to the folate cofactor.
Modeling
When specificity was considered in both series, 12 out of 18
compounds were 1 order of magnitude more specific for LcTS
than hTS (SI ) 9-306). In particular, compounds 3, 5, 8, 11,
13, and 16-18 were at least 100-fold more active against LcTS
with respect to hTS (SI ) 94-306), 16 being the best
representative. The same 12 compounds were also more active
against EcTS than hTS (SI ) 9-365). Compounds 3, 4, 8, 13,
and 16-18 show a specificity index greater than or around 100,
the best being 17. Finally, 10 compounds (1, 3-5, 8, 11, 13,
and 16-18) were specific for CnTS vs hTS, with a specificity
index of 13-330, compound 1 having the highest value.
However, when selectivity between hTS and hDHFR was
considered, only compound 2 showed a significant index (SSI
) 88), all the other derivatives being either inactive toward hTS
or equipotent against the two enzymes.
It is important to note that any new compound to be
considered as a potential antibacterial should not be active
against hDHFR. Unfortunately, in our case the compounds with
specific affinity vs LcTS and EcTS (13, 16-18) with respect
to hTS also showed measurable activity against hDHFR.
Altogether, 1,2- and 1,8-naphthalein derivatives showed some
similarity in their enzyme inhibition profile. In particular, both
series were species-specific against EcTS with respect to hTS.
The three-dimensional models of compounds 1 and 2 were
built using InsightII and then docked in the active site of the
three-dimensional crystallographic structure of LcTS and hTS
in the presence of a dUMP molecule, using the AutoDock 3.0.5
program.
Thirty-seven distinct conformational clusters, out of 100 runs,
were obtained for compounds 1 and 2 docked against LcTS. In
the case of hTS, 26 and 24 distinct conformational clusters, out
of 50 runs, were obtained for compounds 1 and 2, respectively.
The difference in the calculated docked energy among these
clusters were around 2 kcal/mol in the case of LcTS and 3.5
and 2.4 kcal/mol in the case of hTS, for 1 and 2, respectively.
These compounds are characterized by a pseudo-3-fold sym-
metry through which they can change their orientations without
changing their patterns of interactions; therefore, even if the
program considered all these clusters as distinct, these binding
conformations can be grouped as multiple orientations for the
two main binding sites previously identified:14,15 the folate-
binding site and the crystallographic binding site of compound
10 (near the small domain). The three-dimensional crystal-
lographic structure of the complex LcTS-1014 showed that this
compound binds 5 Å away from the folate binding site in a
less conserved region of the enzyme and interacts with many

5962 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Costi et al.
Table 2. Docking Results for Compounds 1 and 2 to LcTS and HTS
Using AutoDock 3.0.5, Showing Percentage of Ligand Conformations
Predicted at the Specified Site (a) and Estimated Free Energy of Binding
(kcal/mol) of the Best-Scored Conformation at the Specified Site (b)
in the 3′ position of the ribose of dUMP and through its carbonyl
function with the side chain amide of N226. It also interacts
through van der Waals interactions with other seven residues,
including M311.
LcTS
small domain
hTS
small domain
A second binding site for compound 1, nearer the small
domain region, was also predicted in hTS. In this binding site,
compound 1 forms hydrogen bonds through its carbonyl function
with the imino group of R176′ and through its phenol hydroxyls
with the carboxyl function of D49, the hydroxyl of Y258, and
the imino group of R175′. It also forms van der Waals
interactions with other eight seridues, including T51 and V313.
Compound 2 binds to hTS in a similar manner. In the folate-
binding site, it forms hydrogen bonds with D218 and I108,
N226, and the hydroxyl group in the 3′ position of the ribose
of dUMP and interacts through van der Waals interactions with
V79, F80, W109, N112, L221, G222, F225, and Y258 through
its phenols and naphthalene rings. The main difference with
respect to compound 1 is that compound 2, through its terminal
naphthalene ring, does not interact with M311 but with V79
and F80.
folate
folate
compd
a
b
a
b
a
b
a
b
1
2
50 -3.14
58 -3.52
40
41
-1.99
-2.46
50 -3.31
62 -3.10
32
26
-1.88
-1.89
Compound 2 also binds in a second binding site of hTS, near
the small domain. It forms hydrogen bonds with R176′, D49,
Y258, and R175′ and van der Waals interactions with R50,
N112, F117, L192, M311, and A312. There is only a slight
difference in the interaction pattern with respect to compound
1 because of the change in the pose of the terminal naphthalene
ring. In the case of compound 1 the terminal naphthalene ring
interacts with T51 and V313, whereas in the case of compound
2 the terminal naphthalene ring interacts with R50. However,
apart from these differential interactions, this part of the
molecule is exposed to the solvent (Figure 4).
Regarding LcTS, in the folate-binding site, compound 1
interacts with D22, T24, E84, Y261, and R178′ by forming
hydrogen bonds through its phenol hydroxyls and with R23,
L195, D221, L224, V314, A315, and V316 through van der
Waals interactions.
In the second binding site of LcTS, near the small domain,
compound 1 forms hydrogen bonds with the side chain amino
group of K92, with the nitrogen of the D103 backbone, with
the C-terminus carboxyl group of V316 through its phenol
hydroxyls, and with the hydroxyl of Y118 and the backbone
nitrogen of G105 through its lactonic function. It also forms
van del Waals interactions with R23, T24, E84, E88, M101,
T1021, F104, H106, and A194.
Compound 2 shows, in the folate-binding site of LcTS, the
same binding mode as compound 1, but it also interacts with
G225.
On the other hand, in the second binding site of LcTS, which
is near the small domain, compound 2 shows a different binding
mode with respect to compound 1 even though it interacts with
the same residues. It forms hydrogen bonds through its phenyl
hydroxyl with the side chains of K92 and E84 and the backbone
carbonyl of M101.
Figure 4. Low-energy binding modes of compounds 1 (green) and 2
(pink) in hTS (A) and LcTS (B). In each enzyme, two possible binding
modes are shown: in the folate binding site (on the right of the images)
and in the binding site near the small domain (on the left of the images).
Only the most important residues (colored by atom) interacting with
ligands are shown.
residues of the small domain. All the conformations that were
outside the active site were disregarded.
The results suggest that these molecules can bind in these
two regions with multiple binding modes in both enzymes. The
folate-binding site is calculated to be more occupied and
energetically more favorable than the second (by about 1-1.4
kcal/mol) (Table 2).
The most favorable binding modes in each site of both
enzymes can be described in detail for compounds 1 and 2
(Figure 4) as follows.
In agreement with experimental inhibition activity data against
LcTS, AutoDock predicts similar behavior in enzyme binding
for both compounds. On the other hand, on the basis of
inhibition activity data against hTS, one could expect that some
differences may be found in binding to hTS. However, the only
difference between the most favorable binding modes of
compounds 1 and 2 in the hTS′ active site is that the terminal
naphthalene ring of compound 2, in its interaction with the folate
binding site, does not interact with M311 but with V79 and
F80. In particular, F80 may change its conformation and form
a stacking interaction with the naphthalene ring of the ligand.
Compound 1 binds in the folate-binding site of hTS, forming
hydrogen bonds through its phenol hydroxyls with the side chain
of D218, the carboxyl function of I108, and the hydroxyl group

Antibacterial Agent DiscoVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5963
Table 3. MIC for Compounds 5, 8, 10, 13-16, 18 (µg/mL)
compd
strains
5
8
10
13
25
>50
>50
10
10
10
10
10
25
10
25
10
25
10
10
10
14
15
16
18
25
>50
>50
10
10
10
10
10
25
10
10
10
10
10
10
10
cpx
Enterococcus faecalis ATCCC29212
Escherichia coli 256
Escherichia coli 292
2.5
25
10
>25
>25
2.5
10
5
10
5
5
5
5
10
>50
>50
5
>50
>50
5
>25
>25
5
12.5
>25
>25
>50
>50
25
25
25
25
25
25
25
25
25
25
25
25
25
>50
25
>25
>25
2.5
Listeria monocytogenes 3
Listeria monocytogenes 4
Listeria monocytogenes 5
Listeria monocytogenes 6
Listeria monocytogenes 7
Listeria monocytogenes 8
Listeria monocytogenes ATCC4428
Staphylococcus aureus 341
Staphylococcus aureus 343
Staphylococcus aureus K28
Staphylococcus aureus ATCC29213
Staphylococcus haemolyticus ATCC2997
Staphylococcus saprophiticus ATCC15305
Citrobacter 224
2.5
2.5
5
5
5
1
5
5
5
5
5
5
5
5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
1
2.5
2.5
2.5
5
2.5
5
2.5
2.5
5
1
2.5
1
2.5
2.5
5
1
2.5
2.5
1
5
1
2.5
2.5
2.5
5
2.5
5
2.5
1.2
0.5
0.5
0.5
2.5
1
5
>25
5
2.5
2.5
>25
2.5
1
>25
>50
10
>50
>50
>25
>50
10
>25
>5
Streptococcus 42
2.5
5
5
5
1
Table 4. Results of MTT Test on VERO Cellsa
This would be an important specific interaction with hTS that
could contribute significantly to the good affinity and inhibition
activity of compound 2 for hTS with respect to compound 1
(Ki vs hTS of 0.4 and .132 µM for compounds 2 and 1,
respectively). Moreover, a deeper analysis of all the predicted
binding modes for each compound reveals that there are few
secondary binding modes predicted by AutoDock for compound
2 that are not predicted for compound 1. These binding modes
allowed the formation of hydrogen bonds with the backbones
of R50 and I108 and with the phosphate group of the substrate;
of electrostatic interactions with the side chains of D49, N112,
D218, R175′, and the 3′-hydroxyl group of the ribose moiety
of dUMP; and of van del Waals interactions with 10 other
residues, including M311 and A312. Compound 1 shows similar,
but not identical, binding modes. In fact, it was not able to
assume these same binding modes because of clashes with M311
and A312.
compd
concn (mg/L)
control
5
10
15
80
>100
>100
>100
>100
>100
16
14
41
84
100
100
100
100
50
25
12.05
6.25
3.12
1.56
100
100
100
100
100
100
> 100
100
100
100
100
58
70
90
90
90
90
61
85
97
97
95
96
100
a Data are expressed as percentage of cellular growth. Control vessels
are without compounds.
Table 5. MIC Values (µg/mL) of Compouds 5, 10, and 15 Tested
against 23 Multiresistant Bacterial Strains
compd
strains
5
10
2.5
2.5
5
2.5
0.5
5
2.5
2.5
5
2.5
2.5
5
10
10
5
15
Staphylococcus epidermidis 45
Staphylococcus epidermidis 39
Staphylococcus epidermidis 42
Staphylococcus epidermidis 33
Staphylococcus epidermidis 35
Staphylococcus epidermidis 30
Staphylococcus epidermidis 32
Staphylococcus epidermidis 50
Staphylococcus epidermidis 21
Staphylococcus epidermidis 48
Staphylococcus epidermidis 36
Staphylococcus aureus 27
1
1
2.5
1
0.5
1
1
1
2.5
1
1
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
>65
Other differences are in regard to secondary binding modes
near the small domain region. Both compounds interact with
R50, R176′, and dUMP; however, while compound 1 also forms
hydrogen bonds with D49, S120, A191, and Y258, compound
2 hydrogen-bonds to M190 and R175′.
Profiling Bacterial Strain Inhibition
A subset of molecules (5, 8, 10, 13-16, and 18) was selected
and tested against a collection of 18 Gram-positive or Gram-
negative bacterial strains. The compounds were selected mainly
on the basis of their enzyme inhibition profile and solubility.
The minimum inhibitory concentrations (MIC), defined as the
lowest concentration at which bacterial growth is no longer
evident, were ascertained. The MIC values of the substances
are listed in Table 3 and compared to ciprofloxacin (cpx), a
member of the fluorquinolone family with a broad spectrum of
action. The active compounds 5, 10, and 13-16 showed a
limited spectrum of action, being active mainly toward Gram-
positive bacteria. The MIC values were often similar to that of
cpx, in the range 0.5-12.5 µg/mL.
The MTT test was performed to evaluate the toxicity of
compounds 5, 10, and 14-16 against VERO cell cultures. The
data in Table 4 clearly show the absence of toxicity, especially
for 5 and 15. Cell growth was not significantly modified by
the presence of different concentrations of the other compounds.
Three compounds (5, 10, and 15) were also tested against 23
clinical isolate multiresistant bacterial strains recovered from
hospitalized patients who have been treated with different
2.5
2.5
5
Staphylococcus aureus 15
Staphylococcus aureus 9
Staphylococcus aureus 17
2.5
5
5
Staphylococcus haemolyticus 51
Staphylococcus haemolyticus 56
Enterococcus faecium 52
5
10
10
>75
10
10
10
5
Enterococcus faecium 28
Enterococcus faecium 57
Enterococcus faecium 6
>100
5
5
Enterococcus gallinarum 37
Enterococcus gallinarum 64
2.5
1
2.5
antibiotics, such as â-lactam, macrolides, and aminoglycosidic.
The measured MIC values ranged between 2.5 and 0.5 µg/mL
(Table 5). The most interesting results are related to compounds
5 and 15, showing a MIC value of 0.5 µg/mL against
Staphylococcus epidermidis 35, a strain that is resistant to 17
antibiotics. No activity is evident for compound 10 against the
multiresistant strains (Table 5).
Figure 5 shows the growth curve recorded biophotometrically
over 24 h of Staphylococcus epidermidis alone (control) and in

5964 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Costi et al.
On the basis of the biolibrary screening results, eight
compounds were selected for multistep cell-based assays. The
first screening was done on 18 wild-type bacterial cell lines,
including Gram-positive and Gram-negative bacteria, such as
Enterobacter, Staphylococcus, and Escherichia. Of eight mol-
ecules, the five most interesting compounds were tested using
MTT tests to evaluate their cytotoxicity. Compounds 5, 15, and
10 showed null or very low toxicity in the assay, suggesting
that these molecules could be selected for further experiments.
Then a second level of cell-based assays was performed against
clinical isolate resistant strains. Different clinical specimens
(blood and purulent exudates) from the microbiology laboratory
of Modena Hospital were processed for cultural analysis.19
Particular attention was paid to the Enterococcus strains with
full or medium sensitivity to teicoplanin and vancomycin. Two
of them were fully resistant and one was medium-resistant to
vancomycin. Among the Staphylococcus strains, 13 showed
medium resistance to teicoplanin, 10 of them belonged to the
species Staphylococcus epidermidis, while another belonged to
Staphylococcus aureus and the other two to Staphylococcus
haemoliticus. Staphylococcus and Enterococcus are among the
Gram-positive that are most frequently involved in infective
processes.
Figure 5. Growth curves of Staphylococcus epidermidis, clinical
isolate, in the presence of compound 5: (blue) control; (pink) 0.1 µg/
mL; (yellow) 0.5 µg/mL; (light-blue) 1 µg/mL; (violet) 5 µg/mL;
(brown) 25 µg/mL.
the presence of different concentrations of compound 5. At 0.1
µg/mL no effect on Staphylococcus epidermidis growth was
observed. By contrast, antibacterial activity was evident after 8
h at a dose of 0.5 µg/mL. The best results were obtained with
1.0 µg/mL (corresponding to the MIC value for most of the
strains). The remaining dosages had a bactericidal effect.
Discussion
In our study, compounds 5 and 15 showed good activity
against most of the multiresistant strains tested, in particular
against Staphylococcus epidermidis. In contrast, the previously
reported compound 10 was completely ineffective against the
resistant strains, even if it was effective against wild-type Gram-
positive bacteria (Table 5).
Compound 5 was tested against all the above-described
multiresistant strains, and the best result was obtained against
Staphylococcus epidermidis 35 (Figure 5), showing an MIC
value of approximately 0.5 µg/mL.
New 1,2-naphthalein derivatives have been designed and
synthesized. All the compounds and a few more previously
reported 1,8-naphthaleins were tested against a TS-based
biolibrary mostly from pathogen TS enzymes, in addition to
hTS and hDHFR. In the 1,2-naphthalein series the most species-
specific was compound 1 (Ki against CnTS of 0.4 µM and SI
K
i hTS/Ki CnTS of 330 with respect to hTS). By contrast,
compound 2 was very active against both pathogen and human
TS (Ki for the latter of 0.4 µM). In the 1,8-naphthalein series,
compounds 10 and 11 were the most active against LcTS (Ki
of 0.7 µM), while compound 16 was the most species-specific
(SI of 306, with respect to this enzyme). The most species-
specific compounds, of great interest as leading antibacterial
agents, were 5, 10, 13-16, and 18. The Ki values ranged from
0.7 to 1.7 as the best result for each compound considered. The
SI Ki hTS/Ki LcTS determined so far ranged from 50 to 330.
Out of all the derivatives, 17 was the most species-specific
(SI Ki hTS/Ki EcTS ) 365).
Conclusions
We have previously synthesized and tested a series of 1,8-
naphthalein analogues, which showed interesting biological
activity profiles against model bacterial cell lines from ATCC
and low toxicity vs human cells. The best inhibitors showed a
Ki of 0.5 µM and a MIC of 1.8 µg/mL. In the present paper we
obtained a compound (5) showing 2 times better efficacy against
clinical isolates bacterial cells. The Ki value was in the same
order as those of the previous naphthaleins; thus, we successfully
improved the cellular pharmacokinetic profile. The clinical
isolates showed drug resistance vs a number of antibiotics and
chemotherapeutic agents including vancomycin, an antibiotic
indicated for the treatment of serious, life-threatening infections
by Gram-positive bacteria, which are unresponsive to other less
toxic antibiotics. Compound 5 is suggested as a possible lead
with promising antibacterial activity.
The chemistry of 1,2-naphthalein derivatives allowed an easier
derivatization of the core skeleton from the perspective of further
functionalization with respect to 1,8-naphthalein analogues.
As an important remark, we observe that an interpretation of
the specificity profile of the two compounds 1 and 2 against
LcTS and hTS enzymes is accessible through modeling studies.
The parallel behavior of the biological profile against bacterial
cell lines does not lead to the conclusion that thymidylate
synthase enzymes are the exclusive intracellular target of the
identified leads. More pharmacology studies are needed for this
conclusion.
The selected species-specific compounds were considered for
further in vitro studies against bacterial wild-type and clinical
isolates.
The data referring to the 1,2-naphthalein series suggest a
relationship between the structural isomeric form and the
observed specificity. In fact, only â-isomer 2 showed affinity
for both pathogen and human enzymes while the R-isomers were
all inactive (1, 3-5, 8) or poorly active (7) toward the latter.
With the aim of understanding the observed enzyme inhibition
profile, molecular modeling studies were performed and the
AutoDock 3.0.5 program was applied to calculate the most
favorable binding site of compounds 1 and 2 toward LcTS and
hTS. The results suggested that these molecules could bind in
both enzymes in two regions, in the folate-binding site and near
the small domain region. As already observed in the case of
other naphthalein derivatives and in agreement with experi-
mental data derived from site-directed mutagenesis and enzyme
kinetic studies,14 for compounds 1 and 2 the AutoDock results
suggest a multiple binding mode that is typical of the phthalein
structure.14,17 The results highlight some differences in the
binding modes of compounds 1 and 2 in the hTS active site. In
particular, compound 2 seems to be able to take advantage of
the favorable interactions with F80, M311, and A312.
The results presented confirm the relevance of the naphthalein
derivatives as antinfective agents and are encouraging further
developments to clarify their full mechanism of action.

Antibacterial Agent DiscoVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5965
solvent, the residue was purified by flash chromatography, eluting
with dichloromethane/methanol, 98/2, to give 0.004 g of 16 (10.2%).
Anal. (C24H13Br2ClO4) C, H, Br, Cl.
Experimental Section
Chemistry. The melting points were determined using Bu¨chi
510 capillary melting point apparatus and are uncorrected. The
elemental analyses for the test compounds were within (0.4 of
the theoretical values. TLC on silica gel plates was used to check
product purity. Silica gel 60 (Merck, 230-400 mesh) was used for
flash chromatography. The structures of all the compounds were
consistent with their analytical and spectroscopic data. NMR spectra
were performed on a Bruker FT-NMR AVANCE400.
NMR data of compounds 1-8, 16, 18 are reported in Table 6.
Synthesis of 1,1-Bis(4-hydroxyphenyl)-1H-naphtho[1,2-c]fu-
ran-3-one (1) and 3,3-Bis(4-hydroxyphenyl)-3H-naphtho[1,2-c]
furan-1-one (2). The compounds were obtained by condensing the
1,2-naphthalic anhydride and phenol according to a known proce-
dure.20 Separation was accomplished by flash chromatography,
eluting with dichloromethane/methanol, 98/2. For 1, the yield was
15%. For 2, the yield was 1%. Anal. (C24H16O4) C, H.
Synthesis of 1,1-Bis(4-hydroxy-3-iodophenyl)-1H-naphtho-
[1,2-c]furan-3-one (3). Iodine monochloride (0.65 mmol) in glacial
acetic acid (1 mL) was added dropwise to a mixture of 1 (0.05 g,
0.14 mmol) suspended in glacial acetic acid (2.5 mL), and the
mixture was stirred overnight at room temperature. After evapora-
tion of the solvent, the residue was purified by flash chromatog-
raphy, eluting with dichloromethane/methanol, 98/2, to give 0.062
g of 3 (71.2%). Anal. (C24H14I2O4) C, H, I.
Synthesis of 1,1-Bis(3-bromo-4-hydroxyphenyl)-1H-naphtho-
[1,2-c]furan-3-one (4). A solution of bromine (0.27 mmol) in
dichloromethane (1 mL) was added dropwise to a solution of 1
(0.05 g, 0.14 mmol) in dichloromethane (3 mL), under stirring at
room temperature. After 5 h the solvent was evaporated and the
residue purified by flash chromatography, eluting with dichlo-
romethane/methanol, 98/2, to give 4 (28%). Anal. (C24H14Br2O4)
C, H, Br.
Synthesis of 1,1-Bis(3,5-dibromo-4-hydroxyphenyl)-1H-naph-
tho[1,2-c]furan-3-one (5). Excess bromine (0.7 mL) was added
dropwise at room temperature to a solution of 1 (0.05 g, 0.14 mmol)
in ethanol (2 mL), and the mixture was stirred overnight. The
solvent was evaporated and the residue purified by flash chroma-
tography, eluting with cyclohexane/ethyl acetate, 80/20, to give 5
(64.5%). Anal. (C24H12Br4O4) C, H, Br.
Synthesis of 1,1-Bis(4-hydroxy-5-isopropyl-2-methylphenyl)-
1H-naphtho[1,2-c]furan-3-one (6). Aluminum chloride (0.67 g,
5 mmol) was added portionwise to a mixture of 1,2-naphthalic
anhydride (0.5 g, 2.5 mmol) and thymol (5 mmol) in s-tetrachlo-
roethane (25 mL), and the mixture was stirred at 115 °C for 3 days.
The still hot mixture was poured onto ice and dichloromethane was
added. After stirring for 0.5 h, the insoluble was filtered off. The
organic layer was separated, dried over sodium sulfate, and
concentrated under vacuum. The residue was purified by flash
chromatography, eluting with cyclohexane/ethyl acetate, 70/30, to
give 0.07 g of 6 (5.8%). Anal. (C32H32O4) C, H.
Synthesis of 3,3-Bis(3-chloro-4-hydroxyphenyl)-3H-naphtho-
[1,2-c]furan-1-one (7) and 3,3-Bis(3-fluoro-4-hydroxyphenyl)-
3H-naphtho[1,2-c]furan-1-one (8). A mixture of 1,2-naphthalic
anhydride (0.5 g, 2.5 mmol), the appropriate phenol (5.0 mmol),
and a few drops of sulfuric acid was heated under stirring at 180
°C for 5 h. After the mixture was cooled, water was added and the
mixture was extracted with dichloromethane (3 × 30 mL). The
organic layer was dried over sodium sulfate, and the solvent was
evaporated. The residue was purified by flash chromatography,
eluting with dichloromethane/methanol, 95/5. For 7: 0.085 g
(7.8%). Anal. (C24H14Cl2O4). For 8: 0.068 g (6.8%). Anal.
(C24H14F2O4) C, H, F.
6(7)-Chloro-3,3-bis(3-fluoro-4-hydroxyphenyl)-1H,3H-naph-
tho[1,8-c,d]pyran-1-one (18). To a mixture of the commercially
available 4-chloro-1,8-naphthalic anhydride (0.5 g, 2.15 mmol) and
2-fluorophenol (0.383 mL, 4.3 mmol) 5 drops of H2SO4 were added.
The mixture was left under stirring at 180 °C for 5 h. After the
mixture was cooled, the residue was purified by flash chromatog-
raphy, eluting with dichloromethane/methanol (98/2) to give 0.02
g of 18 (2.12%). Anal. (C24H13CIF2O4) C, H, Cl, F.
It is noted that because of the presence of a chlorine atom on
the naphthalene ring, in the cases of both 16 and 18, a mixture of
6 or 7 substituted isomers was obtained, which was tested as such.
Biological Evaluation. Purification of the Enzymes. All the
strains and plasmids were provided by Dr. D. V. Santi, from the
University of California, San Francisco. The expression and
purification of LcTS, EcTS, CnTS, and hTS were done following
known procedures.21-24 The enzyme LcTS was purified by column
chromatography using phosphocellulose (P11, Biorad) and hy-
droxyapatite (HAP, Biorad) resin, using phosphate buffer as the
eluent. EcTS, CnTS, and hTS were purified as previously reported.13
The enzyme preparations were >95% homogeneous as visualized
by SDS (sodium dodecyl sulfate) polyacrylamide gel electrophore-
sis. The purified enzymes were stored at -80 °C in 10 mM
phosphate buffer, pH 7.0, and 0.1 mM EDTA. The enzyme activity
was determined spectrophotometrically by steady-state kinetic
analysis, following the increasing absorbance at 340 nm due to the
oxidation reaction of N5,N10-methylenetetrahydrofolate (mTHF) to
dihydrofolate (DHF).25 An amount of 1 mL of reaction solution
was formed by standard assay buffer, pH 7.4, dUMP (100 µM),
6(R,S)-1-CH2CH4-folate (140 µM) (Sigma), and enzyme (0.07 µM).
Assays were performed at 20 °C in the standard assay buffer formed
by TES (N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid)
(50 mM) at pH 7.4, MgCl2 (25 mM), formaldehyde (6.5 mM),
EDTA (1 mM), and 2-mercaptoethanol (75 mM). hDHFR (DHFR,
E.C. 1.5.1.3) was purified through alternate steps of column
chromatography and ammonium sulfate precipitations as reported.26
DHFR catalyzes the NADPH-dependent reduction of dihydrofolate
(DHF) to tetrahydrofolate (THF). The assays were run spectropho-
tometrically by measuring the decrease in absorbance at 340 nm
upon DHF reduction at 25 °C. One enzyme unit is defined as 1
nmol of DHF reduced per minute. The reaction mixture was formed
by 50 µM TES, pH 7.0, 1 µM EDTA, 75 µM 1-mercaptoethanol,
100 µM NADPH, and 58.03 µM DHF.27
Assays. Stock solutions of the inhibitors were prepared in DMSO
(dimethyl sulfoxide) and stored at -20 °C until use. Compound 4
was tested as a racemic mixture. The inhibition pattern for all the
compounds was determined by steady-state kinetic analysis of the
dependence of enzyme activity on folate concentration at varying
inhibitor concentrations. All the compounds showed competitive
inhibition. Ki values were obtained from the linear least-squares
fit of the residual activity as a function of inhibitor concentration,
using suitable equations for competitive inhibition.28 Each experi-
ment was repeated at least three times, and no individual measure-
ment differed by more than 20% from the mean. The maximum
solubility is reported when a molecule does not show any inhibition
of TS activity at the solubility limit. To better understand the
biological activity profile, we simulated the IC50 values assuming
an inhibition of 2% at the solubility limit and calculating a projected
IC50 for the compounds studied.9,17 This number underestimates
the potency of the inhibition, but it is helpful in the comparative
analysis of the specificity index (SI),9 which was determined by
the ratio Ki hTS/Ki other enzyme. SSI was determined by the ratio Ki hTS
/
Ki hDHFR
.
Synthesis of 1,8-Naphthaleins. 6(7)-Chloro-3,3-bis(3-bromo-
4-hydroxyphenyl)-1H,3H-naphtho[1,8-c,d]pyran-1-one (16). Bro-
mine (7.5 × 10-3 mL, 0.146 mmol) in 0.5 mL of dichloromethane
was added dropwise, under stirring, to a solution of 6-chloro-3,3-
bis(4-hydroxyphenyl)-1H,3H-naphtho[1,8-c,d]pyran-1-one13 (0.03
g, 0.07 mmol) in 2 mL of dichloromethane. The reaction mixture
was stirred at room temperature for 4 h. After evaporation of the
The effect of increasing DMSO concentration in the TS assay
mixture was studied, and it was observed that no change in TS
activity was seen at concentrations up to 8% DMSO.29
Microbiology. A collection of 18 Gram-positive and Gram-
negative wild-type bacterial strains were tested against the selected
molecules (Table 3).

5966 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Table 6. NMR Data (δ, ppm) of Compounds 1-8, 16, 18
Costi et al.
1
3
4
5
6
1
3
4
5
6
1
2
3
4
5
6
7
8
9
10
11
12
13
7.94 d
7.69 t
7.84 t
7.97 d
7.77 t
7.88 t
8.34 d
8.40 d
8.08 d
7.96 d
7.74 t
7.87 t
8.32 d
8.38 d
8.07 d
7.00 d Jo
8.12 d
7.75 t
7.84 t
8.28 d
8.34 d
8.14 d
7.79 s
7.70 d
7.58 t
7.80 t
8.30 d
8.35 d
8.00 d
7.05 d Jm
14
15
6′
7′
8′
1.10 t
)H15
8.30 d
8.35 d
8.05 d
7.31 d Jo
6.99 d Jo
)H8
)H7
)H8
)H9
)H10
)H7
7.60 d Jm
)H7
)H7
6.60 d Jm
6.90 d Jm
9′
)H9
)H10
7.32 over.
7.15 d Jo
7.54 d Jm
10′
11′
12′
13′
14′
15′
9.55 s
2.22 s
3.05 m
0.65 t
0.77 t
7.20 d Jo
7.37 dd Jo-Jm
)H8
)H7
6.53, d Jm
)H7
)H7
9.50 s
1.72 s
3.22 m
1. COSY: H1-H2; H2-Hl3; H3-H4; H5-H6; H7-H8. NOE: H4-H5; H1-H7. Solvent: acetone-d6. Temp ) 25 °C.
3. COSY: H1-H2,H3; H2-H3; H4-H3,H2; H5-H6; H9-H10; H7-H10. NOE: H4-H5; H1-H2,H7,H10. Solvent: acetone-d6.
Temp ) 25 °C.
4. COSY: H1-H2,H3; H2-H3; H4-H3,H2; H5-H6; H7′-H10′; H9′-H10′; NOE: H4-H5; H1-H7. Solvent: acetone-d6.
Temp ) 25 °C.
5. COSY: H1-H2; H2-H3; H3-H4; H5-H6. Solvent: acetone-d6 + benzene-d6. Temp ) 25 °C.
6. COSY: H1-H2; H2-H3; H3-H4; H5-H6; H7-H8. NOE: H4-H5; H1-H7. Solvent: acetone-d6. Temp ) 25 °C.
2
7
8
16
18
1
2
3
4
5
6
7
8
7.95 d
9.14 d
7.95 t
7.85 t
8.26 d
8.50 d
8.02 d
7.54 d Jm
9.14 d
7.97 t
7.87 t
8.29 d
8.52 d
8.02 d
1
2
3
4
5
6
7
8
8.51 d
8.03 t
8.68 d
7.95 d
7.20 d
7.28 d Jm
8.49 d
7.69 t
7.41 t
8.30 d
8.35 d
8.05 t
7.31 d
6.99 d
)H8
8.01 t
8.65 d
7.93 d
7.17 d
7.20 dd 3JHF-Jm
7.29 dd 3JHF-Jm
7.03 d Jo
6.97 dd Jo-Jm
)H6
7.05 t Jo
7.10 dd Jo-Jm
)H6
9
7.20 d Jo
7.37 dd Jo-Jm
)H7
7.17 dd 4JHF-Jo
7.22 dd Jo-Jm
)H7
9
10
7′
8′
9′
10′
)H7
6′
7′
8′
9′
)H7
)H8
)H8
)H9
)H8
)H9
)H9
)H9
)H10
)H9
)H10
)H10
16. COSY: H1-H2; H2-H3; H4-H5; H6-H9; H8-H9;
H6′-H9′; H8′-H9′. Solvent: DMSO-d6. Temp ) 25 °C.
18. COSY: H1-H2; H2-H3; H4-H5; H6-H9; H8-H9;
H6′-H9′; H8′-H9′, Solvent: DMSO-d6. Temp ) 25°C.
2. COSY: H1-H2, H2-H3; H3-H4, H5-H6; H7-H9; H7-H12;
H9-H12; H14,15-H13; H7′-H9′; H7′-H12′; H9′-H12′;
H14′-H13′; H15′-H13′. NOE: H1-H7; H9-H11;
H7-H14,H15; H1-H7′; H9′-H11′; H7′-H14′; H7′-H15′.
Solvent: DMSO-d6. Temp ) 25 °C.
7. COSY: H1-H2; H2-H3; H3-H4; H5-H6; H9-H10; H7-H10.
NOE: H4-H5; H6-H10; H6-H7.
Solvent: acetone-d6. Temp ) 25 °C.
8. COSY: H1-H2; H2-H3; H3-H4; H5-H6; H9-H10; H7-H10.
NOE: H4-H5. Solvent: acetone-d6. Temp ) 25 °C.
Twenty-three strains, comprising strains of Staphylococcus
epidermidis (11 strains), Staphylococcus aureus (4 strains), Sta-
phylococcus haemoliticus (2 strains), Enterococcus faecium (4
strains), and Enterococcus gallinarum (2 strains), were tested. All
the strains showed multidrug resistance and were clinical isolates
recovered from blood or urine of patients without malignancies.
The strains were resistant to at least 17 of the most used antibacterial
agents including vancomycin19,30-36 (Supporting Information).

Antibacterial Agent DiscoVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5967
Sensitivity Test. The Gram-positive and Gram-negative bacteria
were cultured in a Mueller-Hinton broth (Difco Milan, Italy) at
37 °C for 24 h. While the bacteria were developing exponentially,
they were diluted in a liquid sterile medium to obtain a final
innoculum of approximately 1 × 104 CFU/mL and subsequently
put in contact with increasing concentrations of the compounds
under study. The minimum inhibitory concentrations (MICs),
defined as the lowest concentration at which bacterial growth was
no longer evident, were ascertained.
Cytotoxicity Test. The Trypan blue exclusion and MTT tests
were performed using VERO cells cultured in MEM medium
supplemented with 5% heat-inactivated (56 °C, 30 min) fetal calf
serum (FCS), 1% of penicillin (50 U/mL), streptomycin (50 µg/
mL), and L-glutamine (1%), according to the method described by
Ishioka.37 The cultures were observed for 2 days under the
microscope with indirect light. Afterward, a count of living cells
was performed. Subsequently, increasing concentrations of the
substances were put in contact with the cultures of VERO cells. A
vessel in which only cells were present was assumed as a control.
After 24 h of incubation, the measurement of active mitochondrial
dehydrogenases of living cells (MTT test) was performed.38
Molecular Docking. The program AutoDock 3.0.539 was used
for the docking of compounds 1 and 2. Since multiple three-
dimensional structures of LcTS were available in the PDB,40 two
different conformations of LcTS (1TSL, 1TSM)14 were considered,
together with a modeled open conformation of hTS.15 Ligands and
water molecules were removed from the active site. A molecule of
dUMP was added instead and considered part of the target for
docking. The structure and coordinates of the dUMP molecule were
taken from the ternary complex structures of LcTS (1LCA)41 and
hTS (1I00),42 respectively, after matching, on the basis of protein
trace atoms, these structures to those used for the AutoDock
calculations. The same starting complexes (1TSL + dUMP, hTS
+ dUMP) were already used in another published work.15 To run
AutoDock, dimeric protein structures were prepared with InsightII.43
Polar hydrogen atoms were added using the WHATIF program.44
Solvation parameters and atomic partial charges were assigned to
the protein using the programs q.kollua and addsol, which are
included in the AutoDock program suite. For dUMP, the solvation
parameters and atom partial charges were manually added on the
basis of the atoms’ similarity to protein atoms and of literature data,
respectively.45 Compounds 1 and 2 were built using InsightII. Their
geometries were optimized using the program Amsol 6.6 (AM1).46
The atomic partial charges were calculated using the program RESP
(Amber 6.0),47 on the basis of electrostatic potentials generated by
Gaussian 98 (HF/6-31G* as basis set).48 The program Babel49 was
used to prepare the input Z matrix of Amsol and Gaussian. A cubic
grid with 22.5 Å long sides, centered on the TS active site, was
defined in the case of hTS, whereas cubic grids with 20.6 and 30.4
Å long sides were used with 1TSL and 1TSM, respectively. These
grids were wide enough to comprise all of the enzymatic active
sites. A Lamarckian genetic algorithm was used to generate 50
bound conformations for compounds 1 and 2 in each protein studied.
All nonterminal rotable bonds were allowed to rotate during the
calculation. All other settings were kept as default. The results were
visually analyzed using the InsightII program. The results of the
calculations performed on the two conformations of LcTS were
considered altogether. On the basis of the clustering histogram
output from the AutoDock program, the lowest energy conformation
of each cluster was selected. The selected conformations were
grouped on the basis of their binding sites. Two main binding sites
were considered, the folate and the crystallographic binding site of
compound 10.14,15 For each binding site, multiple binding modes
were considered. All the conformations that were outside the active
site were disregarded.
Supporting Information Available: Multidrug resistance data
of clinical isolates multidrug-resistant strains used and elemental
analysis results of the synthesized compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Carreras, C. W.; Santi, D. V. The catalytic mechanism and structure
of thymidylate synthase. Annu. ReV. Biochem. 1995, 64, 721-762.
(2) Chu, E.; Callender, M. A.; Farrell, M. P.; Schmitz, J. C. Thymidylate
synthase inhibitors as anticancer agents: from bench to bedside.
Cancer Chemother. Pharmacol. 2003, 52 (Suppl. 1), S80-S89.
(3) Liu, J.; Schmitz, J. C.; Lin, X.; Tai, N.; Yan, W.; Farrell, M.; Bailly,
M.; Chen, T.; Chu, E. Thymidylate synthase as a translational
regulator of cellular gene expression. Biochim. Biophys. Acta 2002,
1587, 174-182.
(4) Chu, E.; Allegra, C. J. The role of thymidylate synthase as an RNA
binding protein. BioEssays 1996, 18, 191-198.
(5) Bruni, P.; Minopoli, G.; Brancaccio, T.; Napolitano, M.; Faraonio,
R.; Zambrano, N.; Hansen, U.; Russo, T. Fe65, a ligand of the
Alzheimer’s beta-amyloid precursor protein, blocks cell cycle
progression by down-regulating thymidylate synthase expression. J.
Biol. Chem. 2002, 277 (38), 35481-35488.
(6) Costi, M. P.; Ferrari, S.; Venturelli, A.; Calo`, S.; Tondi, D.; Barlocco,
D. Thymidylate synthase structure, function and implication in drug
discovery. Curr. Med. Chem. 2005, 12 (19), 2241-2258.
(7) Tondi, D.; Venturelli, A.; Ferrari, S.; Ghelli, S.; Costi, M. P.
Improving specificity vs bacterial thymidylate synthases through
N-dansyl modulation of didansyltyrosine. J. Med. Chem. 2005, 48,
913-916.
(8) Calo`, S.; Tondi, D.; Venturelli, A.; Ferrari, S.; Pecorari, P.; Rinaldi,
M.; Ghelli, S.; Costi, M. P. A step further in the discovery of phthalein
derivatives as thymidylate synthase inhibitors. ARKIVOC 2004, 382-
396.
(9) Ghelli, S.; Rinaldi, M.; Barlocco, D.; Gelain, A.; Pecorari, P.; Tondi,
D.; Rastelli, G.; Costi, M. P. Ortho-halogen naphthaleins as specific
inhibitors of Lactobacillus casei thymidylate synthase. Conforma-
tional properties and biological activity. Bioorg. Med. Chem. 2003,
11, 951-963.
(10) Costi, M. P.; Tondi, D.; Rinaldi, M.; Barlocco, D.; Pecorari, P.;
Soragni, F.; Venturelli, A.; Stroud, R. M. Structure-based studies on
species-specific inhibition of thymidylate synthase. Biochim. Biophys.
Acta 2002, 1587, 206-214.
(11) Fritz, T. A.; Tondi, D.; Finer-Moore, J. S.; Costi, M. P.; Stroud, R.
M. Predicting and harnessing protein flexibility in the design of
species-specific inhibitors of thymidylate synthase. Chem. Biol. 2001,
8, 981-995.
(12) Tondi, D.; Slomczynska, U.; Costi, M. P.; Watterson, D. M.; Ghelli,
S.; Shoichet, B. K. Structure-based discovery and in-parallel opti-
mization of novel competitive inhibitors of thymidylate synthase.
Chem. Biol. 1999, 6, 319-331.
(13) Costi, M. P.; Rinaldi, M.; Tondi, D.; Pecorari, P.; Barlocco, D.; Ghelli,
S.; Stroud, R. M.; Santi, D. V.; Stout, T. J.; Musiu, C.; Marangiu, E.
M.; Pani, A.; Congiu, D.; Loi, G. A.; La Colla, P. Phthalein
derivatives as a new tool for selectivity in thymidylate synthase
inhibition. J. Med. Chem. 1999, 42, 2112-2124.
(14) Stout, T. J.; Tondi, D.; Rinaldi, M.; Barlocco, D.; Pecorari, P.; Santi,
D. V.; Kuntz, I. D.; Stroud, R. M.; Shoichet, B. K.; Costi, M. P.
Structure-based design of inhibitors specific for bacterial thymidylate
synthase. Biochemistry 1999, 38, 1607-1617.
(15) Ferrari, S.; Costi, M. P.; Wade, R. Inhibitor specificity via protein
dynamics: insights from the design of antibacterial agents targeted
against thymidylate synthase. Chem. Biol. 2003, 10, 1183-1193.
(16) Hardy, L. W.; Finer-Moore, J. S.; Montfort, W. R.; Jones, M. O.;
Santi, D. V.; Stroud, R. M. Atomic structure of thymidylate
synthase: target for rational drug design. Science 1987, 235, 448-
455.
(17) Shoichet, B. K.; Stroud, R. M.; Santi, D. V.; Kuntz, I. D.; Perry, K.
M. Structure-based discovery of inhibitors of thymidylate synthase.
Science 1993, 259, 1445-1450.
(18) Hershberg, E. B.; Fieser, L. F. 3,4-Dihydro-1,2-naphthalic anhydride.
Org. Synth., Collect. Vol. 1969, II, 194-196, 423-424.
(19) Murray, P. R.; Baron, E. J.; Pfaller, M. A.; Tenover, F. C.; Yolken,
R. H. Manual of Clinical Microbiology, 7th ed.; AMS Press:
Washington, DC, 1999.
(20) Hubacher, M. H. The phthaleins from phenol and 1,2-naphthalene-
dicarboxylic acid. J. Am. Chem, Soc. 1944, 66, 255-256.
(21) Kealey, J. T.; Santi, D. V. Purification methods for recombinant
Lactobacillus casei thymidylate synthase and mutants: a general,
automated procedure. Protein Expression Purif. 1992, 4, 380-385.
(22) Maley, G. F.; Maley, F. Properties of a defined mutant of Escherichia
coli thymidylate synthase. J. Biol. Chem. 1988, 263, 7620-7627.
Acknowledgment. This work was supported by MIUR Cofin
2004 (Grant Costi 2004030405_004) and FIRST 2004. We thank
Dr. D. V. Santi (University of California, San Francisco) for
giving us plasmids and strains for hTS, CnTS, EcTS, and LcTS

5968 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Costi et al.
(23) Livi, L. L.; Edman, U.; Schneider, G. P.; Greene, P. J.; Santi, D. V.
Cloning, expression and characterization of thymidylate synthase from
Cryptococcus neoformans. Gene 1994, 150, 221-226.
(24) Davisson, V.; Sirawaraporn, W.; Santi, D. V. Expression of human
thymidylate synthase in Escherichia coli. J. Biol. Chem. 1989, 264,
9145-9148.
(39) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W.
E.; Belew, R. K.; Olson, A. J. Automated docking using a Lamarckian
genetic algorithm and an empirical binding free energy function. J.
Comput. Chem. 1998, 19, 1639-1662.
(40) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.;
Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data Bank.
Nucleic Acids Res. 2000, 28, 235-242.
(41) Finer-Moore, J. S.; Fauman, E. B.; Foster, P. G.; Perry, K. M.; Santi,
D. V.; Stroud, R. M. J. Refined structure of substrate-bound and
phosphate-bound thymidylate synthase from Lactobacillus casei. J.
Mol. Biol. 1993, 232, 1101-1116.
(42) Almog, R.; Waddling, C. A.; Maley, F.; Maley, G. F.; Van Roey, P.
Crystal structure of a deletion mutant of human thymidylate synthase
∆(7-29) and its ternary complex with Tomudex and dUMP. Protein
Sci. 2001, 10, 988-996.
(43) InsightII, Modelling Environment. 2000. http://www.accelrys.com.
(44) Vriend, G. WHATIF: A molecular modelling and drug design
program. J. Mol. Graphics 1990, 8, 52-56.
(45) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.,
Jr.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.;
Kollman, P. A. A second generation force field for the simulation of
proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc.
1995, 117, 5179-5197.
(46) Hawkins, G. D.; Giesen, D. J.; Lynch, G. C.; Chambers, C. C.; Rossi,
I.; Storer, J. W.; Li, J.; Winget, P.; Rinaldi, D.; Liotard, D. A.; Cramer,
C. J.; Truhlar, D. G. AMSOL, version 6.6; University of Minnesota:
Minneapolis, MN, 1997.
(47) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham, T. E., III;
Ross, W. S.; Simmerling, C. L.; Darden, T. A.; Merz, K. M.; Stanton,
R. V.; Cheng, A. L.; Vincent, J. J.; Crowley, M.; Tsui, V.; Radmer,
R. J.; Duan, Y.; Pitera, J.; Massova, I.; Seibel, G. L.; Singh, U. C.;
Weiner, P. K.; Kollman, P. A. AMBER 6; University of California:
San Francisco, CA, 1999.
(48) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.,
Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.;
Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P.
Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M.
A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.;
Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J.
L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98;
Gaussian, Inc.: Pittsburgh, PA, 1998.
(25) Pogolotti, A. L., Jr.; Danenberg, P. V.; Santi, D. V. Kinetics and
mechanism of interaction of 10-propargyl-5,8-dideazafolate with
thymidylate synthase. J. Med. Chem. 1986, 29, 478-482.
(26) Dann, J. G.; Ostler, G.; Bjur, R. A.; King, W. R.; Scudder, P.; Roberts,
G. C. K.; Burgen, S. V. Large-scale purification and characterization
of dihydrofolate reductase from a methotrexate resistant strain of
Lactobacillus casei. Biochem. J. 1976, 157, 559-571.
(27) Meek, T. D.; Garvey, E. P.; Santi, D. V. Purification and character-
ization of bifunctional thymidylate synthetase-dihydrofolate reductase
from methotrexate-resistant Leismania tropicae. Biochemistry 1985,
24, 678-686.
(28) Segel, I. H. Enzyme Kinetics. BehaViour and Analysis of Rapid
Equilibrium and Steady-State Enzyme Systems; John Wiley and
Sons: New York, 1975; p 105.
(29) Soragni, F.; Costi, M. P. Unpublished data.
(30) Centers for Disease Control and Prevention. Staphylococcus aureus
resistant to vancomycin. United States. MMWR Morbid. Mortal. Wkly
Rep. 2002, 51 (26), 565-567.
(31) Tabaqchali, S. Vancomycin-resistant Staphylococcus aureus: apoca-
lypse now? Lancet 1997, 350, 1644-1645.
(32) Noble, W. C.; Virani, Z.; Cree, R. G. A. Co-transfer of vancomycin
and other resistance genes from Enterococcus faecalis NCTC 12201
to Staphylococcus aureus. FEMS Microbiol. Lett. 1992, 93, 195-198.
(33) Hiramatsu, K.; Hanak, H.; Ino, T.; Yabuta, K.; Oguri, T.; Tenover,
F. C. Methicillin-resistant Staphylococcus aureus clinical strain with
reduced vancomycin susceptibility. J. Antimicrob. Chemother. 1997,
40, 135-136.
(34) Mainardi, J. l.; Shalaes, D. M.; Goering, J. H.; Shalaes, J. H.; Acar,
J. F.; Goldtein, F. W. Decreased teicoplanin susceptibility of
methicillin-resistant strains of Staphylococcus aureus. J. Infect. Dis.
1995, 171, 1646-1650.
(35) Diekema, D. J.; Pfaller, M. A.; Jones, R. N. The SENTRY Participants
Group. Age-related trends in pathogen frequency and antimicrobial
susceptibility of bloodstream isolates in North America. SENTRY
Antimicrobial Surveillance Program, 1997-2000. Int. J. Antimicrob.
2002, 20, 412-418.
(36) Stinear, T. P.; Olden, D. C.; Johnson, P. D. R.; Davies, J. K.; Gryson,
M. L. Enterococcal vanB resistance locus in anaerobic bacteria in
human faeces. Lancet 2001, 357, 855-856.
(37) Ishioka, C.; Kanamaru, R.; Sato, T.; Dei, T.; Konishi, Y.; Asamura,
M.; Wakui, A. Inhibitory effects of prostaglandin A2 on c-myc
expression and cell cycle progression in human leukemia cell line
HL-60. Cancer Res. 1988, 48, 2813-2818.
(38) Mosman, T. Rapid colorimetric assay for cellular growth and
survival: application to proliferation and cytotoxicity assays. J.
Immunol. Methods 1983, 65, 55-63.
(49) Walters, P.; Stahl, M. Babel, version 1.6; http://smolg.com/chem/babel.
JM051187D