Imidazole analogues of fluoxetine, a novel class of anti-Candida agents

Article abstract of DOI:10.1021/jm049856v

Imidazole analogues of fluoxetine have been obtained by replacing the methylamino terminus of aminopropane chain with the imidazole ring. The newly designed imidazoles showed potent anti-Candida activity, superior to those of miconazole and other antifungal agents of clinical interest. 1-(4-Chlorophenyl)-l-(2,4-dichlorophenoxy)-3-(1H-imidazol-1-yl)propane (16), the most active among test imidazoles, Was about 2-fold more active and as much less cytotoxic than miconazole. High increase of activity was observed with methyl, nitro, fluorine, and chlorine (Cl > F > CH₃ > NO₂ > CF₃).

Full text of DOI:10.1021/jm049856v

3924
J . Med. Chem. 2004, 47, 3924-3926
Ch a r t 1. Commonly Used Antifungal Drugs
Im id a zole An a logu es of F lu oxetin e, a
Novel Cla ss of An ti-Ca n d id a Agen ts
Romano Silvestri,*,† Marino Artico,†
Giuseppe La Regina,† Alessandra Di Pasquali,†
Gabriella De Martino,† Felicia Diodata D’Auria,‡
Lucia Nencioni,‡ and Anna Teresa Palamara*,‡
Istituto Pasteur - Fondazione Cenci Bolognetti, Dipartimento
di Studi Farmaceutici, Universita` degli Studi di Roma
“La Sapienza”, P.le A. Moro 5, I-00185 Roma, Italy, and
Istituto di Microbiologia, Universita` degli Studi di Roma
“La Sapienza”, P.le A. Moro 5, I-00185 Roma, Italy
Received February 17, 2004
Abstr a ct: Imidazole analogues of fluoxetine have been ob-
tained by replacing the methylamino terminus of aminopro-
pane chain with the imidazole ring. The newly designed
imidazoles showed potent anti-Candida activity, superior to
those of miconazole and other antifungal agents of clinical
interest. 1-(4-Chlorophenyl)-1-(2,4-dichlorophenoxy)-3-(1H-imi-
dazol-1-yl)propane (16), the most active among test imidazoles,
was about 2-fold more active and as much less cytotoxic than
miconazole. High increase of activity was observed with
methyl, nitro, fluorine, and chlorine (Cl > F > CH3 > NO2 >
CF3).
Over the past two decades a significant increase in
fungal infections has been observed. Among these, the
widespread diffusion of topical and systemic infectious
diseases caused by the opportunistic pathogen Candida
albicans is often related to the use of broad-spectrum
antibiotics, immunosuppressive agents, anticancer, and
anti-AIDS drugs.1,2 One of the principal problems in the
treatment of C. albicans infections is the spread of
antifungal drug resistance mainly in patients chroni-
cally subjected to antimycotic therapy such as HIV-
infected people.3,4
Ch a r t 2
Treatment of Candida infections is based on azoles,
such as ketoconazole (1), itraconazole (2), econazole (3),
miconazole (4), fluconazole (5), and nonazole antifungal
agents, i.e., the allylamines naftifine (6) and terbinafine
(7), polyene antibiotics, and flucytosine (Chart 1).
Despite appreciable antifungal activities have been
registered in the in vitro test against C. albicans,
candidemia remains an unfulfilled matter with main-
tenance of the mortality rate.5 Azoles are endowed with
fungistatic activity and are responsible of nausea and
hepatotoxicity as side effects. Allylamines are regarded
as active as azoles and polyenes. The latter cause
adverse side effects such as chills, nausea, and vomiting
during acute treatment and nephrotoxicity and anemia
during chronic treatment.6
documented by about 200 patents registered within
1998-2000.
Recently, Lass-Flo¨rl et al. reported the in vitro anti-
fungal properties of selective serotonin reuptake inhibi-
tors (SSRIs) against Aspergillus species and Candida
parapsilosis.7 Among five tested SSRIs, fluoxetine (8)
and sertraline (9) showed the highest activity against
these fungi with differences in susceptibility of the
various isolates tested.
This report and our personal interest on azole anti-
fungal agents8-10 has induced us to investigate how the
antifungal activity of 8 would be influenced by the
replacement of the NHCH3 terminus with an imidazole
ring. Thus, we planned the synthesis of derivatives 10-
18 which are imidazole analogues of fluoxetine (Chart
2).
The increasing number of therapeutic failures due to
the widespread diffusion of C. albicans infections to-
gether with the concomitant increase in drug resistant
strains foregrounded the need for new effective anti-
Candida agents. Efforts in this direction were well
* Corresponding authors. For R.S.: phone +39 06 4991 3800; fax
+39 06 491 491, e-mail: romano.silvestri@uniroma1.it; for A.T.P.:
phone +39 06 446 8626; fax +39 06 4991 4637, annateresa.palamara@
uniroma1.it.
Compounds 10-18 were prepared starting from 1-aryl-
3-chloropropanones 19-22 which were reduced to 1-aryl-
† Istituto Pasteur - F. C., Dipartimento di Studi Farmaceutici.
‡ Istituto di Microbiologia.
10.1021/jm049856v CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/02/2004

Letters
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 16 3925
Ta ble 1. Antifungal Activity of Compounds 10-18 against 20 Candida Albicans Strains
a
b
d
e
h
i
compd CC50 CC50
MIC ( SDc MIC50 MIC90 MIC range QCf MIC MFC ( SDg MFC50 MFC90 MFC range QCj MFC
10
11
12
13
14
15
16
17
18
1
86
71
38.4 ( 27.2
64
32
8
2
4
32
2
2
2
4
128
32
16
4
8
64
4
4
8
8
8
16-128
8-64
4-16
1-8
4
4
2
4
2
8
0.5
0.5
1
0.06
2
1
0.5
8
76.8 ( 57.2
67.2 ( 66.4
13.6 ( 3.7
9.8 ( 9.6
64
32
16
4
8
64
4
128
>128
16
16
8
128
4
8
16
32
32
16
- -
128
32-128
16-128
8-16
2-32
4-16
32-128
1-8
2-16
2-16
8-32
8.32
16
32
16
32
8
32
8
4
4
0.5
2
0.5
8
97.2 102.3 26.8 ( 12.2
140.8
84.9 221
56.5 96.7
95.9
9.4 ( 4.16
3.15 ( 1.9
5.25 ( 2.4
1-8
8 ( 2.25
219.2 298.3 44.2 ( 19.6
61
43.3
40
70
60
40
316
33.3
4-64
0.5-4
0.5-4
1-4
2-16
2-8
1-8
88 ( 34.2
3.62 ( 2.06
6.25 ( 3.7
7.12 ( 4.95
16.3 ( 5.6
15.04 ( 7.4
12.2 ( 6.2
54
2.1 ( 1.24
2.55 ( 1.4
2.65 ( 1.3
5.9 ( 3.1
4.3 ( 1.7
3.25 ( 1.6
33.5
31.4
61.7
54.7
39.3
8
4
16
16
8
- -
128
3
4
5
8
4
4
2
64
4
64
128
4-32
385.2 15.32 ( 23.2
25 76.8 ( 31.8
0.125-64
32-128
>64
112 ( 28.4
>64
64-128
64
a,b
CC50: Drug concentration (µM) required to reduce the human histiocytic lymphoma U937 (a) and chronic myeloid leukemia K562
(b) cells viability by 50%. c MIC ( SD: Arithmetic mean of minimum inhibitory concentration (µg/mL) values ( standard deviation.
d,e MIC50 and MIC90: MICs at which 50% and 90% of isolates, respectively, are inhibited. f QC C. parapsilosis ATCC22019 MIC values.
g
MFC ( SD: Arithmetic mean of minimum fungicidal concentration (µg/mL) values ( standard deviation; when MIC/MFC ratio is lower
than three double dilutions, the drug is regarded as fungicidal. MFC50 and MFC90: MFCs required to cause 99% reduction of surviving
cells in 50% and 90%, respectively, of isolates. QC C. parapsilosis ATCC22019 MFC values.
h,i
j
Sch em e 1a
to a microdilution tray as described in NCCLS docu-
ment.11 The compounds were tested at different con-
centrations ranging from 0.25 to 128 µg/mL. Ketocon-
azole, econazole nitrate, miconazole nitrate, fluconazole,
and fluoxetine were used as reference drugs. Cytotox-
icity of test compounds was evaluated on human his-
tiocytic lymphoma (U937) and chronic myeloid leukemia
(K562) cell lines obtained from American type culture
Collection (ATCC, Rockville; MD). The significance of
differences in mean values was determined by using a
paired t test (SigmaStat 2.03 statistical software for
Windows). P values of <0.05 were considered statisti-
cally significant.
First we focused our attention mainly on the synthesis
of 3-phenyl-3-(4-trifluomethylphenoxy)-1-(1H-imidazol-
1-yl)propane (10), the fluoxetine-like imidazole neces-
sary for a direct comparison between its anti-Candida
activity and that of the lead fluoxetine (8).
When tested against C. albicans (Table 1) compound
10 showed minimum inhibitory concentration (MIC)
about 2-fold greater than that of fluoxetine (P < 0.001),
thus suggesting the imidazole as the moiety responsible
for the increase of activity. However, the inhibitory
potency of 10 differed greatly from those of econazole
(3) and miconazole (4), two chloro derivatives of the
imidazole family largely used for clinical treatment of
antifungal disease. We therefore attempted to improve
the antifungal potency of 10 by replacing the 4-trifluo-
romethyl substituent with a 4-nitro group (compound
11), but only a very little variation of the inhibitory
potency was observed as a result of this chemical
modification. On the contrary, a remarkable increase
in activity was obtained when methyl (12) or chlorine
(13) were inserted instead of the trifluoromethyl group
of 10 (P < 0.001).
The 4-methyl derivative 12 was found about 8- and
4-fold more active in MIC values than fluoxetine (8)
(P < 0.001) and 10 (P < 0.001), respectively. The best
activity was displayed by the 4-chloro derivative 13
which was about 24-times more active than fluoxetine
and showed against C. albicans the same inhibitory
activity of miconazole (4). Compared with ketoconazole
(1) compound 13 was found to be 2-times more active
(P ) 0.002, n ) 20).
a
Reagents: (a) R3-Ph, ClCOCH2CH2Cl, AlCl3, r.t., 2 h; (b):
NaBH4, THF-H2O, r.t., 2 h; (c): PBr3, Et2O, r.t., 2 h; (d): 2-R1-
4-R2-PhOH, NaH, DMF, r.t., 60 h; (e): imidazole, DMF, reflux,
24 h.
3-chloro-propanols 23-26 with sodium borohydride and
then transformed into the corresponding 1-aryl-1-bromo-
3-chloropropanes 27-30 with phosphorus tribromide.
Reaction of 27-30 with proper phenols in the presence
of sodium hydride afforded the corresponding 1-(2-R1,4-
R2-aryloxy)-1-(4-R3-aryl)-3-chloropropanes 31-39 which
were reacted with imidazole in DMF at 100 °C to give
the required compounds 10-18. 1-Phenyl-1-(1H-imida-
zol-1-yl)-2-propene (40) was obtained by heating 27 with
imidazole in DMF at 100 °C (Scheme 1).
Test compounds were evaluated in vitro for their
antifungal activity against 20 strains of C. albicans, a
pathogenic opportunistic dimorphic fungus, according

3926 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 16
Letters
The surprisingly high degree of activity showed by
the monochloro derivative 13 suggested to attempt a
progressive introduction of chlorine atoms in the phenyl
rings in order to gradually improve the antifungal
activity of this compound. In fact, previous studies on
antifungal imidazoles claimed the maximum of activity
for compounds bearing three/four chlorine atoms, as
proved with the couple econazole/miconazole.
Unexpectedly, the introduction of a second atom of
chlorine in the phenoxy moiety resulted in a slight
decrease of activity (compare 14 with 13 P ) 0.004),
which was restored when a third chlorine atom was
linked to the 4 position of the phenyl bound directly to
the propane chain (compare 16 with 13 and 14, P )
0.005 and P < 0.001 respectively).
Due to the high chemotherapeutical interest shown
by some fluoro derivatives, such as fluconazole, flutri-
mazole and fluoroquinolones, we synthesized 15, the
difluorocounterpart of 14, but without success with
regard to the antifungal activity. Therefore the prepara-
tion of difluorophenoxy derivatives was discontinued
indefinitely.
Replacement of the third chlorine of 16 with fluorine
(17) or methyl (18) gave new derivatives as potent as
16. It is noteworthy that the 2,4-dichlorophenoxy moiety
led to potent antifungal derivatives independently from
the nature of the substituents at 4 position of the
phenylpropane moiety (compare 16-18 with 14 P <
0.001). This result is clearly different from that observed
for the couple econazole/miconazole, which showed the
highest antifungal activity in the presence of three/four
chlorine atoms.
In conclusion, the introduction of substituents differ-
ent from trifluoromethyl in the structure of 10 led to a
substantial increase of activity with the maximum
potency for chlorine (13), followed in decreasing order
by methyl (12) and nitro (11). In general, all new
derivatives were more active than fluoxetine, the com-
pound selected as lead compound for the present search.
As a rule, monosubstituted derivatives were less
active than those bearing two or three substituents.
Data against C. albicans showed five compounds more
active or as potent as reference clinical drugs ketocona-
zole, econazole, and miconazole (compare 13, 16-18
with 1 and 14 with 3 and 4), the trichloro derivative 16
being the most potent. As regards the fungicidal activity
(MFC90), it is worthy to note that compound 16 was 4
(miconazole, P < 0.001) to 8 (ketoconazole, P < 0.001)
times more potent than the references. Furthermore,
imidazole 13 and 16 were less cytotoxic than reference
compounds, including fluconazole, with a selectivity
index CC50/MIC50 of 76.5 and 28.75, the highest among
test derivatives 10-18.
As shown by the examples reported here, proper
manipulations of the number, nature, and position of
substituents in the phenyl ring and phenoxy groups of
imidazole derivative 13 can be useful for the design and
synthesis of novel potent antifungal agents active
against C. albicans. Trichloro substitution seems at the
present to be the most favorable chemical modification
with the formation of the potent econazole-like deriva-
tive 16, but other substituents than chlorine, for ex-
ample methyl and fluorine, can act as suitable bioiso-
sters.
Our preliminary results on the antifungal activity of
new compounds against a strain of Aspergillus fumiga-
tus and a strain of Aspergillus niger demonstrate that
14, 16, and 17 were active at a concentration of 16 µg/
mL and more active than fluoxetine (100 µg/mL for A.
fumigatus and 32 µg/mL for A. niger).
Besides the high potency displayed by some members
of the novel class of antifungal imidazoles discovered
in the present preliminary work, we are aware that
further progress is obtainable to acquire compounds
much more active than miconazole. For this reason,
searches are now ongoing on novel fluoxetine-like azoles
for developing more wide and deep structure-activity
relationships (SAR) investigations also guided by mo-
lecular modeling and QSAR studies. Furthermore,
comparative microbiological tests between title com-
pounds and other classical azoles against a panel of
different fungal pathogens will be performed soon to
complete the present search.
Ack n ow led gm en t. The authors thank italian MIUR
for financial aid. Thanks are also due to Progetto Firb
2001, grant n. RBNE01P4B5-006.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails. This material is available free of charge via the Internet
at http://pubs.acs.org.
Refer en ces
(1) Weinberg, E. D. Antifungal Agents. Burger’s Medicinal Chem-
istry and Drug Discovery, 5th ed.; Wiley-Interscience: New York
1996; Vol. 2, pp 637-652.
(2) Fungal Infections: Dermatomycoses. Drug R&D Backgrounder
2001, 1, 57-66.
(3) Wildfeuer, A.; Seidl, H. P.; Paule, I.; Haberreiter, A. In vitro
evaluation of voriconazole against clinical isolates of yeast,
moulds, and dermatophytes in comparison with itraconazole,
ketoconazole, amphotericin and griseofulvin. Mycoses 1998, 41,
306-319.
(4) Georgopapadakou, N. H. Antifungals: mechanism of action and
resistance, established and novel drugs. Curr. Opin. Microbiol.
1998, 1, 547-557.
(5) Sheehan, D. J .; Hitchcock, C. A.; Silbey, C. M. Current and
emerging azole antifungal agents. Clin. Microbiol. Rev 1999, 12,
501-517. (b) Watkins, W. J .; Renau, T. E. Progress with
antifungal agents and approaches to combat fungal resistance.
In Annual Reports in Medicinal Chemistry; Doherty, A. M., Ed.,
Academic Press: New York, 1999; Vol. 35, pp 157-166.
(6) Ryder, N. S.; Squalene epoxidase as a target for the allylamines.
Biochem. Soc. Trans. 1991, 19, 774-777.
(7) Lass-Flo¨rl, C.; Dierich, M. P.; Fuchs, D.; Semenitz, E.; J enewein,
I.; Ledochowski, M. Antifungal properties of selective serotonin
reuptake inhibitors against Aspergillus species in vitro. J .
Antimicrob. Chemother. 2001, 48, 775-779.
(8) Massa, S.; Stefancich, G.; Corelli, F.; Silvestri, R.; Mai, A.; Artico,
M.; Panico, S.; Simonetti, N. Research on antibacterical and
antifungal agents. Synthesis and antifungal activities of 1-{p-
methyl-alpha-[4-(1H-pyrrol-1-yl)phenyl]benzyl}azoles and some
related products. Arch. Pharm. (Weinheim) 1989, 322, 369-373.
(9) Stefancich, G.; Artico, M.; Ortar, G.; Silvestri, R.; Simonetti, G.;
Apuzzo, G.; Artico, M.. Synthesis and antifungal activity of
structural analogues of bifonazole and ketoconazole. Arch.
Pharm. (Weinheim) 1992, 325, 687-694.
(10) Artico, M.; Ragno, R.; Porretta, G. C.; Massa, S.; Musiu, C.;
Spiga, M. G.; Corrias, S.; La Colla, P. Iso-Miconazole and related
derivatives: synthesis and antifungal activities. Med. Chem. Res.
1996, 137-147.
(11) Villanova, P. A. Reference method for broth dilution antifungal
susceptibility testing of yeast: aprroved standard. National
Committee for Clinical Laboratory Standards 2002, Document
M27-A2.
J M049856V