The discovery and structure-activity relationships leading to CE-156811, a difluorophenyl cyclopropyl fluoroether: A novel potent antibacterial analog derived from hygromycin A

Article abstract of DOI:10.1016/j.bmcl.2010.11.022

SAR studies and optimization of various modified Hygromycin A fluoroalkyl ethers, which led to the discovery of the highly potent 4′-(2-cyclopropyl- 2-fluoroethyl ether) antibacterial CE-156811 (1) derived from truncation of the ribose ring and difluorination of the phenyl found in Hygromycin A, are discussed.

Full text of DOI:10.1016/j.bmcl.2010.11.022

Bioorganic & Medicinal Chemistry Letters 21 (2011) 276–279
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery and structure–activity relationships leading to CE-156811,
a difluorophenyl cyclopropyl fluoroether: A novel potent antibacterial analog
derived from hygromycin A
Phuong T. Le a, , Steven J. Brickner c, Sarah K. Wade b, Katherine Brighty b, Rhonda Monahan b,
⇑
Gregory G. Stone d, Dennis Girard e, Steve Finegan b, Joan Duignan b, John Schafer b, Meghan Maloney b,
Richard P. Zaniewski b, Ann G. Connolly b, Jennifer Liras b, Jon Bordner b, Ivan Samardjiev b
a Pfizer Worldwide Research & Development, La Jolla Laboratories, Pfizer Inc., La Jolla, CA 92121, United States
b Pfizer Worldwide Research & Development, Groton/New London Laboratories, Pfizer Inc., Groton, CT 06340, United States
c SJ Brickner Consulting, LLC, Ledyard, CT 06339, United States
d AztraZeneca Pharmaceuticals, Waltham, MA 02451, United States
e Dennis Girard, Preclinical Pharmacology & PK/PD Consulting, Salem, CT 06420, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
SAR studies and optimization of various modified Hygromycin A fluoroalkyl ethers, which led to the dis-
covery of the highly potent 40-(2-cyclopropyl-2-fluoroethyl ether) antibacterial CE-156811 (1) derived
from truncation of the ribose ring and difluorination of the phenyl found in Hygromycin A, are discussed.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 3 September 2010
Revised 31 October 2010
Accepted 2 November 2010
Available online 10 November 2010
Keywords:
Hygromycin
CE-156811
Antibacteria
Infections with bacterial pathogens that have developed resis-
tance to a number of important classes of antibiotics continue to
present highly challenging treatment problems for physi-
cians.1,2,8–10 As a consequence, discovery of novel antibacterial
agents to combat these resistant strains remains of high interest
due to high medical need.
The natural product Hygromycin A, discovered in 1953, has only
weak activity against medically important Gram-positive organ-
isms.3,4 It demonstrates potent activity against the organism impli-
cated in swine dysentery. From mode of action studies carried out
a number of years ago, it was established that Hygromycin A inter-
feres with bacterial growth by inhibition of protein synthesis
involving inhibition of peptide bond formation.5 Our study con-
cerning Hygromycin A analogs as potential new leads within the
Human Health anti-infective group at Pfizer was initiated as a con-
sequence of an intensive examination of a broad selection of
known natural products published in the literature. Earlier efforts
reported a series of modified Hygromycin A analogs, resulting from
investigations during a Pfizer Animal Health research program
that was ultimately terminated. Hecker, Jaynes, and coworkers
established that the Hygromycin A furanose ring could be effec-
tively replaced by simple alkyl ether substituents, for example,
40-allyl ether, and that the phenol hydroxyl could be replaced with
fluorine. The 2,5-difluorophenyl analog in particular demonstrated
improved activity against the animal pathogens of interest, as well
as significant improvement against key Gram-positive human
pathogens.6,7
CE-156811 (1) is a novel truncated Hygromycin A derivative
possessing potent in vitro activities against important Gram-posi-
tive bacterial pathogens. It also demonstrated efficacy commensu-
rate with linezolid in several in vivo infection models with
subcutaneously (sc) dosing (Fig. 1).
The preparation of
5 started with condensation between
phosphonate 37 and commercially available aldehyde 2 under
basic conditions (Scheme 1). Different ether substituents at the
4-aryl position of 5 were appended in a reasonably straightforward
manner using nucleophilic aromatic substitution with a variety of
alcohols on the trifluorophenyl intermediate 4. This typically gave
the desired para substituted analogs as the primary products in
low-to-modest (15–45%) yields; minor amounts of the product of
ortho substitution were also isolable. The selectivity can be
inferred to arise from steric hindrance at the ortho site. As the
isomers were separable by chromatography, the use of this method
allowed us to prosecute the SAR in an expeditious manner.
⇑
Corresponding author.
E-mail address: phuong.le@pfizer.com (P.T. Le).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.11.022

P. T. Le et al. / Bioorg. Med. Chem. Lett. 21 (2011) 276–279
277
O
O
O
HO
O
HO
O
O
F
OH
N
H
HO
O
OH
N
H
H
OH
F
O
OH
F
O
O
OH
HO
HYGROMYCIN A
CE-156811 (1)
Figure 1. The structures of Hygromycin A and CE-156811 (1).
O
O
HO
O
O
O
P
EtO
O
HO
H
O
EtO
OH
N
H
F
F
F
F
OH
O
OH
2-20 eq ROH
KOt-Bu, DMAC, 70oC
N
H
3
OH
KOt-Bu
t-BuOH
F
F
4
2
O
O
HO
O
F
OH
N
4
OH
H
5
F
RO
Scheme 1. Synthesis of 2,5-difluorophenyl 4-ethers.
Using this approach, the compounds in Table 1 were prepared.
In the acyclic 40-alkyl ether series, as exemplified in the generic
structure 5, going from ethyl ether 6 to hydroxyl ethyl ether 7,
potency dropped across species. However, when fluorine was
introduced in place of the hydroxyl found in 7 to give 8, the
potency was regained; 3-fluoropropyl ether 9 provided the optimal
antibacterial activity in this acyclic series. Based on this finding, we
prepared additional 3-fluoropropyl analogs in an attempt to
broaden the SAR. These efforts were generally unsuccessful, as a
number of additional compounds demonstrated an overall weaker
antibacterial activity spectrum (e.g., compounds 10 and 11).
A more promising avenue became evident when we examined
the addition of substituents to the C2-carbon of the 2-fluoroethyl
ether moiety of 8; it is to be noted that the parent 2-fluoroethyl
ether itself has only modest antibacterial activity. However, place-
ment of small alkyl moieties at this site proved to considerably
boost the in vitro potency; for example, 12 demonstrated an
8- to 20-fold improvement in activity compared with 8. In a similar
manner, the importance of fluorination at the 2-ethyl ether posi-
tion in potentiating activity was demonstrated; for example, 13
is 2-fold more potent than cyclobutylmethyl ether 14 (Table 2).
Due to these observations, we hypothesized that there was
potentially a small binding pocket at the site of activity that
accommodated these small alkyl substituents and the 2-fluoro
heteroatom on the ethyl ether template. We therefore targeted a
broader examination of related compounds to better understand
Table 1
In vitro antibacterial activities (MIC,
l
g/mL) of early analogs
Compound
R
S. aureus 1095a
E. faecalis 1085b
E. faecium 1022c
S. pneumo 1016d
6
100
100
25
12.5
NA
O
HO
7
8
9
>100
25
>100
25
NA
6.2
2
O
F
50
O
F
1.56
1.56
0.78
O
F
F
F
10
11
12.5
100
12.5
100
3.13
50
1.56
25
O
O
a
b
c
S. aureus, Staphylococcus aureus.
E. faecalis, Enterococcus faecalis.
E. faecium, Enterococcus faecium.
d
S. pneumo, Streptococcus pneumoniae.

278
P. T. Le et al. / Bioorg. Med. Chem. Lett. 21 (2011) 276–279
Table 2
In vitro antibacterial activities (MIC,
l
g/mL) of follow-up analogs
Compound
R
S. aureus 1095
E. faecalis 1085
60.2
E. faecium 1022
S. pneumo 1016
60.2
F
12
6.25
3.13
O
O
F
13
14
6.2
6.2
6.2
6.2
3.1
6.2
12.5
12.5
3
O
F
15
1.56
1.6
1.6
0.4
O
a
OH
O
MeO
MeO
OCH2Ph
MeO
MeO
b
MeO
c
O
OCH2Ph
MeO
OMe
19
16
17
18
H
F
F
F
e
g
f
OH
d
O
+
HO
OCH2Ph
OCH2Ph
F
F
22
20
21
O
O
H
O
HO
O
F
O
h
HO
O
O
F
O
EtO
P
EtO
N
H
OH
OH
F
+
O
F
N
OH
F
O
F
H
OH
HO
15
3
23
chiral separation
O
O
O
+ 32
F
OH
N
H
H
OH
F
F
O
1
Scheme 2. First generation synthesis of 1. Reagents and conditions: (a) LAH, THF, ꢀ78 °C to rt (78%); (b) benzyl bromide 60% NaH, DMF; (c) 1 N HCl, MeOH/H2O (1:3), rt; (d)
LAH, THF, ꢀ78 °C to rt (44% for three-steps); (e) DAST, CH2Cl2, ꢀ78°C to rt (50%); (g) K2CO3, 65 °C, neat (22%); (h) KOt-Bu, t-BuOH, rt (36%).
F
HO
Br
O
O
b
c
O
a
O
+
H
24
25
O
O
O
H
H
F
F
F
Chiral Separation
d
e
OH
O
O
H
F
O
Cl
+
27
26
28
29
O
O
O
O
HO
HO
H
O
O
F
g
OMs
f
F
F
OH
N
H
O
N
H
+
H
H
OH
OH
F
F
HO
F
O
1
31
30
Scheme 3. Second generation synthesis of 1. Reagents and conditions: (a) tert-butyllithium in pentane THF, ꢀ78 °C to rt; (b) DAST, CH2Cl2 (68–73% for two-steps); (c) 10% Pd/C,
2 , MeOH, rt; (d) TEA, CH2Cl2, rt (87% for two-steps); (e) K2CO3, MeOH, H2O, rt; (f) MsCl, TEA, CH2Cl2, rt; (g) K2CO3, DMF, 75 °C, (33.6% yield for three-steps).
H
the SAR associated with other fluorinated cyclobutyl analogs, and
search for improved activity.
sual but precedented13 rearrangement to give cyclopropyl fluoro-
ethyl ether 21 (Scheme 2). This compound represented an
attractive potential side chain for our examination. Preparation of
1 is shown in Scheme 2.
Cyclobutanol 20 was readily afforded by LAH reduction of the
cyclobutanone 19. This requisite ketone was obtained in a straight-
forward three-step transformation of the known cyclobutanone
One of these targeted cyclobutyl analogs had
a fluorine
introduced at the 3-position in the cyclobutyl methyl ether 14.
We envisioned this would be readily accessible in a straightfor-
ward manner from reaction of the corresponding alcohol 20 with
DAST. In practice, however, this manipulation resulted in an unu-

P. T. Le et al. / Bioorg. Med. Chem. Lett. 21 (2011) 276–279
279
Table 3
Table 5
In vitro activity of 1 [MIC90
(
l
g/ml)]
In vivo efficacy of 1 (sc dosing) versus linezolid, (p.o. dosing) PD50
(mg/kg)
Organism
Compound 1
Linezolid
1
Linezolid
MSSA (28)a
MRSA (18)b
GISA (6)c
1
0.5
0.125–0.5
0.25
0.5
0.5
1
2
1
Acute S. aureusa
4
8
4–7
4–8
3–5
10–13
3
0.5–1
Acute E. faeciuma
Acute E. faecalisa
Lung S. pneumoniaeb
Acute S. pyogenesa
MSSE (23)d
1
1
1
2
10
23
3
PRSP (21)e
E. faecalis (27)
E. faecium (21)
a
Dosed BID for one day.
Dosed BID for three days.
a
b
Methicillin-sensitive S. aureus (MSSA).
Methicillin-resistant S. aureus (MRSA).
Glycopeptide-intermediate S. aureus (GISA).
b
c
d
e
Methicillin-sensitive Staphylococcus epidermidis (MSSE).
Penicillin-resistant Streptococcus pneumoniae (PRSP).
dosed subcutaneously (sc) BID for one day and three days
(Table 5). When compared to vancomycin, linezolid, and levoflox-
acin, 1 demonstrated activity comparable to that of linezolid.12
Conclusions: We report the SAR in the 40 position of the difluo-
rophenyl ring of truncated Hygromycin A derivatives that led to
the synthesis of 1, a novel and highly potent inhibitor of Gram-po-
sitive bacteria with a very good PK profile in multiple animal spe-
cies and excellent in vivo efficacy. Unfortunately, further
development with this compound was precluded upon finding that
1 was insufficiently stable in water. When 1 was studied in water,
it degraded to the extent of 4% per day at rt. After 5.5 days, com-
pound 1 showed almost 25% conversion of the cyclopropyl fluoro-
ethyl ether to the corresponding cyclopropyl hydroxyethyl ether.12
Further lead optimization in the aminocyclitol has recently been
reported.15
Table 4
In vivo PK data for 1
PK species
Cl (mL/
min/kg)
Vdss
(L/kg)
T1/2
(h)
F%a
Plasma protein
binding (% unbound)c
Mouse
Rat
Monkey
Dog
12.1
10.5
8.6
0.52
0.5
0.59
0.71
1.4
1.3
1.0
1.9
55
10.4
10.7
17.2
18.6
56
NDb
ND
5.4
a
b
c
F% oral bioavailability. Oral dose in rat = 10 mg/kg, iv dose = 5 mg/kg.
ND = Not determined.
CE-156811 concentration = 3 lg/ml.
Acknowledgement
ketal ester 16.11 Upon hydrogenation of 21 (Pd/C, H2) to give the
alcohol 22, nucleophilic aromatic substitution of the alcohol 22
onto the 2,4,5-trifluorobenzaldehyde gave the 4-substituted ether
benzaldehyde 23 after separation of isomers. This condensed with
the phosphonate 3 under basic conditions to give 15 as a mixture
of diastereomers. Compound 15 demonstrated highly potent anti-
bacterial activity (Table 2). Upon chiral column HPLC purification,
diastereomer mixture 15 was separated to give 1 and its corre-
sponding diastereomer 32. The absolute stereochemistry of the
most active of these two (1) was inferred from an X-ray crystal
structure.14
Because of the overall low yield and the fact that the chiral sep-
aration was done in the last step in this original route, a second
generation synthesis was executed (Scheme 3). The approach
began with t-BuLi metallation of bromocyclopropane and addition
to 2-(benzyloxy)acetaldehyde to afford the secondary alcohol 24.
Compound 24 was treated with DAST in methylene chloride to give
the desired fluorinated benzyl ether 25 in 68–73% yield over the
previous two steps. After hydrogenation of benzyl ether 25,
2-cyclopropyl-2-fluoroethanol 26 was then converted to benzoate
27 for ease of chiral separation, serving to both provide a UV
chromophore and being less prone to loss during evaporative
workup of column fractions than the more volatile alcohol 26. After
saponification of 28, the unstable and low boiling alcohol was
immediately converted to the mesylate 30, which was then used
to alkylate the difluorophenol 31 shown to provide 1.
The authors thank Dr. Kevin Freeman-Cook for reviewing the
manuscript.
References and notes
1. Levy, S. B.; Marshall, B. Nat. Med. 2004, 10, 2122.
2. Norrby, R. S.; Nord, C. E.; Finch, R. Lancet Infect. Dis. 2005, 5, 115.
3. Pittenger, R. C.; Wolfe, R. N.; Hoehn, P. N.; Daily, W. A.; McGuire, J. M. Antibiot.
Chemother. 1953, 3, 1268.
4. Mann, R. L.; Gale, R. M.; Van Aberlee, R. F. Antibiot. Chemother. 1953, 3, 1279.
5. Guerrero, M. D.; Modolell, J. Eur. J. Biochem. 1980, 107, 409.
6. Jaynes, B. H.; Elliott, N. C.; Schicho, D. L. J. Antibiot. 1992, 45, 1705.
7. Hecker, S. J.; Cooper, C. B.; Blair, K. T.; Lilley, S. C.; Minich, M. L.; Werner, K. M.
Bioorg. Med. Chem. Lett. 1993, 3, 289.
8. Levy, S. B.; Marshall, B. Nat. Med. 2004, 10, S122.
9. Tomasz, A. N. Engl. J. Med. 1994, 330, 1247.
10. Fridkin, S. K.; Edwards, J. R.; Courval, J. M.; Hill, H.; Tenover, F. C.; Lawton, R.;
Gaynes, R. P.; McGowan, J. E., Jr. Ann. Intern. Med. 2001, 135, 175.
11. (a) Pigou, P. E.; Schiesser, C. H. J. Org. Chem. 1988, 53, 3841; (b) Christian
Stevens, C.; De Kimpe, N. J. Org. Chem. 1996, 61, 2174.
12. Stone, G. G.; Girard, D.; Finegan, S.; Duigan, J.; Schafer, J.; Maloney, M.;
Zaniewski, R.; Brickner, S. J.; Wade, S. K.; Le, P. T.; Huband, M. D. Antimicrob.
Agents Chemother. 2008, 52, 2663.
13. Kirihara, M.; Takuwa, T.; Kambayashi, T.; Momose, T.; Takeuchi, Y. J. Chem. Res.
(S) 1998, 652.
14. An X-ray crystal structure was obtained from the following structure, which
was obtained from reacting the isomer of 29 with the corresponding
substituted benzoate ester.
Compound 1 demonstrated potent in vitro activity across multi-
ple drug-resistant organisms S. aureus, S. pneumoniae, E. faecium,
E. faecalis and was comparable with linezolid (Table 3).
In in vivo studies across the four species examined (mouse, rat,
monkey, and dog), 1 exhibited a good PK profile (Table 4). The
O
F
O
N
O
H
S
O
O
human predicted T based upon simple allometric scaling is
approximately 2 h.
In vivo efficacy studies also showed that 1 had comparable
activity to linezolid across several different acute models when
½
15. Visser, M. S.; Freeman-Cook, K. D.; Brickner, S. J.; Brighty, K. E.; Le, P. T.; Wade,
S. K.; Monahan, R.; Martinelli, G. J.; Blair, K. T.; Moore, D. E. Bioorg. Med. Chem.
Lett. 2010, 6730.