Synthesis and screening of small molecule inhibitors of anthrax edema factor

Article abstract of DOI:10.1016/j.bmcl.2008.05.059

The synthesis and development of a novel class of molecules that inhibit anthrax edema factor, an adenylyl cyclase, is reported. These molecules are derived from the initial discovery that histidine and imidazole adducts of the prostaglandin PGE₂ reduce the net secretory response of cholera toxin-challenged mice and act directly on the action of anthrax edema factor, a calmodulin-dependent adenylyl cyclase. The simple enones examined in this letter were prepared by palladium-catalyzed Suzuki reaction.

Full text of DOI:10.1016/j.bmcl.2008.05.059

Bioorganic & Medicinal Chemistry Letters 18 (2008) 4215–4218
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and screening of small molecule inhibitors of anthrax edema factor
Maria Estrella Jimenez a, Kathryn Bush b, Jennifer Pawlik b, Laurie Sower b, Johnny W. Peterson b,
Scott R. Gilbertson a,
*
a Chemical Biology Program, Department of Pharmacology and Toxicology, The University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-0650, USA
b Department of Microbiology and Immunology, The University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-1070, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and development of a novel class of molecules that inhibit anthrax edema factor, an aden-
ylyl cyclase, is reported. These molecules are derived from the initial discovery that histidine and imid-
azole adducts of the prostaglandin PGE2 reduce the net secretory response of cholera toxin-challenged
mice and act directly on the action of anthrax edema factor, a calmodulin-dependent adenylyl cyclase.
The simple enones examined in this letter were prepared by palladium-catalyzed Suzuki reaction.
Ó 2008 Elsevier Ltd. All rights reserved.
Received 15 April 2008
Revised 13 May 2008
Accepted 15 May 2008
Available online 20 May 2008
Keywords:
Antrhax
Edema factor
Adeylyl cyclase
Anthrax toxin produced by Bacilus anthracis, the causative agent
of anthrax, consists of three proteins; lethal factor (LF), protective
antigen (PA), and edema factor (EF). Working in concert these pro-
teins are responsible for the virulence associated with inhalation
anthrax infections, that if not promptly treated, are often fatal.1
Individually, the three proteins are not toxic, but the combination
of PA and LF forms lethal toxin (LT), which causes death when in-
jected into experimental animals.2 The combination of PA and EF
forms edema toxin (ET), which produces tissue edema at the site
of infection. Edema and lethal toxins synergize their action against
a host’s innate immunity. It has been shown that the deletion of
the EF or LF gene results in a reduction of virulence of anthrax bac-
teria.2 Given that EF is an adenylyl cyclase, edema toxin’s role in
pathogenesis is thought to impair phagocyte function. For that rea-
son small molecule inhibitors of EF may play a role in the treat-
ment of inhalation anthrax infections, as well as other diseases
that involve edema, due to increased cyclic AMP production. Here-
in we report the synthesis and testing of a series of inhibitors of an-
thrax edema factor.
mediate (Scheme 1). These adducts were found to be responsible
for the observed antisecretory activity. Subsequent studies re-
vealed that not only were the imidazole-prostaglandin adducts
active against mammalian adenylyl cyclases, but also the edema
factor of anthrax (IC50 = 21 lm). With these derivatives as the
lead compounds, an effort has been made to develop novel small
molecule inhibitors of EF through the synthesis of a series of re-
lated compounds.
With PGE2-L-histidine, and PGE2-imidazole in mind, a series of
simple enone adducts of imidazole were examined. One problem
that had to be considered in the development of new inhibitor
molecules was that the initial lead compounds were unstable to
elimination of the amine at the b-position with the subsequent ref-
ormation of PGA2 (6 Scheme 1). While it was understood that the
reversible nature of the amine adducts would present a problem,
the decision was made to examine such adducts, since they were
readily available. Interestingly, a number of these very simple com-
pounds (Fig. 2) exhibited adenylyl cyclase inhibition activity in the
same potency range as the PGE2 imidazole adducts (100–500 lM).
The initial lead for the development of the molecules re-
ported here came from the report by Peterson that histidine
and imidazole adducts of prostaglandin E2 (PGE2) (Fig. 1) re-
duced the net secretory response of the small intestine of mice
challenged with cholera toxin (CT).3 It was reported that upon
In the initial experiments the cyclohexenone adducts appeared to
be more active than the five-membered ring compounds. However,
these observations were not conclusive, since they may be biased
by the fact that the six-membered ring adducts appeared to be
more stable to elimination of imidazole than the five-membered
ring derivatives.
in vitro incubation of PGE2 with either L-histidine or imidazole,
a covalent adduct formed, presumably with PGA2 as the inter-
Because of the labile nature of the amine adducts, the decision
was made to examine molecules with the heterocycle attached via
a carbon–carbon bond. There were a number of methods that could
have been used in the synthesis of such molecules. The addition of
the imidazolium cuprate was attempted on cyclohexenone.4,5
* Corresponding author. Tel.: +1 409 772 9703; fax: +1 409 772 9700.
E-mail address: srgilber@utmb.edu (S.R. Gilbertson).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.05.059

4216
M. E. Jimenez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4215–4218
O
O
COOH
COOH
N
N
H2N
O
N
HO
HO
2 PGE2-imidazole
HO
N
1 PGE2-His
Figure 1.
O
O
O
O
COOH
COOH
COOH
OH
HN
+
N
NH2
4
N
N
HO
HO
or
HO
3 PGE2
HO
6 PGA2
R
HN
O
1 R = H
2 R =
N
5
OH
NH2
Scheme 1.
activities comparable to the saturated adducts. Consequently it
was decided to synthesize unsaturated versions of the primary tar-
get. This allowed the use of the Suzuki reaction for their synthesis
(Scheme 2). For ensembles of the desired molecules, the Suzuki
reaction appeared to be a good choice. There are over 500 commer-
cially available boronic acids and esters, and the necessary vinyl
bromo enones are available from 1,3-diones in a single step.6 The
ready availability of the two principal starting materials allowed
for a variety of different structural types to be synthesized and
tested (Table 1).
O
O
O
O
O
N
N
N
N
N
N
N
N
7
8
11
9
10
N
N
N
N
O
O
O
O
In our initial work we synthesized twenty-four enones. The Su-
zuki reaction was catalyzed by a variety of different palladium cat-
alysts with FC1007 providing the most consistent results.7,8 The
reactions were conducted under microwave conditions (250 W at
110 °C for 10 min). The yields ranged from low to quite good with
no attempt to optimize the reaction conditions. In all cases the
samples were purified by column chromatography to remove any
metal-based contaminants and other impurities.
The molecules were tested in whole cells to determine their
ability to inhibit the production of cAMP by anthrax edema factor.9
As can be seen methoxypyridine adducts of 2-methylcyclopente-
none (32 and 39) had good activity, while the analogous adduct
of cyclohexenone (27) was not active. The indole adduct of cyclo-
hexenone (35) also exhibited comparable activity to the best
compounds.
Through this limited study we have established some struc-
ture–activity relationships. It was determined that only cyclic ke-
tone and enone adducts are active acyclic versions of active
structures such as the imidazole adduct of methylvinyl ketone
are not. The ketone functionality appears to be necessary in that
molecules where the ketone has been reduced to an alcohol are
inactive. Additionally, only aromatic heterocyclic adducts are ac-
tive. The addition of simple amines did not provide inhibitors.
In summary, through the targeted synthesis of a small collec-
tion of compounds moderately active structures were obtained.
The Suzuki reaction of b-bromoenones with boronic acids provided
the desired compounds for screening. Additional versions of these
molecules, including structures with the prostaglandin side chains,
are being synthesized and evaluated.
N
NH
N
N
N
O
12
13
14
Figure 2.
15
However our initial use of cuprate chemistry demonstrated that
each cuprate addition required its own set of reaction conditions.
This precluded it as a method to provide ready access to a number
of different derivatives. Attempts to use the Heck reaction also
proved problematic. The Heck reaction between cyclohexenone
and vinyl or arylbromides often provided mixtures of saturated
and unsaturated products. These materials proved to be difficult
to separate, and consequently this approach was not considered
appropriate for a system where ready access to a variety of com-
pounds was desired.
Despite the initial problems with the Heck reaction, a number of
select compounds were synthesized and assayed. We observed
that the unsaturated versions of these molecules had inhibitory
1.) 3 mol% FC1007
O
O
K2CO3 (aq. 1.2 equiv)
N
N
Microwave, 110 ºC,10 min
+
55%
B
N
Br
O
18
O
N
16
17
Scheme 2.

M. E. Jimenez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4215–4218
4217
Table 1
Yield and select activity data for enone adducts
1.) 3 mol% FC1007
K2CO3 (aq. 1.2 equiv)
O
O
Microwave, 110ºC, 10 min
+
(HO)2B-Heterocycle
Br
n
Heterocycle
n = 0 or 1
n
Yielda (%)
IC50
(l
M)
Compound
Yielda (%)
IC50 (lM)
b
b
Compound
O
O
O
22
75
>100
34
88
> 100
N
O
23
54
>100
35
83
56
N
H
N
O
O
O
O
24
83
89
>100
36
37
85
64
53
O
O
O
*
25
>100
S
OH
O
O
26
29
>100
38
95
>100
CH3
N
N
H
N
O
O
O
O
O
O
27
28
29
30
31
89
28
72
83
16
>100
>100
>100
>100
>100
39
40
41
42
43
69
78
16
31
92
30
N
OCH3
N
OCH3
*
N
S
N
*
*
*
N
N
S
O
O
O
O
S
N
N
O
H
(continued on next page)

4218
M. E. Jimenez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4215–4218
Table 1 (continued)
b
b
Compound
Yielda (%)
IC50
24
(
lM)
Compound
Yielda (%)
IC50 (lM)
O
O
O
32
92
44
96
>100
35
N
OCH3
O
O
OCH3
33
88
30
45
98
OCH3
*
IC50 was not determined.
a
Spectral data for compounds the purification information is available in the supplementary material.
b
The assay was performed with murine monocyte/macrophage (RAW 264.7) cells. Media was aspirated from the cells and replaced with varying concentrations of PGE2-
M) containing PA and EF and incubated for 4 h at 37 °C in 5% CO2. Following the incubation period, culture
supernatants were removed (extracellular cAMP) and transferred to a 96-well plate for cAMP determination by ELISA.
imidazole and test inhibitors (100, 50, 10, 5, 1, and 0.1
l
8. Sauer, D. R.; Kalvin, D.; Phelan, K. M. Org. Lett. 2003, 5, 4721.
Acknowledgment
9. Cell Culture: Murine monocyte/macrophage cells (RAW 264.7) were propagated
in T75 flasks containing Dulbecco’s modified Eagle’s medium (DMEM)
(Mediatech, Inc., Herndon, VA) at 37 °C with 5% CO2. The culture media
This work was supported by the Mission Pharmacal and the
Robert A. Welch Foundation and Mission Pharmcal.
contained 10% fetal bovine serum (FBS), 100
-glutamine.
lg/ml penicillin/streptomycin and
L
Cell assay: Murine monocyte/macrophage (RAW 264.7) cells (ATCC, Manassas,
VA) were plated at 1 Â 106 cells per well in DMEM containing 10% fetal bovine
Supplementary data
serum (FBS), 5 mL of 100 lg/ml penicillin/streptomycin and 5 mL of 2 mM L-
glutamine (Mediatech, Herndon, VA) in 48-well tissue culture plates. The cells
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2008.05.059.
were then allowed to adhere to the wells overnight at 37 °C in 5% CO2. PGE2-
imidazole, Protective Antigen (2.5 lg/mL PA), edema factor (0.625 lg/mL EF)
and test inhibitors were reconstituted using sterile water and then diluted to
the desired concentrations using DMEM assay media (without phenol red)
containing 50 lM of 3-isobutyl-1-methylxanthine (IBMX) (BIOMOL, Plymouth
Meeting, PA). Media was aspirated from the cells and replaced with varying
concentrations of PGE2-imidazole and test inhibitors (100, 50, 10, 5, 1, and
References and notes
1. Brachman, P. S. Anthrax; Harper & Rowe: New York, 1972.
2. Brossier, F. M.; Weber-Levy, M.; Mock, M.; Sirard, J.-C. Infect. Immun. 2000, 68,
1781.
3. Peterson, J. W.; King, D.; Ezell, E. L.; Rogers, M.; Gessell, D.; Hoffpauer, J.; Reuss,
L.; Chopra, A. K.; Gorenstein, D. Biochim. Biophys. Acta 2001, 1537, 27.
4. Tatsuta, K.; Miura, S.; Ohta, S.; Gunji, H. J. Antibiot. 1995, 48, 286.
5. Kirk, K. L. J. Org. Chem. 1978, 43, 4381.
0.1
lM) containing PA and EF and incubated for 4 h at 37 °C in 5% CO2.
Following the incubation period, culture supernatants were removed
(extracellular cAMP) and transferred to
determination.
a new 48-well plate for cAMP
cAMP determination: The extracellular cAMP concentration in the supernatants
was measured with a cAMP-specific ELISA from Assay Designs, Inc. (Ann Arbor,
Michigan) per manufacturer directions.
6. Piers, E.; Nagakura, I. Sny. Commun. 1975, 5, 193.
7. Wang, Y.; Sauer, D. R. Org. Lett. 2004, 6, 2793.