Potent and selective inhibitors of bacterial methionyl tRNA synthetase derived from an oxazolone-dipeptide scaffold

Article abstract of DOI:10.1016/j.bmcl.2004.01.094

The preparation and structure-activity relationships (SARs) of potent and selective small molecule inhibitors of bacterial methionyl-tRNA synthetase (MetRS) derived from an oxazolone-dipeptide scaffold are described. Examples combine Staphylococcus aureus MetRS (SaMetRS) potency with selectivity over human MetRS. As a result of the SAR expansion compound 14a was identified, as a potent SaMetRS inhibitor (IC₅₀=18 nM) having moderate inhibition of MetRS derived from Enterococci faecalis (IC₅₀=3.51 μM).

Full text of DOI:10.1016/j.bmcl.2004.01.094

Bioorganic & Medicinal Chemistry Letters 14 (2004) 1909–1911
Potent and selective inhibitors of bacterial methionyl tRNA
synthetase derived from an oxazolone–dipeptide scaffold
Manish Tandon,a David L. Coffen,a Paul Gallant,b Dennis Keithb
and Mark A. Ashwella,*
aArQule Inc., 19 Presidential Way, Woburn, MA 01801, USA
bCubist Pharmaceuticals Inc., 65 Hayden Ave., Lexington, MA 02421, USA
Received 12 December 2003; revised 27 January 2004; accepted 28 January 2004
This communication is dedicated to the memory of David L. Coffen
Abstract—The preparation and structure–activity relationships (SARs) of potent and selective small molecule inhibitors of bacterial
methionyl-tRNA synthetase (MetRS) derived from an oxazolone–dipeptide scaffold are described. Examples combine Staphylo-
coccus aureus MetRS (SaMetRS) potency with selectivity over human MetRS. As a result of the SAR expansion compound 14a was
identified, as a potent SaMetRS inhibitor (IC50=18 nM) having moderate inhibition of MetRS derived from Enterococci faecalis
(IC50=3.51 mM).
# 2004 Elsevier Ltd. All rights reserved.
The increasing frequency of bacterial resistance and the
associated rise in healthcare costs to successfully treat
patients with infections due to resistant bacteria is pro-
viding renewed impetus for the pursuit of antimicrobial
agents with novel modes of action. Methicillin-resistant
Staphylococcus aureus (MRSA) and vancomycin resis-
tant enterococci (VRE) are two members of an increas-
ingly long list of resistant Gram-positive pathogens.1
We have screened a number of our small molecule
libraries against such bacterial targets. One campaign
targeted bacterial amino acyl–tRNA synthetases, which
were available for screening at Cubist Pharmaceuticals,
Inc. These enzymes are essential for bacterial growth,
being responsible for the charging of tRNA molecules
with the cognate amino acids required in an essential
step of protein synthesis.2 The clinical antibiotic mupir-
ocin (pseudomonic acid), a topically active agent for the
eradication of nasal carriage of MRSA,3 is an inhibitor
of bacterial isoleucyl tRNA synthetase.4 This paper
describes work and early SAR investigations of a novel
series of potent and selective small molecule inhibitors
of methionyl tRNA synthetase (MetRS) derived from a
high-throughput screening hit.
MetRS from Staphylococcus aureus (SaMetRS). Sig-
nificantly, follow-up screening of 1a and b showed them
to be selective for SaMetRS versus human MetRS
(hMetRS).5 In the case of 1a the selectivity for
SaMetRS was more than 100-fold (Table 1). Thus, these
novel low molecular weight hits represented a promising
starting point for SAR expansion.
Three regions of structural diversity were explored
simultaneously in the first phase of SAR determination
as shown in Figure 1. As illustrated in Scheme 1, reac-
tion of the aromatic carboxylic acids 2 (activated as the
acid chlorides) with glycine provided glycine amides 3.
Cyclization of 3 with a tri-substituted orthoformate in
acetic anhydride at elevated temperature provided the
High-throughput screening identified two oxazolone–
dipeptides (1a and b) as low micromolar inhibitors of
* Corresponding author. Tel.: +1-781-994-0598; fax: +1-781-994-
0690; e-mail: mashwell@arqule.com
Figure 1. HTS hits from small molecule library screening.
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.01.094

1910
M. Tandon et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1909–1911
Table 1. Inhibition of MetRS by compounds with linker modifications
Compd
R1
R2
Linker
IC50 (mM)
SaMetRS
EfMetRS
hMetRS
1a
1b
6a
6b
7a
7b
8a
8b
9a
9b
Cl
Br
Cl
Br
Cl
Br
Cl
Br
Cl
Br
Cl
H
0.68
14.0
48.0
0.73
>100
1.1
5%@10 mM
>100
Cl
H
0.07
.4
0.511
>100
Cl
H
73%@10 mM
55%@10 mM
43%@10 mM
39%@10 mM
3%@10 mM
1%@10 mM
60%@10 mM
7%@10 mM
2%@10 mM
ia
88%@10 mM
19%@10 mM
20%@10 mM
23%@10 mM
5%@10 mM
6%@10 mM
Cl
H
Cl
H
5%@10 mM
7%@10 mM
In the next phase of SAR expansion alternatives to
l-proline in Region 2 were explored. Four building
blocks were prepared using standard peptide coupling
methodology. Region 1was limited to two variants (2,4-
dichloro- and 4-bromophenyl) and Region 3 was fixed
as an l-tyrosine amide. The four l-tyrosine amides were
prepared from l-azetidine carboxylic acid, pipecolic,
nipecotic, and isonipecotic acids.
In order to provide an initial assessment of the Gram-
positive enzymatic spectrum for this and subsequent hit
expansion libraries, screening was extended to include
the MetRS from Enterococcus faecalis (EfMetRS). The
results are presented in Table 1. Ring expansion to a 6-
membered ring resulted in a substantial loss of activity
against SaMetRS and ring contraction resulted in a 5–
10-fold decrease in IC50 against SaMetRS with both
l-azetidine compounds (6a and 6b) displaying moderate
activity for EfMetRS. In addition no significant activity
against hMetRS was observed for the l-azetidine and
l-proline analogues and we therefore concentrated our
attention on additional structural variants containing
these amino acids.
Scheme 1. (a) Oxalyl chloride, triethylamine, CH2Cl2, then glycine; (b)
R6C(OEt)3, Ac2O, reflux; (c) THF, DMSO, R4R5NH, 50 ꢀC.
key intermediate substituted 4-ethoxy-5-oxazolones 4.6
The oxazolones 4 were then treated with a series of
dipeptides, amino acids and amino acid amides to pro-
vide 1 in good yields.7
Key structural features were readily determined. In
Region 1only 2,4-dichloro- or 4-bromo phenyl sub-
stiutents were tolerated. Of the more than 100 dipep-
tides used in the libraries the preferred Region 2 amino
acid was the cyclic amino acid l-proline. When the tyr-
osine in Region 3 was replaced with tryptophan or
In the final phase of SAR expansion, the effect of the
exomethylene (R6=H) was investigated. A series of 4-
ethoxy-5-oxazolones 4 were prepared, maintaining the
4-bromo and 2,4-dichlorophenyl substitution pattern in
Region 1while the R 6-substituent was varied in size
(methyl, ethyl, n-propyl) and electronic character (phenyl).
Both l-azetidine and l-proline dipeptide amides were
phenylalanine
a substantial loss in activity was
observed. Additionally, the importance of the C-term-
inal of the initial hits 1a and b was confirmed when the
corresponding acids were screened (1c, R1=R2=Cl,
R3=OH; SaMetRS IC50=3.1 mM and 1d, R1=Br,
R2=H, R3=OH; SaMetRS IC50=9.5 mM).

M. Tandon et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1909–1911
1911
Table 2. Inhibition of MetRS by l-azetidine and l-proline sub-
stituted at R6
R6=H is tolerated since 6a remains the most potent
inhibitor of the set (EfMetRS IC50=730 nM, Table 2).
The most active SaMetRS inhibitor identified in this
SAR exploration was 14a (SaMetRS IC50=18 nM).
The compound had low micromolar activity at
EfMetRS and was devoid of activity at hMetRS.
In conclusion, novel low molecular weight high-
throughput screening hits 1a and b were the starting
point for a three-phase SAR expansion. Successive
improvements in activity against SaMetRS led to 14a,
which had an IC50 of 18 nM against SaMetRS. This was
a 40-fold improvement in IC50 from the original hit 1a.
A 4-fold improvement against the bacterial MetRS
enzyme from the Gram-positive pathogen Enterococuss
faecalis was also obtained for 14a. In addition small
modifications to the structures of 1a and b were shown
to elicit changes in inhibition at the two enzyme targets
and as such may provide additional insight into the
putative binding pockets of MetRS derived from S.
aureus and E. faecalis. Disappointingly compounds
from this series lacked cellular activity and were not
pursued further. However, 14a and the series from
which it was obtained represent a new class of potent
and selective inhibitors of S. aureus methionyl tRNA
synthetase and may provide the basis for the design of
therapeutically useful antibacterials.
Compd R1 R2
R6
n
IC50 (mM)
SaMetRS EfMetRS hMEtRS
10a
10b
11a
11b
12a
12b
13a
13b
14a
14b
15a
15b
Cl
Br
Cl
Br
Cl
Br
Cl
Br
Cl
Br
Cl
Br
Cl
H
Cl
H
Cl
H
Cl
H
Cl
H
Cl
H
Me
Me
Et
2
2
2
2
2
2
2
0.34
1.67
0.18
0.85
0.17
0.76
0.43
0.816.1
6.63
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
1.28
5.93
99.0
3.11
27.4
Et
nPr
nPr
Ph
Ph
Et
Et 7 10.1
nPr 10.04
nPr 10.087
2 4
810.01
3.51
99
3.52
>100
retained. The screening results for this small array of
compounds are shown in Table 2.
Generally increasing the size of the R6 substituent is
tolerated with all analogues showing comparable or
improved potency against SaMetRS and no activity at
hMetRS. Where Region 2 was set at l-proline both the
2,4-dichlorophenyl and the 4-bromophenyl analogues
are most potent when the R6 substituent is an n-propyl
group 12a, SaMetRS IC50=170 nM and 12b, SaMetRS
IC50=760 nM. However, 13b (R6=Ph) has a compar-
able IC50 to 12b (IC50=810 nM) suggesting that the 4-
bromophenyl analogue is more tolerant of an aromatic
ring at R6 than is the case with corresponding 2,4-
dichlorophenyl analogue (13a, IC50=430 nM). In this
case there is a more than a two-fold loss in activity.
Acknowledgements
We thank Drs. M. Harris for the preparation of the
general screening library and P. Pallai for managerial
support.
References and notes
1. Bush, K.; Macielage, M.; Clancy, J. Emerg. Drugs 2000, 5,
347.
2. Ibba, M.; Soll, D. Annu. Rev. Biochem. 2000, 69, 617.
¨
The most potent SaMetRS inhibitors were found when
Region 2 was the l-azetidine linker. This holds true for
both the 2,4-dichloro- and the 4-bromophenyl analo-
gues. However, the 2,4-dichlorophenyl is present in the
two most potent analogues, 14a and 15a. Both com-
pounds have IC50’s below 50 nM.
3. Bertino, J. S.; Am, J. Health Syst. Pharm. 1997, 54, 2185.
4. Parenti, M. A.; Hatfield, S. M.; Leyden, J. J. Clinical
Pharmacy 1987, 6, 761.
5. Enzymatic inhibition was determined by measuring the
amount of radiolabeled methionine incorporated into the
product, the charged methionyl–tRNA complex. The
enzymatic reaction was run at Km concentrations for ATP
and methionine and a saturating level of 90 mM for
tRNA. A compound’s IC50 was determined by fitting the
results from a 10-point dose–response curve. All results
are the average of at least two measurements.
While none of the analogues in this third set showed
sub-micromolar EfMetRS inhibition the observed SAR
shows that when Region 2 is l-proline an R6 substituent
the size of an ethyl group confers maximal EfMetRS
inhibition (11a, EfMetRS IC50=1.3 mM). However,
when Region 2 is l-azetidine no substituent larger than
6. Behringer, H.; Horst, T. Chem. Ber. 1957, 90, 1398.
7. Benoiton, N. L.; Hudecz, F.; Francis, M. F.; Int, J. Pept.
Prot. Res. 1995, 45, 266.