ortho-halogen naphthaleins as specific inhibitors of Lactobacillus casei thymidylate synthase. Conformational properties and biological activity

Article abstract of DOI:10.1016/S0968-0896(02)00541-2

Thymidylate synthase (TS) (EC 2.1.1.45), an enzyme involved in the DNA synthesis of both prokaryotic and eukaryotic cells, is a potential target for the development of anticancer and antinfective agents. Recently, we described a series of phthalein and na

Full text of DOI:10.1016/S0968-0896(02)00541-2

Bioorganic & Medicinal Chemistry11 (2003) 951–963
ortho-Halogen Naphthaleins as Specific Inhibitors of Lactobacillus
casei Thymidylate Synthase. Conformational Properties and
Biological Activity
Stefano Ghelli,a,y Marcella Rinaldi,b Daniela Barlocco,c Arianna Gelain,c
Piergiorgio Pecorari,b Donatella Tondi,b Giulio Rastellib and Maria Paola Costib,*
aDipartimento di Chimica, Universita’ degli Studi di Modena e Reggio Emilia, Via Campi 183 41100, Modena, Italy
bDipartimento di Scienze Farmaceutiche, Universita’ degli Studi di Modena e Reggio Emilia, Via Campi 183, 41100 Modena, Italy
c
`
Istituto Chimico Farmaceutico, Universita degli Studi di Milano, Viale Abruzzi 42, 20131 Milan, Italy
Received 3 July2002; revised 1 October 2002; accepted 18 October 2002
Abstract—Thymidylate synthase (TS) (EC 2.1.1.45), an enzyme involved in the DNA synthesis of both prokaryotic and eukaryotic
cells, is a potential target for the development of anticancer and antinfective agents. Recently, we described a series of phthalein and
naphthalein derivatives as TS inhibitors. These compounds have structures unrelated to the folate (Non-Analogue Antifolate
Inhibitors, NAAIs) and were selective for the bacterial versus the human TS (hTS). In particular, halogen-substituted molecules
were the most interesting. In the present paper the halogen derivatives of variously substituted 3,3-bis(4-hydroxyphenyl)-1H,3H-
naphtho[2,3-c]furan-1-one (1–5) and 3,3-bis(4-hydroxyphenyl)-1H,3H-naphtho[1,8-c,d]pyran-1-one (6–14) were synthesized to
investigate the biological effect of halogen substitution on the inhibition and selectivityfor the TS enzymes. Conformational prop-
erties of the naphthalein series were explored in order to highlight possible differences between molecules that show species-specific
biological profile with respect to non species-specific ones. With this aim, the conformational properties of the synthesized com-
pounds were investigated byNMR, in various solvents and at different temperatures, and bycomputational analysis. The apparent
inhibition constants (Ki) for Lactobacillus casei TS (LcTS) were found to range from 0.7 to 7.0 mM, with the exception of the weakly
active iodo-derivatives (4, 10, 13); all] the compounds were poorlyactive against hTS. The di-halogenated compounds 7, 8, 14
showed the highest specificitytowards LcTS, their specificityindex (SI) ranging between 40 and
>558. The di-halogenated
1,8-naphthalein derivatives (7–10) exhibited different conformational properties with respect to the tetra-haloderivatives. Though a
clear explanation for the observed specificitybymeans of conformational analysis is difficult to find, some interesting confor-
mational effects are discussed in the context of selective recognition of the compounds investigated bythe LcTS enzyme.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
via Thymidine Kinase (TK) (salvage pathway), the
reaction catalyzed by TS is the rate-limiting step in the
DNA synthesis, for it is the sole de novo pathway for
the synthesis of thymidine triphosphate (dTTP). Inhibi-
tion of TS would therefore prevent cell multiplication.
Thymidylate synthase (TS) (EC 2.1.1.45), an enzyme
involved in the DNA synthesis of both prokaryotic and
eukaryotic cells, is a potential target for the develop-
ment of antibacterial, antimycotic, anticancer and,
recently, antiviral agents.1À6 TS catalyzes the conversion
of deoxyuridine monophosphate (dUMP) to deoxy-
thymidine monophosphate (dTMP) by a reductive
methylation involving N5,N10-methylentetrahydrofolate
(mTHF) as cofactor.1,2 In the absence of dTMP formation
Compounds structurallyrelated to the substrate
(dUMP) or to the cofactor (mTHF) have been devel-
oped as classical antimetabolites.3 A huge number of
derivatives have been tested and some are on clinical
trial or on the market for anticancer therapy, such as
FdUMP (5-fluoro-20-deoxyuridine-50-monophosphate)
and the folate analogues ZD1694, BW1843U89 (Fig. 1).
These compounds prevent the substrate or the cofactor
from binding the TS active site and, as a consequence,
theyblock the snythesis of dTMP. Nevertheless, the
*Corresponding author. Tel.: +39-059-205-5134; fax: +39-059-205-
5131; e-mail: costimp@mail.unimo.it
yPresent address: Spin, Via Tamagno 3, Rubiera, 42048 Reggio Emilia,
Italy.
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00541-2

952
S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
excess of free TS, due to failures of feedback regulation,
and its capabilityto bind mRNA can induce more complex
long-term reactions.7À9 Owing to the structural similarity
between classical antimetabolites and natural metabolites,
these compounds are generallypoorlyselective and give
rise to resistance through different mechanisms.10,11
between TS from various sources, as recentlydemon-
strated byX-raycrystallographyof Lactobacillus casei
TS (LcTS),14 Escherichia coli TS (EcTS),15Pneumocistis
carinii TS (PcTS)16 and human TS (hTS),17,18 either as
apo-enzymes and complexes with inhibitors.3 A com-
parative analysis of the mentioned structures suggests
that the folate analogue inhibitors bind in the folate
binding site and the binding mode is highlyconserved
among TS complexes3 from different species.
Recently, Non Antifolate Analogue Inhibitors (NAAIs)
compounds not structurallyrelated to the folate cofactor
but possessing antifolate activity, have been receiving
increasing attention. In an attempt to overcome resis-
tance, especiallythat arising from the metabolic path-
wayactivation tpyical of the folate analogues, many
inhibitors with non-classical folate structure have been
developed, and some of them, such as AG331 (Fig. 1)
and AG33712,13 have been tested in clinical trials.
Recently, TS has become an interesting target for anti-
infective drugs. That is due to structural differences
Figure 2. Phthalein and naphthalein derivatives.4
Figure 1. N5,N10-Methylentetrahydrofolate, folate analogues and
NAAI of Thymidylate synthase of recent development in anticancer
chemotherapy.
Figure 3. Chemical structure of the compounds studied in the present
paper.

S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
953
In a previous work,4 we studied a series of phthalein
and naphthalein derivatives as NAAIs, (A,B, Fig. 2).
These compounds have structures not related to the
folate but displaya competitive inhibition pattern with
respect to the cofactor. In particular, compound 3,3-
bis(3-chloro-4-hydroxyphenyl) - 1H,3H - naphtho[1,8 -
c,d]pyran-1-one (R=Cl, R0=H, 8, Fig. 3) was 60 times
more active for LcTS with respect to hTS.
phenyl)-1H,3H-naphtho[1,8-c,d]pyran-1-one (6–14) were
synthesized (Fig. 3) and the biological role of the halo-
gen substitution on the inhibitoryprofile towards TS
enzymes was investigated. Conformational properties of
the naphthalein derivatives synthesized were explored to
highlight possible differences in molecular properties
between molecules showing species-specific and non spe-
cies-specific biological profiles. Studies on the conforma-
tional properties of the naphthalein derivatives have been
performed byNMR, in various solvents and at different
temperatures, and bycomputational analysis. Inhibition
assays versus hTS and LcTS were performed for all the
compounds, and their specificitywas estimated.
The X-raystructure of the complex between compound
8 and LcTS showed that it binds at the entrance of the
active site pocket and makes veryfew specific hydrogen
bonds with the enzyme.4 In particular, the hydrogen
bonds between the phenol hydroxyl groups and Trp85,
His106 and Val316 are shown in Figure 4. Compound 8
shows some interactions with an area of the enzyme
surface called small domain; this area, formed byresi-
dues 90–139, is typical of LcTS and other pathogens TS
such as Staphylococcus aureus and Enterobacter faecalis.
The abilityof the compound 8 to bind to the ‘small
domain’ of TS, which is not present in the human
enzyme (Fig. 4), has been suggested to explain the spe-
cificityof compound 8 and its derivatives. However, the
molecular bases of the observed species-specificity
remain unclear, at the moment. Therefore better-
designed naphthalein derivatives have to be synthesized
to better explore structure–specificityrelationships.
Chemistry
ortho-Halogen substituted naphthaleins can be obtained
either through Friedel–Crafts reaction between the
naphthalidic compound (2,3-naphthalic anhydride and
1,8-naphthalic anhydride) and the required substituted
phenol or through direct halogenation of the appropriate
naphthalein compound, namely[3,3-bis(4-hdyrox-y
phenyl)-1H,3H-naphtho[2,3-c]furan-1-one (1) and 3,3 -
bis(4-hydroxyphenyl)-1H,3H-naphtho[1,8-c,d]pyran-1-one
(6)] bybromine, or iodine chloride. It should be noted
that in no case examples of substitutions on the naph-
thalene ring were evidenced.
In the present paper, two series of halogen derivatives of
variously substituted 3,3-bis(4-hydroxyphenyl)-1H,3H-
naphtho[2,3-c]furan-1-one (1–5) and 3,3-bis(4-hydroxy-
Compounds 1–3 (Scheme 1) were obtained byheating
the anhydride at 180 ꢀC for 5 h with the appropriate
Figure 4. Detail of the X-raycrystal structure of the binarycomplex Lactobacillus casei TS-compound 8. The small domain is colored in yellow.

954
S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
phenol (molar ratio 1:2), in the presence of a few drops
of sulphuric acid,19 and purifying the crude by silica-gel
chromatography. Compound 4 was obtained bystirring
1 and an excess iodine chloride in glacial acetic acid at
room temperature overnight.20 Compound 5 was prepared
bybromination of 1, using 4 equivalents of bromine in
ethanol overnight at room temperature21,22 (Scheme 1).
ubstituted at position 3 bytwo phenolic rings, is sensi-
tive to the average angle formed bythe phthalidic and
the aromatic ring planes. Moreover, we reported that
the resonances of chemicallyequivalent protons
belonging to the phenolic rings split at low temperature
as a result of conformational constrains due to steric
hindrance. These aspects were considered relevant also
in the present studyin order to investigate the con-
formational properties.
Compounds 6–8 were prepared byheating 1,8-naph-
thalic anhydride at 110–115 ꢀC for 3 days with the
appropriate phenol, in the presence of aluminum chlo-
ride.23,24 Treatment of 6 with excess bromine in di-
chloromethane at room temperature for 6 h gave a
mixture of 9 and 14,21 which were separated silica-gel
chromatography. Compound 14, the onlychiral mole-
cule of the series, was tested as a racemic mixture. The
tetra-bromo derivative 12 was prepared from 6 as
reported for 5, while the tetra-chloro derivative 11 was
obtained byheating 6 with chlorine in DMF at 60 ꢀC.
The iodo derivatives 10 and 13 were synthesized by
treating 6 with iodine chloride at room temperature or
with a mixture of iodine/mercuryacetate at 95 ꢀC,
respectively( Scheme 2).
Proton NMR spectra of compounds 1–14 in DMSO-d6 at
300 K showed small and meaningless differences in the d
of proton H4 (values ranging between 8.36 and 8.50ppm)
and onlyone signal for each couple of chemicallyequiva-
lent protons belonging to the phenolic rings. This beha-
viour was also observed for compounds 1–5, 6 and 11–13
in acetone-d6, at temperature ranging between 300 and
170 K (freezing point of the solutions). This demonstrates
that in all these cases, the two phenolic rings lie in the
same average conformational state in the NMR time scale
although, in some cases (see tetra-bromo derivative 5),
bulkysubstituents are present. As a consequence, it is
possible to suggest that the presence of the penta-atomic
lactonic ring does not promote anysignificant steric con-
strain at the phenolic rings that maintain unimpaired high
rotational freedom around the C3–C10 and C3–C17
bonds. On the contrary, the low temperature NMR spec-
tra of compounds 7–10 shown separate peaks for protons
H11, H14, H15, H18, H21 and H22 whose differences in
chemical shift increased with the lowering of the tem-
perature, as shown in Figure 5.
1H NMR Spectroscopy
Room temperature one-dimensional proton NMR
spectra of compounds 1–14, dissolved in DMSO-d6,
were fullyassigned on the basis of the chemical shifts
and the signal multiplicityanalsyis. In a few cases,
homonuclear scalar and dipolar correlations obtained,
respectively, by COSY and NOESY spectra were
required. All the data are summarized in Table 1.
To verifyif the polarityof the solvent could influence
the conformational properties of the molecules, CDCl3
and CS2 were used. Onlycompound 8 was soluble
enough to allow the collection of NMR spectra, so the
experiments were limited to this compound. Up to 203
In a previous paper,25 we reported that the chemical
shift (d) of phthalidic proton H4, of phthaleins dis-
Scheme 1.

S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
955
and 170 K, the freezing points in CDCl3 and CS2
respectively, the spectra did not displayanysplitting of
signals. This behavior, compared with that observed in
aceton-d6, suggests that the polarityof the solvent plays
a role in conformation. Since in acetone-d6, compound
8 behaves like all the other di-halogen derivatives 7, 9
and 10, one can reasonablysuppose compounds 7, 9
and 10 to behave in similar fashion in CDCl3 and CS2.
tion (o < <1) and signals resulted well detectable
although, in some cases, theyfall in crowded area of the
spectra (in compound 9 all the phenolic protons reso-
nate in only0.5 ppm of spectral width). In the case of
the tetra-iododerivative 10, all the possible NOEs were
observed: between protons H4-H11, H4-H18, H4-H14,
H4-H15 and between all the phenolic protons (Fig. 6).
It is worth noting that the splitting of oxydrilic protons
was also observed. On the basis of these data, a con-
formation for compound 10 can be drawn in which the
two phenolic rings appear as shown in Figure 7.
Low temperature NOESY spectra of 8–10, dissolved in
acetone-d6, were also recorded to gain deeper insight
into the conformational differences observed among the
di-halogen derivatives and all the other compounds. In
all cases onlypositive NOEs were detected, i.e., the
molecules maintained their extreme narrowing condi-
Although a complete set of NOEs was observed onlyfor
compound 10, it is reasonable to suppose that the other di-
halogen derivatives 7–9 also exhibit, at low temperature,
Scheme 2.

956
S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
Table 1. NMR data of proton spectra of compounds 1—14
p.n.
Chemical shift (ppm)
1
2
3
4
5
6
7
H4
H5
H6
H7
H8
H9
H11
H12
H14
H15
H16
H18
H19
H21
H22
H23
8.36
8.23
7.83
7.77
8.34
8.73
7.24
6.86
=H12
=H11
9.72
=H11
=H12
=H14
=H15
=H16
s
d
td
td
d
s
dt
dt
8.46
8.25
7.85
7.78
8.35
8.73
7.17
—
7.06
7.11
10.10
=H11
—
s
d
td
td
d
8.50
8.25
7.86
7.79
8.35
8.77
7.34
—
s
d
td
td
d
8.45
8.25
7.85
7.78
8.35
8.76
7.65
—
=12
=11
10.73
=H11
=H12
=H14
=H15
=H16
s
d
t
t
d
s
8.19
8.22
7.84
7.87
8.32
8.75
7.89
—
—
=11
—
=H11
—
—
=H15
—
s
d
t
t
d
s
7.17
7.75
8.19
8.47
7.84
8.36
6.96
6.81
7.04
6.99
9.71
=H11
=H12
=H14
=H15
=H16
dd
dd
dd
dd
dd
dd
d
d
d
dd
s
7.24 dd
7.77 dd
8.23 dd
8.49 dd
7.86 dd
8.39 dd
s
dd
s
d
d
d
6.98
—
o
t
dd
s
7.09
7.26
d
dd
s
6.79
=11
9.83
=H11
—
=H14
=H15
=H16
s
10.61
=H11
=H12
=H14
=H15
=H16
s
—
=H15
=H16
8
9
10
11
12
13
14
H4
H5
H6
H7
H8
H9
H11
H12
H14
H15
H16
H18
H19
H21
H22
H23
7.24
7.78
8.24
8.50
7.87
8.39
7.10
—
dd
dd
dd
dd
dd
dd
d
7.23
7.78
8.24
8.50
7.87
8.40
7.25
—
dd
dd
dd
dd
dd
dd
d
7.21
7.78
8.23
8.49
7.87
8.39
7.46
—
dd
dd
dd
dd
dd
dd
dd
7.36
7.80
8.27
8.52
7.89
8.42
7.26
—
d
t
d
d
t
7.33
7.89
8.28
8.53
7.81
8.43
7.31
—
dd
dd
d
dd
dd
dd
s
7.35
7.95
8.32
8.57
7.86
8.46
7.54
—
d
t
d
d
t
7.20 dd
7.77 dd
8.22 dd
8.48 dd
7.86 dd
8.38 dd
d
s
d
s
7.24
—
d
6.97
7.05
d
dd
s
7.04
6.99
d
dd
s
. . .
6.96
d
d
s
—
—
—
7.00
o
=H11
10.70
=H11
—
=H11
10.50
=H11
—
—
=H15
=H16
=H11
10.06
=H11
—
—
=H15
=H16
=H14
9.76
6.82
10.62
=H11
=H12
=H14
=H15
=H16
10.70
=H11
=H12
=H14
=H15
=H16
10.70
=H11
=H12
=H14
=H15
=H16
s
s
s
d
o
7.00
—
=H19
=H18
10.65
=H15
=H16
Correlations
s
1)
6)
8)
C: H5–H6;H6–H7;H7–H8;H11–H12–N: H4–H11;H4–H5; H8–H9
C: H4–H5;H4–H6;H5–H6;H7–H8; H7–H9;H8–H9;H11–H12; N: H4–(H11,H18); H4–(H15,H22) ; H6–H7
C:H4–H5;H4–H6;H5–H6;H7–H8;H7–H9;H8–H9;H11–H14; H11–H15;H14–H15; N: H4–H11;H4–H15;H6–H7
10) C: H4–H5;H4–H6;H5–H6;H7–H8;H7–H9;H8–H9,H11–H14, H11–H15
12) C: H4–H5; H5–H6; H7–H8; H8–H9; N: H6–H7; H4–H11
Chemical shifts (d), spin multiplicities (s=singlet, d=doublet, t=triplet), scalar and dipolar correlations (C=correlation in 2D-COSY spectra;
N=correlation in 2D-NOESY spectra; Hn–Hm means correlation between proton Hn and proton Hm) of compounds 1–14 dissolved in DMSO-d6.
All spectra were recorded at room temperature. All chemical shifts are referred to the residual signal of DMSO-d6 fixed at 2.6 ppm.
the conformation found for 10 (data not shown),
because the splitting pattern of the signals is verysimilar
to that observed for 10.
symmetry/asymmetry of the substitution, as confirmed
bythe different behavior of di-halogen with respect to
the tetra-halogen derivatives.
Finally, compounds 6 and 14 exhibited neither splitting
nor broadening of the resonance of the phenolic protons
at low temperature in acetone-d6. This proofs that the
conformation of compounds 1–14 also depends on the
Quantum chemical calculations
In order to investigate the conformational properties
and confirm the interpretation of the NMR results,

S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
957
Figure 5. 1H NMR spectra of compounds 7 (left panel) and 10 (right panel) in acetone-d6 at different temperatures. The splitting protons are indi-
cated as follows: a=H14, b=H11, g=H15. With decreasing temperature, the splitting of the proton signal of the phenol rings is evident.
conformational analysis was performed on a selected
set of compounds (6, 8, 9, 11 and 12). These com-
pounds were chosen to investigate the effects of the
halogens on conformational barriers in symmetrically-
and asymmetrically-substituted compounds. Con-
formations were investigated byrotating the phenolic
rings in steps of 20ꢀ, with complete geometryoptimiza-
tion at each step, using the PM3Hamiltonian.
pounds have low rotational barriers (close to 1 Kcal/
mol), in agreement with the high conformational freedom
of the two phenolic rings demonstrated byNMR experi-
ments. Interestingly, one significant difference was found
between symmetrically- and asymmetrically-substituted
compounds. The symmetrically-substituted compounds
6, 11 and 12, have almost identical conformational
profiles. Therefore, these compounds exhibit almost
identical conformational properties regardless of substitu-
tion and of the nature of the halogen substituent (chloro or
bromo). The asymmetrically-substituted compounds 8 and
9 still have verysimilar conformational profiles, thus
The rotational profiles obtained for these compounds
are graphicallyreported and superimposed in Figure 8.
Inspection of these profiles shows that all the com-

958
S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
Figure 6. 1H/2D-NOESY spectrum of compound 10 recorded at 170 K in acetone-d6 using 200 ms of mixing time.
affording further evidence that conformational properties
are not significantlyaffected bythe tpye of halogen
substituent, but theyexhibit somewhat higher rotational
barriers in the 150–270ꢀ range of the dihedral angle with
respect to the symmetrically substituted compounds.
This increase in rotational barrier for compounds 8 and
9 nicelysupports the NMR experimental evidence that
all the di-halogen derivatives assume a well-defined
conformation at verylow temperature, at variance with
the non-substituted (6) and symmetrically substituted
(11, 12) compounds. In conclusion, the calculated con-
formational profiles are in agreement and support the
NMR results; theyalso suggest that conformation, in
this type of molecules, is more dependent on symmetry/
asymmetry of substitution rather than on the nature of
substitution. Similar conclusions have also been reached
bycalculating the dipole moments of the compounds as
a function of the conformational angle (Fig. 9). Even for
the dipoles, identical trends were observed for the non-
substituted compound 6 and the symmetrically-sub-
stituted compounds 11 and 12, regardless of the sub-
stituent, while a completelydifferent trend was observed
for the asymmetrically-substituted compounds 8 and 9.
Enzyme assays
All the compounds were evaluated as inhibitors of L.
casei TS and human TS (Table 2) and IC50 values were
determined. In order to studytheir inhibition pattern,
apparent inhibition constant values (Ki) were measured
against LcTS (Table 2); the specificityindexes (SI) were
calculated bythe ratio IC 50hTS/IC50LcTS (Table 2).
All the compounds proved to be competitive inhibitors
of LcTS with respect to the folate cofactor. 2,3-Naph-
thalein derivatives (1–5) are generallyless active
towards LcTS than the 1,8-naphthalein derivatives (6–
14). Their apparent Ki values range from 1.5 mM (2), to
7.0 mM (1), while no inhibition is seen for compound 4
at 23 mM, which corresponds to the solubilitylimit for
Figure 7. Compound 10 is reported with the conformation supported
by 1H/2D-NOESY spectrum at 170 K in acetone-d6 (see Fig. 6). The
protons referring to NOEs are indicated byarrows.

S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
959
this compound. Their IC50 values at high concentrations
towards hTS were also measured; theyrange from 6.0 to
41 mM; the iodo-derivative (4) is inactive at the concen-
tration measured (23 mM). The most selective com-
pound is the ortho-fluorine derivative 2, which shows a
specificityindex (SI) of 4.
Table 2. Inhibitoryactivityof compounds 1–14 against LcTS and
hTS. Specificityindexes (SI) are given bythe ratio of IC 50 for hTS and
IC50 for LcTS
Compd
LcTS
Ki (mM)
LcTS IC50
(mM)
hTS IC50
(mM)
SI
1
2
3
4
5
6b
7
8b
9
10
11
12
13
14
CB3717
7.0
1.5
1.6
NIa
5.3
2.8
1.6b
0.7
6.8
31
2.0
0.9
—
1.0
0.06
20
7.2
5.3
—
20
27
7.2
1.3
32
111
7.4
4.3
NIa
5.4
0.09
40
29
2.0
4.0
1.1
Inhibition data analysis for the 1,8-naphthalidic di-
halogen series (7–10,14) (Table 2) showed that the best
inhibitor towards LcTS is compound 8 (Ki=0.7 mM);
compound 12 and the chiral compound 14 are also good
inhibitors with Ki values of 0.9 and 1.0 mM, respectively.
All the other compounds had higher Ki values ranging
from 1.6 to 6.8 mM; the iodo derivative 10 exhibited the
lowest inhibitoryactivity( Ki=31 mM). All the com-
pounds are weak or veryweak inhibitors of hTS, with
IC50 between 6 and well above 4 mM. Compounds 7, 10
and 14 did not demonstrate anyinhibitoryactivityat
the solubilitylimit ( Table 2); the IC50 values were
therefore calculated assuming 2% of inhibitoryactivity
to be present at the solubilitylimit owing to the experi-
mental error.4 This approximation clearlyunder-
estimates the true IC50 value of the compounds.
6.0
NIa
41
—
2.1
2.5
>558c
67
>4018c
52
47
40
1.4
>12c
7.6
6.1
—
>298c
0.34
>1317c
57
32
NIa
>1627c
0.03
aNI, no inhibition at 23 mM, corresponding to the solubilitylimit of
the compound.
bRef 4.
cCalculated supposing 2% inhibition at the maximum solubilitycon-
centration, 7, 82 mM; 10, 27 mM; 14, 33 mM, respectively.
The most specific inhibitors are 7, 8 and 14. The speci-
ficityrange that we were able to measure was higher
than 558 for 7, 298 for 14 and 40 for 8 (Table 2).
The 1,8-naphthalidic tetra-halogen derivatives (11, 12)
are good inhibitors of LcTS but also inhibit hTS, hence
the SI value is around 7 for both the compounds. The
iodo derivative 13 is inactive at the solubilitylimit
(23 mM). The well-known TS inhibitors with classical
folate structure, CB371726 (5,8-dideaza folic acid), did not
demonstrate anyspecies-specificityamong the two species
considered (LcTS and hTS), displaying an SI of 0.34.
Discussion
1
The conformational studies performed with H NMR
showed that compounds 1–5 behave similarlyto that of
25
the previouslyinvestigated phthalein derivatives, with
high rotational freedom at room and low temperature.
On the contrary, differences were noted within the series
of compounds 6–14. The derivatives having two halo-
gens on both phenolic rings in ortho position to the
hydroxyl group (11–13) still exhibit high rotational
freedom up to 170 K, regardless of the nature of the
halogen atom. In contrast, the analogues having only
one halogen substituent on the phenolic ring (7–10)
assume a well-defined conformation when the tempera-
ture is lowered to 179 K. This conformation is similar to
the one observed from X-raycrystallographyof similar
compounds.27 In that conformation the phenolic rings
are on perpendicular planes with the chlorine atom on
opposite direction. It is noteworthythat, at least in the
case of compound 8, a well-defined conformation could
not be observed in less polar solvents such as CDCl3
and CS2. This compound was the onlyone soluble in
apolar solvents and it was not possible to perform a
similar experiment for the other di-halogen derivatives
(7, 9, 10). These data suggest that the 1,8-di-halogen
substituted compound, 8, likelyundergoes specific
interactions with the polar solvent leading to a well-
defined conformation; this effect is not seen in apolar
solvents. Since the NMR results proved to be independent
Figure 8. Rotational profile for compounds 6, 8, 9, 11, 12. The solid
line is obtained from interpolation of the calculation results every20 ꢀ,
as described in Experimental.
Figure 9. Dipole moment profiles for compounds 6, 8, 9, 11, 12.

960
S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
from the nature of the halogen, it is likelythat the other
di-halogen derivatives (7, 9, 10) would also behave
similarlyin apolar solvents.
formed at room and low temperature in different solvents
and quantum chemical calculations were run to study
their conformational properties. Their biological activity
profile against LcTS and hTS was then determined.
In agreement with NMR, quantum-chemical calcu-
lations suggest that the symmetrically-substituted com-
pounds 6, 11 and 12, exhibit almost identical
conformational properties regardless of substitution
and of the nature of the halogen substituent (chloro or
bromo). The asymmetrically-substituted compounds 8
and 9 still have verysimilar conformational profiles,
thus affording further evidence that conformational
properties are not significantlyaffected bythe type of
halogen substituent. The low conformational barriers
are in agreement with the high conformational freedom
observed byNMR. Interestingl,y calculations suggest
that conformation is more dependent on symmetry/
asymmetry of substitution rather than on the nature of
substitution. Similar conclusions have also been reached
bycalculating the dipole moments of the compounds as
a function of the conformational angle.
We found an agreement between the conformational
behaviour of the di-halogen derivatives in solution as
determined by H NMR analysis and the quantum che-
1
mical calculations performed on the conformational
properties of the molecules. Low-temperature NMR in
acetone showed that compounds 7–10 with onlyone
halogen on the phenol ring have a preferred confor-
mation corresponding to the low-energyconformation,
even though in NMR terms it is an average conforma-
tion. Correspondingly, conformational analysis based
on quantum chemical calculations demonstrated that no
preferred conformation exists for the di-halogen deriva-
tives with respect to the tetra-halogen substituted deri-
vatives and that rotational freedom is allowed at room
temperature. Thus, these compounds exhibit almost
identical conformational properties regardless of the
substitution and the nature of the halogen substituent
(chloro or bromo). The asymmetrically substituted com-
pounds 8 and 9 still displayverysimilar conformational
profiles. This adds further evidence that the type of
halogen does not significantlyaffect conformational
properties. The compounds exhibit somewhat higher
rotational barriers in the 150–270ꢀ range of the dihedral
angle with respect to the symmetrically substituted
compounds.
Considerations on structure–activityand structure–spe-
cificityrelationships show that the 2,3-naphthalein ser-
ies (1–5) is less interesting than the 1,8-naphthalein
series (6–14). In the first series the Ki ranged between 1.5
and 7.0 mM, and no species-specificitycould be observed
(selectivityindex, SI=4, being the best result). The sec-
ond series is more interesting in terms of species-specifi-
cityand, in particular, the di-halogen derivatives ( 7–10).
Their Ki values range between 0.7 and 31 mM, and the SI
values were 100 times higher with respect to the SI
values found in the first series, the SI being >558 for
compound 7. One exception is the chiral compound 14,
that shows a Ki higher than 298.
This suggests that the behaviour of the compound 8 in
solution at low temperature is governed byspecific
interaction with the solvent. In fact, unlike the other di-
halogens derivatives, compound 8 is soluble in two sol-
vents with different polarity, namely, acetone and car-
bon disulfide.
The findings that di- and tetra-halogenated compounds
displaydifferent conformational behaviour in polar/
apolar solvents, and that their dipole moments follow
opposite trends as a function of conformation, suggest
that these two classes of compounds could have differ-
ent molecular states also in a biological environment. In
a previous work,4 it was proposed that the molecules
showing a species-specific inhibitoryprofile bind to non-
conserved binding site such as the small domain in
LcTS, as showed bythe X-raystructure of the binary
complex LcTS-a156. The different molecular state of the
present molecules might influence the binding con-
formation, shifting the molecule from a well-conserved
site to a non-conserved one, therefore contributing to
the observed specificity.
In acetone, a highlypolar solvent, preferred conforma-
tions were found at low temperature. On the other
hand, these preferred conformations were not detected
in the apolar solvent.
Structure–specificityanalysis was difficult to be assessed
because factors other than conformation can playa
role. It is also true that the conformational analysis was
performed in conditions far from those found in the
biological environment. Nevertheless, it was clearly
highlighted that some molecules (7–10) have confor-
mational spaces that are different from the others. We
suggest that those molecules with different confor-
mational properties can bind to different sub-sites in the
recognition process, thus expressing a different bio-
logical activityprofile.
Obviously, conformational behavior is not the only
factor determining activityand specificity. The impor-
tance of one ortho substituent can also affects the
polarity of the phenolic hydroxyls and influences the
interactions with polar aminoacids of the active site.
This conclusion agrees with the results of a previous
studyon the binding mode of naphthalein derivatives:
the folate pocket surface is a highlyplastic area in the
LcTS enzyme where the naphthalein derivatives can
easilybind in more than one mode (multiple binding
mode). A deeper analysis of the mode of interaction of
the mentioned compounds is required to clearlyexplain
the observed specificity.
Conclusions
In the present study, some ortho-halogen di- and tetra-
substituted derivatives were synthesized, NMR were per-

S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
961
3,3-Bis(4-hydroxyphenyl)- (6), 3,3-bis(3-fluoro-4-hydroxy-
phenyl)- (7) and 3,3-bis(3-chloro-4-hydroxyphenyl)-
1H,3H-naphtho[1,8-c,d]pyran-1-one (8). Anhydrous alu-
minum chloride (0.02 mol) was added portionwise to a
mixture formed by1,8-naphthalic anhydride (0.01 mol)
and the appropriate phenol (0.02 mol) in anhydrous
s-tetrachloroethane (100 mL); the suspension was stirred
for three days at 100–115 ꢀC. The hot mixture was then
poured onto ice and dichloromethane (200 mL) was
added. After stirring for 0.5 h the precipitate was filtered
off and the organic layer separated and dried over
sodium sulfate. After evaporation of the solvent, the
residue was purified bysilica gel chromatographyelut-
ing with dichloromethane/methanol (95/5 v/v) to give,
first, the unreacted anhydride, followed by an uni-
dentified compound and, finally, the desired product.
The final compounds were crystallized from dichloro-
methane as a pale-brown solid.
Table 3. Elemental analyses of compounds 1–14
Compd
Calcd %
Found %
C
H
X
C
H
X
1
2
3
4
5
6
7
8
78.25
71.29
65.92
46.48
42.14
78.25
71.29
65.92
54.78
46.48
56.95
42.14
33.06
64.45
4.38
78.37
70.98
66.01
46.37
42.23
78.37
71.09
66.01
54.67
46.52
57.09
42.28
33.11
64.51
4.11
3.49
3.23
2.28
1.77
4.38
3.49
3.23
2.68
2.28
2.39
1.77
1.39
3.38
9.40 (F)
16.22 (Cl)
40.93 (I)
3.56
3.11
1.99
1.86
4.44
3.27
3.41
2.79
2.32
2.24
1.95
1.29
3.46
16.09 (Cl)
41.05 (I)
46.55 (Br)
46.73 (Br)
9.40 (F)
16.22 (Cl)
30.37 (Br)
40.93 (I)
28.02 (Cl)
46.73 (Br)
58.22 (I)
16.35
30.51 (Br)
41.05
27.95 (Cl)
46.03 (Br)
58.34 (I)
17.54 (Br)
9
10
11
12
13
14
17.86 (Br)
6: yield 18%; mp 265–267 ꢀC. 7: yield 32%; mp 248–
250 ꢀC. 8: yield 38%; mp 190–195 ꢀC.
Experimental
Compounds commerciallyavailable were used without
further purification. Br2 (Merck), ICl (Aldrich), mercury
acetate (J.T. Baker), I2 (Acros) were bought as indicated
in parentheses. Melting points were determined on a
Buchi 510 apparatus and are uncorrected. Thin-layer
chromatography(TLC) was carried out with pre-coated
silica gel F254 plates (Merck).
3,3-Bis(3,5-dibromo-4-hydroxyphenyl)-1H,3H-naphtho
[2,3-c]furan-1-one (5) and 3,3-bis(3,5-dibromo-4-hydroxy-
phenyl)-1H,3H-naphtho[1,8-c,d]pyran-1-one (12). Bromine
(0.60 mmol) was added dropwise at room temperature
to compound 1 or 6 (0.14 mmol) dissolved in 2 mL of
ethanol and the reaction mixture was stirred overnight.
After evaporation of the solvent, the residue was puri-
fied bysilica gel chromatographyeluting with ccylo-
hexane/ethyl acetate (70/40 v/v). Crystallization from
ethylacetate gave the desired compounds as a yellowish
crystalline solid.
The IR spectra were recorded in Nujol suspension on a
Perkin-Elmer Spectrophotometer Mod 681. Silica Gel
60 (Merck 70–230 mesh) was used for purification of the
synthesized compounds using the solvent system indi-
cated in the section dealing with synthesis.
5: yield 50%; mp 155–158 ꢀC. 12: yield 65%; mp 257–
260 ꢀC.
Microanalysis was carried out by Rossella Gallesi in the
Microanalysis Laboratory of Dipartimento di Scienze
Farmaceutiche (Universita di Modena e Reggio Emilia).
The results (Table 3) were within Æ0.4% of the theore-
tical value and the structure of the compounds were
consistent with their spectroscopic data.
3,3-Bis(3-bromo-4-hydroxyphenyl)-1H,3H-naphtho[1,8-c,d]
pyran-1-one (9) and 3-(3-bromo-4-hydroxyphenyl)-3-(4-
hydroxyphenyl)-1H,3H-naphtho[1,8-c,d]pyran-1-one (14).
Bromine (0.27 mmol), dissolved in dichloromethane, was
added dropwise to a stirred solution of 6 (0.14 mmol) in
3 mL of dichloromethane at room temperature. After
stirring for 5 h at rt the solvent was evaporated and the
residue purified bysilica gel chromatography(dichloro-
methane/methanol 98/2 v/v). 14 was eluted as the first
run followed by 9. Compounds were crystallized from
ethyl acetate as yellowish crystalline solids.
Chemistry
3,3-Bis(4-hydroxyphenyl)- (1), 3,3-bis(3-fluoro-4-hydroxy-
phenyl)-(2) and 3,3-bis(3-chloro-4-hydroxyphenyl)-3H-
naphtho[2,3-c]furan-1-one (3). General procedure. A
mixture formed by2,3-naphthalic anhydride (0.01 mol),
the appropriate phenol (0.02 mol) and a few drops of
H2SO4 concd was heated at 180–190 ꢀC for 5 h with
constant stirring. After cooling, water (30 mL) was
added to the reaction mixture; the new mixture was then
extracted with dichloromethane (3 Â 20 mL). The
organic phase was dried on anhydrous Na2SO4 and
concentrated in vacuum. The solid residue was purified
bysilica gel chromatographywith dichoromethane/
methanol (95/5 v/v) as eluent. The first collected com-
pound was the unreacted anhydride, followed by the
desired compound. The final compounds crystallized
from dichloromethane as a pale-brown powder.
9: yield 28%; mp 98–100 ꢀC. 14: yield 25%; mp 228–
230 ꢀC.
3,3-Bis(3-iodo-4-hydroxyphenyl)-3H-naphtho[2,3-c]furan-
1-one (4) and 3,3-bis(3-iodo-4-hydroxyphenyl)-1H,3H-
naphtho[1,8-c,d]pyran-1-one (10). Iodine monochloride
(0.61 mmol), dissolved in 1 mL of glacial acetic acid, was
added to a mixture of 1 or 6 (0.14 mmol) suspended in
2.5 mL of glacial acetic acid, and the mixture was stirred
at room temperature overnight. After evaporation of
the solvent, the residue was purified bysilica gel
chromatography(dichloromethane/methanol 98/2 v/v).
Compounds were crystallized from ethyl acetate as
brown crystalline solids.
1: yield 35%; mp 298–300 ꢀC. 2: yield 38%; mp 138–
140 ꢀC. 3: yield 40%; mp 200–203 ꢀC.

962
S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
4: yield 59%; mp 230–232 ꢀC. 10: yield 71%; mp 143–
Conformational analysis was performed using the
MOPAC33 program in the PM3 parameterization.34
The rotational profiles were obtained byrotating one of
the two substituted phenyl rings in steps of 20ꢀ, with all
other geometric parameters completelyoptimized.
145 ꢀC.
3,3-Bis(3,5-diiodo-4-hydroxyphenyl)-1H,3H-naphtho[1,8-
c,d]pyran-1-one (13). Iodine (0.54 mmol) was added
portionwise to a stirred mixture of 6 (0.14 mmol) and
mercuryacetate (0.54 mmol) in 2 mL of glacial acetic
acid heated to 95 ꢀC; the mixture was allowed to react
under stirring at the same temperature for 30 min. After
cooling, the mixture was diluted with water and extrac-
ted with ethyl ether. The organic phase was dried over
anhydrous Na2SO4, evaporated and the residue purified
bysilica gel chromatography(dichloromethane/metha-
nol 98/2 v/v). 13: yield 8.5; mp 160–163 ꢀC (ethyl ace-
tate, dec).
All calculations were performed on Silicon Graphics O2
workstations.
Enzyme assay
Plasmids that express L. casei TS in the ThyÀ Escher-
35
ichia coli strain w2913 have been previouslydescribed.
The enzyme was purified by column chromatography
using phosphocellulose (P11, Biorad) and hydroxy-
apatite ( HAP, Biorad) resin, with phosphate buffer as
eluent.35,36 Human Thymidylate synthase (hTS) was
purified, as reported, through affinitycolumn chroma-
tography.37 The enzyme preparations were >95%
homogeneous, as shown by SDS-polyacrylammide gel
electrophoresis. The LcTS purified enzyme was stored
at À80ꢀ C in 10 mM phosphate buffer, pH 7.0,
0.1 mM EDTA until use. hTS was used immediately
after purification.
3,3-Bis(3,5-dichloro-4-hydroxyphenyl)-1H,3H-naphtho
[1,8-c,d]pyran-1-one (11). A solution of 6 (2.72 mmol) in
anhydrous DMF (10 mL) was saturated with chlorine,
stirred at room temperature for 60 min and then
decomposed with ice. The precipitate was crystallized by
acetone/petroleum as a pale brow solid. Yield 32%; mp
95–100 ꢀC.
NMR spectra
Enzymatic assays were run with a Beckman DU-640 UV-
Vis spectrophotometer, equipped with multicell system
thermostated with a Haake F3C circulating bath.
All NMR measurements were performed on a Bruker
AMX 400 WB spectrometer operating at 9.395 Tesla.
Spectra were performed on 16 mM samples dissolved in
DMSO-d6 for analysis at 303 K, or in acetone-d6,
CDCl3 and CS2 for studies at low temperatures. 1D
proton, carbon-13 and carbon-13 DEPT spectra were
obtained byoptimizing the spectral width and the digi-
tal resolution. No weighting of the FIDs was performed
before the Fourier transformation.
The activity of Thymidylate synthase was determined
spectrophotometricallybysteady-state kinetic analysis,
following the increasing absorbance at 340 nm due to
the oxidation reaction of N5,N10-methylenetetrahydro-
folate to dihydrofolate at 5,6 bond.38 Assays were per-
formed at 20 ꢀC in the standard assaybuffer, which
contained 50 mM TES (tris ethylaminosulfonic acid) at
pH 7.4, 25 mM MgCl2, 6.5 mM formaldehyde, 1 mM
EDTA (ethylendiaminetetracetic acid) and 75 mM 2-
mercapto-ethanol. 1 mL of reaction mixture consists of
standard TES buffer pH 7.4, dUMP 120 mM, 6-(R,S)-l-
CH2CH4-folate 180 mM, enzyme 0.07 mM.
2D-COSY28 spectra were recorded using 512 or 1024
time-domain points and 128 or 256 increments (F1
dimension), depending on the spectral width employed,
1 s of relaxation delayand 8 transients per increment.
Data were doubled in the F1 dimension byzero filling
and weighted bya sine-bell function in both dimensions
before magnitude-mode Fourier transformation.
The stock solutions of the inhibitors were prepared in
DMSO (dimethylsulfoxide) and kept at À20 ꢀC. Control
reactions were run in order to measure the effect of
DMSO on enzyme activity. DMSO never exceeded 10%
in the enzyme assay mixture.39
2D-NOESY29À32 spectra were obtained with TPPI
Phase Cycle, 256(F1)\1024(F2) or 512(F1)\1024(F2) data
points, depending on the spectral width used, 1 s of
relaxation delay, 8 scans per transient and using mixing
times of 200 ms. Data were weighted before Fourier
transformation using a sine square-bell function shifted
by _2 radiants in both dimensions.
TS inhibition studies were carried out under the condition
of the standard assay. Enzyme concentration was 0.07mM,
dUMP concentration was 120mM, while CH2H4folate
concentration was 60 mM for IC50 calculations and was
varied for Ki calculations. Reactions were initiated bythe
addition of the enzyme. Steady-state kinetic analysis of the
Lineweaver–Burk plot of the enzyme activity upon differ-
ent inhibitor concentrations was carried out byvarying the
folate concentration in the case of compounds 6 and 8. All
the compounds displayed a competitive inhibition pattern.
When the inhibition pattern was competitive, Ki values
were determined with the appropriate equation.40 All the
experiments were repeated at least three times, and the
standard errors from the linear least-squares fit of the
experimental data are within the 20%.
Quantum chemical calculations
A subset of compounds in Figure 3 was selected for
quantum chemical calculations: 6 (unsubstituted), 11
(tetra-chloro) and 12 (tetra-bromo) were chosen to
investigate the effects of halogen substitution on
conformational barriers in symmetrically substituted
compounds. The di-bromo (9) and di-chloro (8) derivatives
were chosen to investigate asymmetrical substitutions,
particularlyas the di-chloro compound is specific for
Lc TS.

S. Ghelli et al. / Bioorg. Med. Chem. 11 (2003) 951–963
963
Acknowledgements
14. Hardy, L. W.; Finer-Moore, J. S.; Montfort, W. R.; Jones,
M. O.; Santi, D. V.; Stroud, R. M. Science 1987, 35, 448.
15. Matthews, D. A.; Applet, K.; Oatley, S. J.; Xuong, N. H.
J. Mol. Biol. 1990, 214, 923.
16. Anderson, A. C.; O’Neil, R. H.; Delano, W. L.; Stroud,
R. M. Biochem. 1999, 38, 13829.
17. Sayre, P. H.; Finer-Moore, J. S.; Fritz, T. A.; Biermann,
D.; Gates, S. B.; MacKellar, W. C.; Patel, V. F.; Stroud, R. M.
J. Mol. Biol. 2001, 313, 813.
18. Phan, J.; Koli, S.; Minor, W.; Dunlap, R. B.; Berger, S. H.;
Lebioda, L. Biochem. 2001, 40, 1897.
This work was supported byCNR and MURST. We
thank CIGS (Centro Interdipartimentale Grandi Stru-
menti) of the Universityof Modena and Reggio Emilia
for supplying NMR spectrometer, CICAIA (Centro
Interdipartimentale di Calcolo Elettronico) for computing
facilities. We thank Professor Daniel V. Santi for supply-
ing the enzyme strains, Soragni Fabrizia for her excellent
technical assistance, and Renata Battini from the Bio-
chemistryDepartment of the Universityof Modena and
Reggio Emilia for helping with biochemical facilities.
19. Haubacher, M. J. Am. Chem. Soc. 1944, 66, 255.
20. Papa, D.; Ginsberg, H.; Lederman, Z.; De Camp, V.
J. Am. Chem. Soc. 1953, 75, 1107.
21. Adams, R.; Marvel, C. S. Org. Synth. Coll. I, 128.
22. Franchetti, P.; Cappellacci, L.; Sheikha, G. A.; Jayaram,
H. N.; Gurudutt, V. V.; Sint, T.; Schneider, B. P.; Jones,
W. D.; Goldstein, B. M.; Perra, G.; De Montis, A.; Loi, A. G.;
La Colla, P.; Grifantini, M. J. Med. Chem. 1997, 40, 1731.
23. Blicke, F. F.; Patelski, R. A. J. Am. Chem. Soc. 1938, 69,
2283.
24. Blicke, F. F.; Swisher, R. D. J. Am. Chem. Soc. 1934, 56, 902.
25. Ghelli, S.; Rastelli, G.; Barlocco, D.; Rinaldi, M.; Tondi, D.;
Pecorari, P.; Costi, M. P. Bioorg. Med. Chem. 1996, 4, 1783.
26. Jones, T. R.; Calvert, A. H.; Jackman, A. L.; Brown, S. J.;
Jones, M.; Harrap, K. R. Eur. J. Cancer 1981, 17, 11.
27. Costi, M. P.; Celentano, G.; Ghelli, S.; Barlocco, D.;
Tondi, D.; Pecorari, P. J. Het. Chem. 1999, 36, 1043.
28. Aue, W. P.; Batholdi, E.; Ernst, R. R. J. Chem. Phys.
1976, 64, 2229.
References and Notes
1. Santi, D. V.; Danenberg, P. V. Folate and Pteridines: Vol. 1,
Chemistryand Biochemistryof Folates. Blakely, R. L., Benkovic,
S. J., Eds.; John Wiley& Sons: New York, 1984; p. 345.
2. Douglas, K. T. Med. Res. 1987, 7, 441.
3. Costi, M. P. Med. Res. Rev. 1998, 1, 21.
4. 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. Biochem. 1999, 38, 1607.
5. Costi, P. M.; Rinaldi, M.; Tondi, D.; Pecorari, P.; Bar-
locco, 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. J. Med. Chem. 1999, 42, 2112.
6. Harrison, J. E.; Scott, M.; Hutchinson, J.; Thompson, R.
Eur. J. Biochem. 1995, 230, 511.
29. Jeener, J.; Meier, B. H.; Bachmann, P.; Ernst, R. R.
J. Chem. Phys. 1979, 71, 4546.
7. Ferguson, P. J.; DeMoor, J. M.; Vincent, M. D.; Kor-
opatnick, J. Br. J. Pharmacol. 2001, 134, 1437.
8. Chu, E.; Ju-J, F.; Schmitz, J. C. Antifolate Drugs Anti-
cancer Ther. 1999, 397.
9. DeMoor, J. M.; Vincent, M. D.; Collins, O. M.; Kor-
opatnick, J. Exp. Cell Res. 1998, 245, 11.
10. Van der Wilt, C. L.; Backus, H. H.; Smid, K.; Comijn, L.;
Veerman, G.; Wouters, D.; Voorn, D. A.; Priest, D. G.;
Bunni, M. A.; Mitchell, F.; Jackman, A. L.; Jansen, G.; Peters,
G. J. Cancer Res. 2001, 61, 3675.
11. Takemura, Y.; Kobayashi, H.; Miyachi, H. Hum. Cell
2001, 14, 185.
30. Redfield, A. G.; Kunz, S. D. J. Magn. Res. 1975, 19, 250.
31. Bodenhausen, G.; Vold, R. L.; Vold, R. R. J. Magn. Res.
1980, 37, 93.
32. Bauer, C.; Freeman, R.; Frenkiel, T.; Keeler, J.; Shanka
J. Magn. Res. 1984, 58, 442.
33. Stewart, J. J. P. MOPAC 6.0, QCPE 455.
34. Stewart, J. J. P. J. Comp. Chem. 1989, 10, 209.
35. Climie, S. C.; Santi, D. V. Proc. Natl. Acad. Sci. U.S.A.
1990, 87, 633.
36. Kealey, J. T.; Santi, D. V. Protein Expression Purif. 1992,
4, 380.
37. Davisson, V.; Sirawaraporn, W.; Santi, D. V. J. Biol.
Chem. 1989, 264, 9145.
38. Pogolotti, A. L.; Danenberg, P. U.; Santi, D. V. J. Med.
Chem. 1986, 29, 478.
39. Soragni, F., Costi, M. P. Unpublished.
12. Webber, S.; Bartlett, C. A.; Boritzki, T. J.; Hillard, J. A.;
Howland, E. F.; Johnston, A. L.; Kosa, M.; Margosiak, S. A.;
Morse, C. A.; Shetty, B. V. Cancer Chemother. Pharmacol.
1996, 37, 509.
13. Greer, J.; Erickson, J. W.; Baldwin, J. J.; Varney, M. D.
J. Med. Chem. 1994, 37, 1035.
40. Segel, I. H. Enzyme Kinetics; Wiley: New York, 1975, p.
109.