Design and synthesis of novel benzofurans as a new class of antifungal agents targeting fungal N-myristoyltransferase. Part 3

Article abstract of DOI:10.1016/S0960-894X(02)00844-2

A new series of acid-stable antifungal agents having strong inhibitory activity against Candida albicans N-myristoyltransferase (CaNmt) has been developed starting from acid-unstable benzofuranylmethyl aryl ether 2. The inhibitor design is based on X-ray crystallographic analysis of a CaNmt complex with aryl ether 3. Among the new inhibitors, pyridine derivative 8b and benzimidazole derivative 8k showed clear antifungal activity in a murine systemic candidiasis model.

Full text of DOI:10.1016/S0960-894X(02)00844-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 87–91
Design and Synthesis of Novel Benzofurans as a New Class of
Antifungal Agents Targeting Fungal N-Myristoyltransferase.
Part 3
Ken-ichi Kawasaki,a Miyako Masubuchi,a Kenji Morikami,a Satoshi Sogabe,a
Tsunehisa Aoyama,a Hirosato Ebiike,a Satoshi Niizuma,a Michiko Hayase,b
Toshihiko Fujii,b Kiyoaki Sakata,b Hidetoshi Shindoh,c Yasuhiko Shiratori,a Yuko Aoki,b
Tatsuo Ohtsukaa,* and Nobuo Shimmaa
aDepartment of Chemistry, Nippon Roche Research Center, 200 Kajiwara, Kamakura, Kanagawa 247-8530, Japan
bDepartment of Mycology, Nippon Roche Research Center, 200 Kajiwara, Kamakura, Kanagawa 247-8530, Japan
cDepartment of Preclinical Science, Nippon Roche Research Center, 200 Kajiwara, Kamakura, Kanagawa 247-8530, Japan
Received 6 August 2002; accepted 1 October 2002
Abstract—A new series of acid-stable antifungal agents having strong inhibitory activity against Candida albicans N-myristoyl-
transferase (CaNmt) has been developed starting from acid-unstable benzofuranylmethyl aryl ether 2. The inhibitor design is based
on X-ray crystallographic analysis of a CaNmt complex with aryl ether 3. Among the new inhibitors, pyridine derivative 8b and
benzimidazole derivative 8k showed clear antifungal activity in a murine systemic candidiasis model.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
inhibitors12,13 of C. albicans Nmt (CaNmt) have been
identified. Among them only the benzofuran inhibitors
have been reported to show clear antifungal activity in
vivo. In our previous papers, we reported the chemical
modification work that yielded new benzofuran Nmt
inhibitors 1 and 2.12,13 The inhibitor design was mainly
based on X-ray crystallographic analysis of CaNmt
complex with a benzofuran inhibitor and the modifi-
cation was guided by various biological investigations
including in vitro antifungal assay in the presence of
80% calf serum (quasi in vivo assay) and pharmaco-
kinetic (PK) study. Such a multi-dimensional optimiza-
tion resulted in the identification of fungicidal
benzofuranylmethyl aryl ether 2, which showed both
strong CaNmt inhibitory activity (IC50: 0.0056 mM) and
in vitro antifungal activity against C. albicans CY1002
(IC50: 0.035 mM). This compound also showed a good
PK profile and in vivo efficacy in a murine systemic
candidiasis model (ED50: 7.1 mg/kg), but it was
unstable in an artificial gastric fluid (pH 1.2). This labile
nature of 2 should be improved for developing an orally
active antifungal drug. To obtain acid-stable antifungal
agents we designed benzofuranyl heteroaromatic
ketones 8a–8k by analyzing the crystal structure of the
N-Myristoyltransferase (Nmt) is an enzyme that trans-
fers the myristoyl group of myristoyl CoA to the N-
terminal glycine of various eukaryotic cellular proteins.
N-Myristoylation of proteins, Gpa1, Arf1, Arf2 and
Vps15, which are essential for fungal growth, has been
reported to be indispensable for their function in Sac-
charomyces cerevisiae.1À4 Nmt has also been proven to
be essential for the viability of fungi, including medi-
cally important pathogenic fungi such as Candida albi-
cans5 and Cryptococcus neoformans.6 Therefore, Nmt is
a good target for the development of novel fungicidal
drugs with a new mode of action. Since the mechanism
of action is novel, Nmt inhibitors might show antifungal
activity against even azole resistant fungal strains and
might not cause the drug-drug interactions that are a
drawback of azole antifungal agents.7
So far, peptidomimetic inhibitors,8 benzothiazole inhibi-
tors,9,10 p-toluenesulfonamide inhibitors11 and benzofuran
*Corresponding author. Tel.: +81-467-45-4356; fax: +81-467-45-
6824; e-mail: ohtsukatto@chugai-pharm.co.jp
0960-894X/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00844-2

88
K. Kawasaki et al. / Bioorg. Med. Chem. Lett. 13 (2003) 87–91
CaNmt complex with 3. In the present paper, we wish to
report the design and synthesis of acid-stable benzo-
furan derivatives having inhibitory activity against
CaNmt, and their biological activities.
was converted into methylene derivative 10 by treat-
ment with sodium borohydride in a mixture of tri-
fluoroacetic acid and dichloromethane.14
Results and Discussion
Synthesis
Design of acid-stable inhibitors
The general synthesis of the designed inhibitors (8a–8k)
and reference compounds (8l, 8m, 9 and 10) is outlined
in Scheme 1.
The crystallographic analysis of a CaNmt complex with
3 and SARs obtained so far suggested that following
interactions are important for the binding: (1) the
hydrogen bonding between amide–NH2 of Asn 392 and
the ether oxygen of 3, and (2) the hydrophobic inter-
action between the difluorophenyl group of 3 and phe-
nylalanine residues, Phe 115, Phe 240 and Phe 339 (Fig.
1).13 Since the arylmethyl ether moiety of 2 was thought
to be unstable under acidic conditions, we designed
compound 8a that has an acid-stable ketone group
instead of the ether group. The nitrogen atom of the
pyridine ring would be a good hydrogen bond acceptor
and the pyridine ring would be able to situate in the
lipophilic pocket composed of the three phenylalanine
residues. To locate the pyridine ring and its nitrogen
atom to the appropriate positions, a carbonyl group
was inserted between the pyridine ring and the benzo-
furan ring. The repulsive Coulomb force between the
carbonyl oxygen and the benzofuran oxygen was
Ester 413 was hydrolyzed by potassium hydroxide, and
the resulting acid was converted into its acid chloride by
treatment with oxalyl chloride, which was immediately
subjected to the next amide formation reaction with
morpholine. The resulting amide 5 was heated to 70 ꢀC
with 3-aminomethylpyridine in EtOH to give amine 6.
Amine 6 was converted to its t-butoxycarbonyl (Boc)
derivative 7 by treatment with Boc2O. The addition of
2-lithiated pyridine generated in situ to 7 at À78 ꢀC and
subsequent deprotection of Boc group with tri-
fluoroacetic acid afforded pyridyl ketone 8a in a good
yield. Other heteroaromatic ketone derivatives 8b–8m
were synthesized in a similar manner from the common
starting material 7 and an appropriate lithiated hetero-
aromatic compound. Treatment of 8a with sodium
borohydride gave alcohol 9 quantitatively and then it
Scheme 1.

K. Kawasaki et al. / Bioorg. Med. Chem. Lett. 13 (2003) 87–91
89
Figure 1. (a) Interaction between Asn 392 of CaNmt and 3 suggested by X-ray crystallographic analysis; (b) hypothetical interaction between Asn
392 of CaNmt and 8a; (c) the crystal structure of the binding site of a CaNmt complex with 3.
Figure 2. Comparison of the lowest energy conformations of A, B and C at the HF/6-31G level. Atoms denoted by red and blue are oxygen and
nitrogen atoms, respectively. Phenyl ether moiety of A is depicted in green. (a) Superimposition of the lowest energy conformations of A and B. The
conformation of A was calculated to be similar to that of the corresponding part of 3 bound with CaNmt.15 (b) Superimposition of the lowest energy
conformations of A and C.

90
K. Kawasaki et al. / Bioorg. Med. Chem. Lett. 13 (2003) 87–91
Table 1. Enzyme inhibitory activity against CaNmta
other heteroaromatic ketones for further improvement
of the activity. The pyridine ring of 8a was replaced by
various heteroaromatic rings as shown in Scheme 1. The
enzyme inhibitory activities of the new analogues
against CaNmt, human Nmt (HsNmt) and Aspergillus
fumigatus Nmt (AfNmt), their in vitro antifungal activ-
ity and PK properties are summarized in Table 2.
Compd
IC50 (mM)
8a
0.0035
9
0.047
10
1.9
8l
0.62
8m
1.2
aPotency against C. albicans Nmt (CaNmt) as assessed by IC50 using
substrate peptide GLTISKLFRR-NH2 (0.5 mM) and myristoyl-CoA
at 0.5 mM.
expected to align these heteroatom s-trans and to make
the conformation rigid. To verify our supposition, con-
formational analysis was performed on structures A, B
and C that are the model compounds of 3, 8a and 10,
respectively. Their stable conformations generated by
ab-initio molecular orbital calculation are shown in
Figure 2. The analysis suggested that the pyridine
nitrogen of B would occupy almost the same position as
the ether oxygen of A. On the other hand, the position
and orientation of the pyridine nitrogen of C were cal-
culated to be completely different from those of the
oxygen of A. Thus, 8a was expected to be a good binder,
whereas 10 was not.
Furan derivative 8g showed very weak CaNmt inhibi-
tory activity, whereas the nitrogen-containing hetero-
aromatic derivatives showed strong inhibitory activity.
Since the furan oxygen is expected to be a poor hydro-
gen acceptor, it is reasonable that the inhibitory activity
of 8g was weaker than that of 8a. Introduction of
methyl group(s) to the thiazole ring enhanced the activ-
ity significantly (8e and 8f vs 8d). A similar increase of
potency was also observed when a methyl group was
introduced to the pyridine ring (8b vs 8a). Introduction
of methyl groups, the electron-donating groups, increa-
ses the electron density on thiazole nitrogen or pyridine
nitrogen and thus would increase the hydrogen bond
interaction with Asn 392. However, there is another
factor that enhanced the activity of the methyl deriva-
tives. The methyl group(s) might fit into a favorable
position in the lipophilic pocket composed of Phe 115,
Phe 240 and Phe 339.
Inhibitory activity against CaNmt
The enzyme inhibitory activity against CaNmt was
measured by the same methods as described in previous
papers12,13 and the results are summarized in Table 1.
As we expected, compound 8a bearing 2-pyridyl car-
bonyl group showed strong inhibitory activity (IC50:
0.0035 mM) against CaNmt, whereas its methylene ana-
logue 10 was 540 times less active. On the other hand,
8l, a position isomer of 8a, was 180 times less active
than 8a. These results suggested that the position of the
pyridine nitrogen is crucial for strong binding to
CaNmt.
Most inhibitors in Table 2 showed very strong enzyme
inhibitory activities against CaNmt with extremely high
selectivity over HsNmt. However, none of them showed
strong inhibitory activity against AfNmt.
Antifungal activity and acid-stability
The strong CaNmt inhibitors, 8a–8f and 8h–8k, also
showed strong antifungal activity against C. albicans
CY1002 in vitro (Table 2). Among them, 8b and 8k
were further evaluated by in vivo experiments, since 8b
Thus, a new type of strong enzyme inhibitor 8a was
obtained and this success prompted us to synthesize
Table 2. Enzyme inhibitory activity, in vitro antifungal activity and pharmacokinetic (PK) properties in rats
Compd
Enzyme
inhibition (mM)
CaNmta
Enzyme
inhibition (mM)
HsNmtb
Enzyme
inhibition (mM)
AfNmtc
Antifungal activity
(mM) serum (À)d
Antifungal activity
(mM) serum (+)e
PK parameters
in ratsf AUC
(ng h/mL)
t1/2 (h)
8a
8b
8c
8d
8e
8f
8g
8h
8i
8j
8k
0.0035
0.00039
0.043
340
>480
>430
25
130
83
280
69
94
99
42
4.1
11
63
18
4.3
150
30
0.065
0.036
0.34
0.54
0.14
0.039
4.6
0.12
0.035
0.019
0.048
0.72
0.47
0.082
5.1
410
440
0.54
0.58
0.56
NT
0.34
0.65
NT
0.45
0.59
0.86
0.80
4.0
590
0.091
8.3
NTg
470
0.0020
0.00071
1.4
0.0013
0.00040
0.0013
0.00058
NT
0.74
0.19
25
3.0
1.2
2.7
0.42
0.5
320
NT
530
450
810
570
13
0.48
5.1
NT
85
NT
Fluconazole
19,600
aPotency against C. albicans Nmt (CaNmt) as assessed by IC50 using substrate peptide GLTISKLFRR-NH2 (0.5 mM) and myristoyl-CoA at 0.5 mM.
bPotency against human Nmt (HsNmt) as assessed by IC50 using substrate peptide GNAASARR-NH2 (0.5 mM) and myristoyl-CoA at 0.5 mM.
cPotency against A. fumigatus Nmt (AfNmt) as assessed by IC50 using substrate peptide GLTISKLFRR-NH2 (0.5 mM) and myristoyl-CoA at
0.5 mM.
dAntifungal activity against C. albicans CY1002 as assessed by IC50 in YNBPB medium (1% glucose, 0.25% K2HPO4, pH 7).
eAntifungal activity against C. albicans CY1002 as assessed by IC50 in 80% calf serum (80% calf serum supplemented with 10 mM FeCl3 6H2O,
.
10 mM deferoxamine, 2% dextrose).
fCassette dosing: Five compounds were intravenously administered to a rat (2 mg/kg) and the plasma concentration of each compound was mea-
sured by LC–MS.16
gNT, not tested.

K. Kawasaki et al. / Bioorg. Med. Chem. Lett. 13 (2003) 87–91
91
showed the strongest antifungal activity against C. albi-
References and Notes
cans CY1002 in the presence of 80% calf serum and 8k
showed a good PK profile. The antifungal assay in 80%
serum should be more predictive for in vivo efficacy
than the conventional in vitro antifungal assay, since the
former assay system includes a factor of drug-protein
binding.17,18 Their in vivo efficacy was determined as fol-
lows: (1) Fisher rats (n=5) were infected intravenously
with a lethal dose of C. albicans (CY1002); (2) The infected
rats were treated iv with multiple doses (three times per
day for 2 days) of the test compound; and (3) efficacy of
the compounds was calculated as the effective dose (mg/
kg) for 50% survival (ED50) on day 7. The compounds
8b and 8k showed clear antifungal activity in this model
with ED50s of 15 and 8.7 mg/kg, respectively. However,
the efficacy was still weaker than that of fluconazole.
This can be explained by a poorer PK profile than that
of fluconazole. We still need to improve the PK profile
to obtain a strong antifungal drug.
1. Gordon, J. I.; Rudnick, D. A.; Adams, S. P.; Gokel, G. W.
J. Biol. Chem. 1991, 266, 8647.
2. Johnson, D. R.; Bhatnagar, R. S.; Knoll, L. J.; Gordon, J. I.
Annu. Rev. Biochem. 1994, 63, 869.
3. Stearns, T.; Kahn, R. A.; Botstein, D.; Hoyt, M. A. Mol.
Cell. Biol. 1990, 10, 6690.
4. Herman, P. K.; Stack, J. H.; DeModena, J. A.; Emr, S. D.
Cell 1991, 64, 425.
5. Weinberg, R. A.; McWherter, C. A.; Freeman, S. K.;
Wood, D. C.; Gordon, J. I.; Lee, S. C. Mol. Microbiol. 1995,
16, 241.
6. Lodge, L. K.; Jackson-Machelski, E.; Toffaletti, D. L.;
Perfect, J. R.; Gordon, J. I. Proc. Natl. Acad. Sci. U.S.A. 1994,
91, 12008.
7. Benedetti, M. S.; Bani, M. Drug Metab. Rev. 1991, 31, 665.
8. Devadas, B.; Gordon, J. I.; Sikorski, J. A. J. Med. Chem.
1998, 41, 996, and reference cited therein.
9. Armour, D. R.; Bell, A. S.; Kemp, M. I.; Edwards, M. P.;
Wood, A. Abstracts, 221st ACS National Meeting, San Diego,
April 1–5, 2001; MEDI-349.
As we expected, these compounds were stable under
acidic conditions. HPLC analysis showed 100, 86, 87
and 99% of 8a, 8b, 8j and 8k remained unchanged even
after 2-h treatment in artificial gastric fluid (pH 1.2) at
37 ꢀC, whereas 94% of 2 decomposed after 2-h treat-
ment under the same condition. Thus, we successfully
designed and synthesized acid-stable inhibitors that
showed clear antifungal activity against C. albicans both
in vitro and in vivo. The PK properties, however, were
not much improved by the modification. Although the
benzimidazole derivatives showed slightly higher AUC
10. Bell, A. S.; Armour, D. R.; Edwards, M. P.; Kemp, M. I.;
Wood, A. Abstracts, 221st ACS National Meeting, San Diego,
April 1–5, 2001; MEDI-350.
11. Ganesh, K. R.; Madhvaro, K. V. Indian Drugs 2001, 38,
406.
12. Masubuchi, M.; Kawasaki, K.; Ebiike, H.; Ikeda, Y.;
Tsujii, S.; Sogabe, S.; Fujii, T.; Sakata, K.; Shiratori, Y.; Aoki,
Y.; Ohtsuka, T.; Shimma, N. Bioorg. Med. Chem. Lett. 2001,
11, 1833.
13. Ebiike, H.; Masubuchi, M.; Liu, P.; Kawasaki, K.; Mor-
ikami, K.; Sogabe, S.; Hayase, M.; Fujii, T.; Sakata, K.;
Shindoh, H.; Shiratori, Y.; Aoki, Y.; Ohtsuka, T.; Shimma, N.
Bioorg. Med. Chem. Lett. 2002, 12, 607.
.
than did the aryl ether 2 (AUC=330 ng h/mL), their
elimination half-lives (t1/2) were shorter than that of 2
(t1/2=2.0 h).
14. Gribble, G. W.; Kelly, W. J.; Emery, S. E. Synthesis 1978,
763.
15. Meunier, F.; Lambert, C.; Van der Auwera, P. J. Anti-
microb. Chemother. 1990, 25, 227.
16. Frick, L. W.; Adkison, K. K.; Wells-Knecht, K. J.;
Woollard, P.; Higton, D. M. Pharm. Sci. Technol. Today 1998,
1, 12.
17. Masubuchi, K.; Okada, T.; Kohchi, M.; Sakaitani, M.;
Mizuguchi, E.; Shirai, H.; Aoki, M.; Watanabe, T.; Kondoh,
O.; Yamazaki, T.; Satoh, Y.; Kobayashi, K.; Inoue, T.; Horii,
I.; Shimma, N. Bioorg. Med. Chem. Lett. 2001, 11, 395.
18. Masubuchi, K.; Okada, T.; Kohchi, M.; Murata, T.;
Tsukazaki, M.; Kondoh, O.; Yamazaki, T.; Satoh, Y.; Ono,
Y.; Tsukaguchi, T.; Kobayashi, K.; Ono, N.; Inoue, T.;
Horii, I.; Shimma, N. Bioorg. Med. Chem. Lett. 2001, 11,
1273.
In summary, crystallographic analysis of the CaNmt
complex with benzofuranylmethyl aryl ether 3 led us to
design a new series of acid-stable antifungal agents,
heteroaromatic ketones 8a–8k. Among them, 8b and 8k
showed clear antifungal activity in a murine systemic
candidiasis model. We have overcome the acid labile
nature of benzofuranylmethyl aryl ether 2, keeping its
strong antifungal activity. Nevertheless, since the inhi-
bitors do not strongly inhibit AfNmt, further modifi-
cation work is required to obtain compounds that
inhibit both AfNmt and CaNmt and thus would inhibit
the growth of both A. fumigatus and C. albicans.