Identification of benzofuran-4,5-diones as novel and selective non-hydroxamic acid, non-peptidomimetic based inhibitors of human peptide deformylase

Article abstract of DOI:10.1016/j.bmcl.2011.05.129

Selective inhibitors of human peptide deformylase (HsPDF) are predicted to constitute a new class of antitumor agents. We report the identification of benzofuran-4,5-diones as the first known selective HsPDF inhibitors and we describe their selectivity pr

Full text of DOI:10.1016/j.bmcl.2011.05.129

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4528–4532
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of benzofuran-4,5-diones as novel and selective
non-hydroxamic acid, non-peptidomimetic based inhibitors of human peptide
deformylase
Christophe Antczak a, David Shum a, Bhramdeo Bassit a, Mark G. Frattini b, Yueming Li a, Elisa de Stanchina a,
David A. Scheinberg a, Hakim Djaballah a,
⇑
a Department of Molecular Pharmacology and Chemistry, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA
b Department of Medicine, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Selective inhibitors of human peptide deformylase (HsPDF) are predicted to constitute a new class of
antitumor agents. We report the identification of benzofuran-4,5-diones as the first known selective
HsPDF inhibitors and we describe their selectivity profile in a panel of metalloproteases. We characterize
their structure–activity relationships for antitumor activity in a panel of cancer cell lines, and we assess
their in vivo efficacy in a mouse xenograft model. Our results demonstrate that selective HsPDF inhibitors
based on the benzofuran-4,5-dione scaffold constitute a novel class of antitumor agents that are potent
in vitro and in vivo.
Received 18 April 2011
Revised 27 May 2011
Accepted 31 May 2011
Available online 12 June 2011
Keywords:
Human peptide deformylase
Benzofuran-4,5-diones
Structure–activity relationships
Fluorescence polarization
Antiproliferative agents
Ó 2011 Elsevier Ltd. All rights reserved.
During protein synthesis in prokaryotes, the N-formyl group of
nascent peptides is removed from most peptides in order to yield
mature proteins. Consequently, PDF activity is essential to bacterial
growth.1,2 Since until recently PDF was thought to be absent from
eukaryotes, PDF has constituted an attractive target for the devel-
opment of antibotics.3 However, the demonstration of the exis-
tence of a functional human analogue of PDF4–6 raises concerns
over the use of non selective PDF inhibitors as antibacterial agents
in humans.
Following the observation that HsPDF inhibition by actinonin (1)
and actinonin analogs or by specific siRNA knockdown of expression
is associated with antiproliferative effect in cancer cells7, we specu-
lated that HsPDF inhibitors could constitute a new class of antitumor
agents. However, most currently known PDF inhibitors such as acti-
nonin consist of a peptidomimetic backbone attached to a hydroxa-
mic acid moiety, and this class of compounds is typically associated
with poor selectivity across metalloproteases.8–11 In addition, their
poor bioavailability precludes their use in vivo as antitumor agents.
The crystal structure of an N-terminal truncated, catalytically active
HsPDF revealed structural differences between HsPDF and EcPDF
such as a characteristic entrance to the active site that provide a
rationale for the identification of selective HsPDF inhibitors.12 For
this reason, we developed and validated a strategy that would allow
us to identify novel non peptidomimetic and non hydroxamic acid
based inhibitors of HsPDF13, and we subsequently embarked in the
screening of a library of 200,000 small molecules using our proven
strategy. Among the confirmed positives identified in this campaign
were five compounds (2–6) belonging to the chemical scaffold of
benzofuran-4,5-diones (Fig. 1). All five compounds induced P75%
inhibition at 10 lM in our fluorescence polarization-based assay
for HsPDF (Fig. 1) in absence of any optical interference, which
was measured as previously described.13 In addition, the five benzo-
furan-4,5-diones identified during primary screening were con-
firmed as functional inhibitors of HsPDF using a methodology
previously described.13 While the benzofuran moiety is included
in inhibitors of various enzymes, to our knowledge, no inhibitory
activity toward any PDF and no antitumor activity has previously
been described for the chemical scaffold of benzofuran-4,5-diones.
In order to expand the limited structure–activity relationships of
benzofuran-4,5-diones gathered during primary screening, we initi-
ated exploratory chemistry efforts aimed at defining the importance
of the halogen substitutions at
orthodione moiety.
a- and b-positions on the 4,5-
For the synthesis of 13 novel benzofuran-4,5-dione derivatives
and three napthofurandione derivatives, we engaged in a strategy
relying on acid catalyzed reaction of substituted enaminones with
appropriately halogenated 1,4-quinones14–17 to provide a general
⇑
Corresponding author. Tel.: +646 888 2198; fax: +646 422 0706.
E-mail address: djaballh@mskcc.org (H. Djaballah).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.05.129

C. Antczak et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4528–4532
4529
O
O
CH3
O
R1
R2
R1
a
N
O
H
N
HO
O
N
R2
N
H
N
H
R3
CH3
R3
O
O
O
H3C
CH3
actinonin,
7a - 7d
8a - 8d
1
7a: R1 = H; R2 = H; R3 = H
8a: R1 = H; R2 = H; R3 = H
7b: R1 = H; R2 = OCH3; R3 = H
8b: R1 = H; R2 = OCH3; R3 = H
O
7c: R1 = OCH3; R2 = OCH3; R3 = H
8c: R1 = OCH3; R2 = OCH3; R3 = H
O
O
7d: R1 = OCH3; R2 = OCH3; R3 = OCH3
8d: R1 = OCH3; R2 = OCH3; R3 = OCH3
O
O
O
O
Scheme 1. Synthesis of the enaminones 8a–8d. Reagents and conditions: (a) DMF-
DMA, 150 °C, 20–30 h.
Br
Br
2
; HTS%=97
3
; HTS%=81
O
O
O
O
O
O
and hydroquinone in acetic acid as a solvent to give the correspond-
ing 5-hydroxybenzofuran derivatives 16i–30i19,20 (Scheme 2), as
well as the corresponding 5-hydroxynaphthofuran derivatives
32i–34i21 (Scheme 3) in variable yields. The oxidation of 5-hydrox-
ybenzofuran derivatives 16i–28i and 5-hydroxynaphthofuran
derivatives 32i–34i was best accomplished via either nitric acid22,23
or with Dess–Martin’s periodinane, to give the corresponding
substituted 4,5-benzofurandiones 16–28 with 40–57% yield and
the 4,5-naphthofurandiones 32–34 with 38 to 43% yield.
We evaluated the potency of the 13 novel benzofuran-4,5-dione
derivatives toward HsPDF and EcPDF as well as their selectivity
profile using a previously validated methodology.10 We found that
all the benzofuran-4,5-dione derivatives we have synthesized
O
O
O
Br
Cl
Cl
4
; HTS%=90
5
; HTS%=90
R1
O
O
O
O
O
O
X
O
Cl
O
Y
O
O
6
; HTS%=75
inhibit HsPDF with an IC50 ranging from 5.2 to 65 lM (Table 1).
In contrast, when we characterized the potency of the 5-hydrox-
ybenzofuran intermediates of synthesis 16i–30i, we found that
they were all inactive toward HsPDF up to 100 lM (Supplementary
Table 1). This result clearly indicates that the orthodione moiety in
the benzofuran-4,5-dione scaffold is important for activity. Simi-
larly, the three napthofurandiones 32, 33 and 34 we have synthe-
sized have no activity toward any of the metalloproteases tested
(Supplementary Table 2), suggesting that substituting the benzofu-
randione moiety by a naphtofurandione moiety abrogates activity.
Interestingly, since the benzofurandiones 16 and 17 lacking a hal-
Figure 1. Chemical structure of actinonin, 1; chemical structure and percentage
inhibition (HTS%) of confirmed positives in primary screen belonging to the
benzofuran-4,5-dione scaffold, 2–6; general chemical structure of the primary hits
belonging to the benzofuran-4,5-dione scaffold.
construct of substituted 5-hydroxy benzofuran and naphthofuran
derivatives, followed by oxidation with a suitable oxidant. Toward
this end, substituted acetophenones 7a–7d were reacted with di-
methyl formamide dimethyl acetal at 150 °C in DMF to give the
enaminones 8a–8d18 in 63–88% yield (Scheme 1). The enaminones
8a–8d were reacted with appropriately halogenated 1,4-quinones
ogen substituent at a- and b-positions on the 4,5-orthodione moi-
ety are potent toward HsPDF (Table 1), we can conclude that
R1
R1
R2
R2
O
O
O
Y
Z
Y
O
O
R3
O
X
R3
HO
X
Z
X
Y
b
a
8a 8b 8c 8d
or or or
O
O
9 15
16i 30i
16 28
-
-
-
Scheme 2. Synthesis of benzofuran-4,5-diones. Reagents and conditions: (a) AcOH, rt; (b) For 16, 17, 19, 21, 28: Dess–Martin periodinane, DMSO, 0 °C?rt, 20 min; For 5, 20,
22–27: HNO3, AcOH, rt?65 °C, 3 h. Z = Br or Cl, matching X and Y.
R1
R1
R2
R2
O
O
O
O
O
R3
a
O
b
HO
R3
O
8b or 8c or 8d
O
31
32i-34i
32-34
Scheme 3. Synthesis of naphtofurandiones. Reagents and conditions: (a) AcOH, rt; (b) HNO3, AcOH, rt?50 °C, 30 min.

4530
C. Antczak et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4528–4532
Table 1
Potency and selectivity profile of benzofuran-4,5-dione derivatives
R1
R2
O
O
Y
O
X
R3
O
Compd
X
Y
R1
R2
R3
HsPDF (IC50
,
l
M)
EcPDF (IC50
,
l
M)
HsPDF (%I)
EcPDF (%I)
APN (%I)
MMP-1 (%I)
1
—
H
H
Cl
Br
H
—
H
H
H
H
Cl
Br
Cl
Cl
Br
Cl
Br
Cl
Br
—
H
H
H
H
H
H
H
H
H
OMe
OMe
OMe
OMe
—
H
—
H
H
H
H
H
H
H
H
H
H
H
OMe
OMe
2.7
59
34
40
32
16
15
45
15
65
6.1
10
25
5.2
0.14*
>100
>100
>100
>100
>100
>100
>100
>100
>100
67
98
99
32
24
11
5
13
26
65
1
22
63
53
27
32
95
5
4
97
0
6
11
26
10
8
41
0
23
46
46
32
38
16
17
20
28
19
21
5
22
25
23
26
24
27
100
100
100
100
100
100
100
88
100
100
100
100
100
OMe
OMe
OMe
OMe
OMe
H
OMe
OMe
OMe
OMe
OMe
OMe
54
22
68
64
101
50
41
99
96
50
50
H
Cl
Cl
Br
Cl
Br
Cl
Br
>100
>100
>100
%I: % Inhibition at 100 lM, average of duplicates; 100: P100 %I; 0: 60 %I.
*
as assessed in the FLUO functional assay13
.
2,000
140
a
DMSOcontrol
27, 15 mg/kg
27, 30 mg/kg
120
22, HsPDF
1,500
1,000
500
100
80
60
40
20
0
22, EcPDF
27, HsPDF
27, EcPDF
-20
-
9
12
15
18
21
24
0.01
0.1
1
10
100
1000
Time posttumor implantation,days
Compound concentration, µM
Figure 3. In vivo efficacy of 27 in a mouse xenograft model using HL-60 cells.
b
HL-60
HL-60/RV+
Jurkat
Molt-3
ALL-3
CWR22
HEK293
Y79
100
80
60
40
20
0
lack halogens, but rather because the benzofuran-4,5-dione moiety
is necessary for activity. While halogen substitutions at
a- and
b-positions on the 4,5-orthodione moiety are not required for
HsPDF activity, they do seem to increase the potency of benzofu-
ran-4,5-diones toward HsPDF, since the top six most potent deriv-
atives we have synthesized harbor at least one chloro or one bromo
substituent in those positions. Similarly, we observe that the top
three most potent derivatives have two or three methoxy substit-
uents on the benzoyl moiety (Table 1), indicating that those sub-
stituents could be responsible for the formation of hydrogen
Meso47
bonds with the target. With an IC50 of 5.2 lM toward HsPDF, 27
is the most potent HsPDF inhibitor we have characterized. When
we assessed the selectivity of benzofuran-4,5-diones, we found
that only one of the derivatives we have synthesized was weakly
0.01
0.1
1
10
100
1000
[27], µM
active toward EcPDF (23, IC50 = 67 lM), while 12 out of 13 had
Figure 2. (a) Dose response curves for 22 and 27 toward HsPDF and EcPDF. (b)
Cytotoxicity profiling of 27.
no measurable activity toward EcPDF (Table 1). This result con-
trasts with the broad activity of benzofuran-4,5-diones toward
the human enzyme and demonstrates that this class of compounds
is selective for the human PDF compared to the bacterial enzyme.
halogens in those positions are not required for activity, and that
the naphtofurandiones 32, 33 and 34 are not inactive because they

C. Antczak et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4528–4532
4531
Table 2
Cytotoxicity profile of benzofuran-4,5-dione derivatives
Compd
HL-60
HL-60/RV+
Jurkat
Molt-3
ALL-3
CWR22
HEK293
Y79
Meso47
1
16
21
17
32
48
17
10
72
10
53
69
42
36
5.4
100
22
27
67
67
40
8.1
8.9
57
20
13
12
46
6.1
55
67
28
5.9
2.8
45
7.3
9.1
47
4
11
11
31
4.1
40
71
25
7.0
2.9
N.D.
22
29
34
48
47
35
69
27
70
69
36
49
10
25
30
35
69
71
59
46
72
50
70
68
100
67
21
15
23
21
100
54
N.D.
30
31
67
100
74
63
100
66
68
66
100
63
100
71
16
17
20
28
19
21
5
22
25
23
26
24
27
71
48
38
40
36
100
100
100
100
100
72
100
100
37
100
43
100
100
64
100
30
100
100
100
67
100
64
33
55
12
16
IC50, lM ; 100: P100 lM; N.D.: not determined.
To our knowledge, benzofuran-4,5-diones constitute the first class
of compounds with such a specificity, since all currently known
HsPDF inhibitors are peptidomimetic- and hydroxamic acid-based
that inhibit both HsPDF and EcPDF6,8,24, and tend to exhibit low
specificity among metalloproteases.8–11 When we further charac-
signed novel and selective HsPDF inhibitors that have cell-based
anticancer activity.
In a pilot study, we assessed the in vivo efficacy of 27 in a mouse
xenograft model using human promyelocytic leukemia HL-60 cells.
Our lead compound delayed the growth of HL-60 tumors by up to
40% (Fig. 3). To put this result in perspective, actinonin – which has
a reported IC50 of 43 nM toward HsPDF7—must be administered
twice a day for two weeks at the large dose of 250 mg/kg to delay
tumor growth.7 This difference in potency could be attributed to a
better bioavailability compared to the peptidomimetic actinonin
but further studies are needed to address this question. In light
of this observation, our results are encouraging considering that
27, a benzofuran-4,5-dione derivative of first generation, was po-
tent in vivo at a dose of 15 mg/kg.
terized the selectivity profile of benzofuran-4,5-diones at 100 lM
in a panel of metalloproteases, we found that the newly synthe-
sized compounds are not only selective for HsPDF compared to
EcPDF, but also demonstrate selectivity for HsPDF among other
metalloproteases such as APN and MMP-1; as an example, 22
had no significant activity toward EcPDF and MMP-1, while being
potent toward HsPDF (IC50 = 15 lM) (Table 1, Fig. 2a). Similarly,
16, 17 and 28 completely inhibited HsPDF while having low activ-
ity toward EcPDF, APN and MMP-1 (Table 1). Altogether, our re-
sults demonstrate that benzofuran-4,5-diones constitute the first
In conclusion, our findings strongly suggest that derivatives of
the benzofuran-4,5-dione scaffold could constitute a new class of
potent antitumor agents selective for HsPDF.
known selective inhibitors of HsPDF. With an IC50 of 5.2 lM to-
ward HsPDF and no significant activity toward the bacterial
enzyme (Table 1, Fig. 2a), 27 is the most potent and selective HsPDF
inhibitor we have characterized.
To confirm that the novel HsPDF inhibitors we have identified
have no antibacterial activity as predicted by their lack of activity
toward EcPDF, we determined the MIC of compounds 5 and 22 in
a panel of 15 bacterial strains. We found that none of the com-
pounds had measurable antibacterial activity toward 14 out of
Acknowledgments
The authors thank Constantin Radu and the other members of
the High Throughput Screening Core Facility for their help during
the course of this study. The authors are greatful to Kumaraswamy
Sorra, Benarjee Velaga, Srinivas Oruganti and Narinder Mohal for
the chemical synthesis of the new derivatives described in this
study. The HTS Core Facility is partially supported by Mr. William
H. Goodwin and Mrs. Alice Goodwin and the Commonwealth Foun-
dation for Cancer Research, the Experimental Therapeutics Center
of MSKCC, the William Randolph Hearst Fund in Experimental
Therapeutics, the Lilian S Wells Foundation and by an NIH/NCI
Cancer Center Support Grant 5 P30 CA008748-44. This study was
funded by NIH grant 1R21NS57008 (HD).
the 15 strains tested (MIC >64 lg/mL), while actinonin was active
in 12 out of the 15 strains tested as expected (Supplementary Table
3). This important result confirms that benzofuran-4,5-diones con-
stitute the first known HsPDF inhibitors that specifically target the
human enzyme and lack antibacterial activity, presumably because
they are structurally different than actinonin and its derivatives and
do not contain a hydroxamic acid moiety.
To verify the hypothesis that HsPDF inhibitors may constitute
novel anticancer agents, we assessed the cytotoxicity profile of
the 13 novel benzofuran-4,5-dione derivatives in a panel of nine
well characterized cancer cell lines using a method previously
described.25 We found that all the benzofuran-4,5-dione deriva-
tives we have synthesized have cytotoxic activity toward at least
five out of nine of the cancer cell lines tested, with an IC50 rang-
Supplementary data
Supplementary data (supplementary tables, experimental pro-
cedures, spectroscopic data and NMR data for all compounds) asso-
ciated with this article can be found, in the online version, at
doi:10.1016/j.bmcl.2011.05.129.
ing from 2.8 to 74 lM (Table 2). Interestingly, the most potent
HsPDF inhibitor we have identified was also the most potent
compound in the viability assay: 27 was active across all the can-
References and notes
cer cell lines tested, with IC50 ranging from 2.8 to 37
lM, includ-
ing the multidrug resistant cell line HL-60/RV+ (IC50 = 12
(Table 2, Fig. 2).
In summary, our findings provide guidelines for the develop-
ment of selective HsPDF inhibitors active at the cellular level and
altogether our results validate our strategy, in that we have de-
l
M)
1. Mazel, D.; Pochet, S.; Marliere, P. EMBO J. 1994, 13, 914.
2. Meinnel, T.; Blanquet, S. J. Bacteriol. 1994, 176, 7387.
3. Meinnel, T.; Patiny, L.; Ragusa, S.; Blanquet, S. Biochemistry 1999, 38, 4287.
4. Serero, A.; Giglione, C.; Sardini, A.; Martinez-Sanz, J.; Meinnel, T. J. Biol. Chem.
2003, 278, 52953.

4532
C. Antczak et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4528–4532
5. Lee, M. D.; Antczak, C.; Li, Y.; Sirotnak, F. M.; Bornmann, W. G., et al Biochem.
Biophys. Res. Commun. 2003, 312, 309.
6. Leeds, J. A.; Dean, C. R. Curr. Opin. Pharmacol. 2006, 6, 445.
7. Lee, M. D.; She, Y.; Soskis, M. J.; Borella, C. P.; Gardner, J. R.; Hayes, P. A.; Dy, B.
M.; Heaney, M. L.; Philips, M. R.; Bornmann, W. G.; Sirotnak, F. M.; Scheinberg,
D. A. J. Clin. Invest. 2004, 114, 1107.
14. Atkinson, R.; Mare, P.; Larsen, D. J. Chem. Soc., Perkin Trans 2 1983, 271.
15. Singh, V.; Sapehiyia, V.; Kad, G. ChemInform 2003, 34.
16. Lopez-Alvarado, P.; Avendano, C.; Menendez, J. ChemInform 2003, 34.
17. Barrero, A.; Alvarez-Manzaneda, E.; Chahboun, R.; Diaz, C. Synlett 2000, 1561.
18. Lin, Y.; Lang, S., Jr. J. Org. Chem. 1980, 45, 4857.
19. Cheng, J. International patent WO/2005/011670.
8. Kontogiorgis, C.; Papaioannou, P.; Hadjipavlou-Litina, D. Curr. Med. Chem. 2005,
12, 339.
9. Saghatelian, A.; Jessani, N.; Joseph, A.; Humphrey, M.; Cravatt, B. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 10000.
10. Antczak, C.; Radu, C.; Djaballah, H. J. Biomol. Screen 2008, 13, 285.
11. Overall, C.; Kleifeld, O. Nat. Rev. Cancer 2006, 6, 227.
12. Escobar-Alvarez, S.; Goldgur, Y.; Yang, G.; Ouerfelli, O.; Li, Y.; Scheinberg, D. A. J.
Mol. Biol. 2009, 5, 1211.
13. Antczak, C.; Shum, D.; Escobar, S.; Bassit, B.; Kim, E.; Seshan, V. E.; Wu, N.;
Yang, G.; Ouerfelli, O.; Li, Y. M.; Scheinberg, D. A.; Djaballah, H. J. Biomol. Screen
2007, 12, 521.
20. Cheng, X.; Liu, X. J. Comb. Chem. 2007, 9, 906.
21. Al-Mousawi, S.; Mohammed Abdel-Khalik, M.; EI-Sherbiny, S.; John, E.; Elnagdi,
M. J. Heterocycl. Chem. 2001, 38, 949.
22. Grinev, A.; Khun’Shchi-tszyun’, A.; Terent’ev, A. J. Gen. Chem. USSR 1962, 32,
1931.
23. Grinev, A.; Arkhangel’skaya, N.; Uretskaya, G.; Vlasova, T. Chem. Heterocycl.
Compd. 1975, 11, 639.
24. Nguyen, K.; Hu, X.; Colton, C. Biochemistry 2003, 42, 9952.
25. Shum, D.; Radu, C.; Kim, E.; Cajuste, M.; Shao, Y.; Seshan, V. E.; Djaballah, H. J.
Enzyme Inhib. Med. Chem. 2008, 23, 931.