Conformational restriction of methionyl tRNA synthetase inhibitors leading to analogues with potent inhibition and excellent gram-positive antibacterial activity

Article abstract of DOI:10.1016/S0960-894X(03)00093-3

Conformationally restricted analogues of the central linker unit of bacterial methionyl tRNA synthetase (MRS) inhibitors have been prepared. The (1S,2R)-cyclopentylmethyl moiety was identified as the preferred cyclic linker, with significant diastereo- and enantioselectivity of activity. Combination of this linker with an optimal substituted aryl right-hand side has resulted in a compound with exceptionally good antibacterial activity against staphylococci and enterococci, including antibiotic resistant strains.

Full text of DOI:10.1016/S0960-894X(03)00093-3

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1265–1268
Conformational Restriction of Methionyl tRNA Synthetase
Inhibitors Leading to Analogues with Potent Inhibition and
Excellent Gram-Positive Antibacterial Activity
Richard L. Jarvest,a,* John M. Berge,a Pamela Brown,a Catherine S. V. Houge-Frydrych,a
Peter J. O’Hanlon,a David J. McNair,a Andrew J. Popea and Stephen Rittenhouseb
aGlaxoSmithKline, New Frontiers Science Park, Third Avenue, Harlow, Essex CM19 5AW, UK
bGlaxoSmithKline, 1250 South Collegeville Road, Collegeville, PA 19426, USA
Received 12 December 2002; revised 23 January 2003; accepted 23 January 2003
Abstract—Conformationally restricted analogues of the central linker unit of bacterial methionyl tRNA synthetase (MRS) inhibi-
tors have been prepared. The (1S,2R)-cyclopentylmethyl moiety was identified as the preferred cyclic linker, with significant dia-
stereo- and enantioselectivity of activity. Combination of this linker with an optimal substituted aryl right-hand side has resulted in
a compound with exceptionally good antibacterial activity against staphylococci and enterococci, including antibiotic resistant
strains.
# 2003 Elsevier Science Ltd. All rights reserved.
New classes of antibiotics acting at novel molecular
targets are sought after to counteract the growing threat
of bacterial resistance. In the hospital context, resistant
pathogens of particular concern are methicillin resistant
Staphylococcus aureus (MRSA) and vancomycin resis-
tant enterococci (VRE). At some locations MRSA is
now responsible for 20–40% of all S. aureus infections,1
whilst many strains of VRE have developed resistance
to all established classes of antibiotics.2 An under-
exploited class of antibacterial targets are the aminoacyl
tRNA synthetases, essential enzymes in protein bio-
synthesis. The topical antibiotic mupirocin (marketed as
Bactroban1) acts by inhibition of bacterial isoleucyl
tRNA synthetase and selective inhibitors of other ami-
noacyl tRNA synthetases are of interest for their
potential as novel-acting antibacterials.
left hand side aryl substitution pattern resulted in 2,3,5-
trisubstituted analogues such as 2 which had potent
antibacterial activity.4 Here we describe the identifica-
tion of a constrained central linker with diastereo- and
enantioselective activity which can confer exceptionally
potent Gram-positive antibacterial activity.
The first constrained analogues to be prepared were the
three cyclohexanes 6–8. Reaction of the methoxyquino-
line 3 with an excess of the appropriate diaminocyclo-
hexane followed by acid hydrolysis of the
methoxyquinoline to the quinolone and subsequent
reductive alkylation afforded the desired cyclohexane
analogues (Scheme 1).
We have reported inhibitors of bacterial methionyl
tRNA synthetase (MRS) as a potential new class of
antibiotic agents with good Gram-positive antibacterial
activity.3,4 The quinolone 1 was an early chemistry lead
for MRS inhibition with improved inhibition, target-
related Gram-positive antibacterial activity, and
increased chemical manipulability.3 Optimisation of the
The cyclohexanes were tested as inhibitors in an MRS
assay3 and tested for antibacterial activity against S.
aureus and Enterococcus faecalis in a standard assay for
minimum inhibitory concentration (MIC), Table 1.
The cis-1,3-diamino 6 and trans-1,4-diamino 8 cyclo-
hexane analogues, both corresponding to extended lin-
ker conformations, were poor MRS inhibitors (Table 1).
*Corresponding author. Tel.: +44-1279-627629; fax: +44-1279-
622260; e-mail: richard_l_jarvest@gsk.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00093-3

1266
R. L. Jarvest et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1265–1268
trans cyclopentylmethyl linkers were predicted to have
the correct geometry and scored highly on the list.
The separate stereoisomers of the cyclopentylmethyl
linker were prepared as shown in Scheme 2. All four
diastereoisomers of the aminocyclopentane ester 9–12
were prepared by the route of Davies et al.7 Each isomer
was separately taken through the reaction sequence as
illustrated for 12, starting with lithal reduction to the
alcohol 13. Mitsunobu coupling of alcohols to amino-
heterocycle carbamates8,9 was then exploited, using the
preferred conditions of tributylphosphine and 1,1-azo-
Scheme 1. Reagents and conditions: (i) 1,3-diaminocyclohexane or
t-1,3-diaminocyclohexane/Et2iPrN or t-1,4-diaminocyclohexane 70 ꢀC;
(ii) c.HCl reflux, 20 h; (iii) 3,4-diCl-benzaldehyde/NaCNBH3/NaOAc/
AcOH/ MeOH, 20 h.
dicarbonylpiperidine
ethoxycarbonyl
(ADDP).9
of
The
trichloro-
derivative
2-amino-4-ethoxy
Table 1. MRS inhibition and antibacterial activity of cyclohexane
linker analogues
quinoline (TAEQ)10 was used as the coupling partner to
afford the intermediate 14. Removal of the Troc and
benzylic groups by catalytic hydrogen transfer followed
by the standard hydrolysis and reductive alkylation
steps afforded the four stereoisomers 17, 18, 19, 20.
A tetrahydrofuran analogue of the preferred cyclopentyl
stereochemistry was prepared from the deoxyribose
derivative 2111 (Scheme 3). The left hand side amine was
introduced as azide which was reduced down with con-
comitant cleavage of the Troc group. The tributylphos-
pine/ADDP method was again used for coupling the
linker to the quinoline moiety.
No.
Stereo
X
IC50 (nM)
MIC (mg/mL)
S. aureus E. faecalis
S. aureus
MRS
Oxford
1
6
7
8
rac
rac
—
150
43
64
32
>64
>64
>64
>64
An unsaturated analogue 34 was prepared from the
alcohol 2812 (Scheme 4). Removal of the protecting
2300
The trans-1,3-diamino 7 was a moderately potent inhib-
itor, with an IC50 value of 43 nM, just 3-fold down on
the propyl linker. This compound corresponds to a
more-folded linker conformation.
The molecular modelling program Catalyst5 was used to
analyse conformational data from a number of linker
analogues, including the cyclohexyl compounds 6–8, to
generate a hypothesis of the biologically active linker
conformation.6 This hypothesis was used to prioritise
further conformationally restricted targets. The 1,2-
Scheme 3. Reagents and conditions: (i) PhCO2H/DEAD/PPh3/THF; (ii)
NaOH/MeOH; (iii) Ph2PON3/DEAD/PPh3/THF; (iv) TFA/CH2Cl2; (v)
ADDP/Bu3P/PhH; (vi) NH4HCO2/Pd-C/MeOH/reflux; (vii) 3,4-diCl-
benzaldehyde/NaCNBH3/AcOH/MeOH; (viii) c.HCl/reflux.
Scheme 4. Reagents and conditions: (i) ADDP/Bu3P/benzene; (ii)
Bu4NF/THF; (iii) 2M NaOH/dioxane/45 ꢀC; (iv) (PhO)2P(O)N3/Ph3P/
DEAD/THF; (v) NaBH4/EtOH; (vi) c.HCl/reflux; (vii) 3,4-diCl-ben-
zaldehyde/NaCNBH3/AcOH/MeOH.
Scheme 2. Reagents and conditions: (i) LiAlH4/Et2O; (ii) TEAQ/
ADDP/Bu3P/PhH; (iii) NH4HCO2/Pd-C/MeOH/reflux; (iv) c.HCl/
reflux; (v) 3,4-diCl-benzaldehyde/NaCNBH3/NaOAc/AcOH/MeOH.

R. L. Jarvest et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1265–1268
1267
Scheme 5. Reagents and conditions: (i) 2-Cl-4-quinolone/Et2iPrN/
dioxane/75 ꢀC; (ii) 3,4-diCl-benzaldehyde/NaCNBH3/NaOAc/AcOH/
MeOH.
Scheme 6. Reagents and conditions: (i) TEAQ/DEAD/PPh3/THF; (ii)
NH4HCO2/Pd-C/MeOH/reflux; (iii) 3M H2SO4; (iv) 3,4-diCl-benzyl-
amine/NaCNBH3/NaOAc/AcOH/MeOH; (v) c.HCl reflux.
groups from 29 resulted in unexpected formation of the
cyclic carbamate 30, which nonetheless could be ring
opened and taken through to the target 34.
mediate (in this instance DEAD/PPh3 was the preferred
reagent combination) gave 39 which was deprotected to
the afford the ketone 40. The reductive amination was
carried out with inverted functionality to the usual
method and afforded a cis/trans mixture 41, which was
separated into the two diastereomers by HPLC. The cis
and trans compounds were separately hydrolysed by the
standard procedure to give the racemates 42 and 43
(Scheme 6).14
Due to the constraint of the acid lability of a benzylic
bond to the quinoline 2-amine, the aromatic analogue
37 was prepared via amine displacement on 2-chloro-4-
quinolone followed by reductive alkylation (Scheme 5).
Cyclobutane analogues were prepared from the acetal
38.13 Mitsunobu coupling to the Troc-quinoline inter-
The cyclic-methyl compounds were tested in the same
way as the cyclohexanes (Table 2). Of the four cyclo-
pentanes, the (1S,2R) isomer 20 was preferred, whilst its
antipode 19 was worst. There is a moderately high
degree of stereoselectivity with a range of about two
orders of magnitude of IC50 and MIC values across the
isomers. The antibacterial activity of 20 is clearly best
and is significantly improved over that of the straight
propyl linker. This suggests that, as predicted, the trans-
cyclopentylmethyl linker is locking the preferred con-
formation of the linker moiety.
Table 2. MRS inhibition and antibacterial activity of cyclic-methyl
linker analogues
No.
Stereo
X
IC50 (nM)
MIC (mg/mL)
S. aureus
MRS
S. aureus
Oxford
E. faecalis
1
The tetrahydrofuran analogue of the preferred stereo-
isomer, 27, had significantly reduced MRS inhibition
and little antibacterial activity. Introduction of planar-
ity—either unsaturation 34 or aromaticity 37—resulted
in loss of antibacterial activity. The cyclobutane analo-
gues were tested in racemic form, 42 and 43. Taking into
account the fact that 42 is racemic, it appears to have
very similar enzyme inhibition and antibacterial activity
to the cyclopentane 20. The cyclobutane did not have
any benefits in other properties over the cyclopentane
and in view of the increased synthetic difficulties, it was
not pursued further.
17
1R,2R
18
16
8
18
19
20
27
34
37
42
43
1S,2S
1R,2S
1S,2R
1S,2R
rac
100
700
9
8
>64
1
1
32
0.125
4
78
64
Table 3. MRS inhibition and antibacterial activity of cyclo-
pentylmethyl left-hand side analogues
10
>64
>64
2
>64
>64
0.25
16
—
56
No.
R
IC50 (nM)
MIC (mg/mL)
rac
12
S. aureus
MRS
S. aureus
Oxford
E. faecalis
1
rac
110
>64
44
45
H
EtO
10
7.6
ꢁ0.06
ꢁ0.06
ꢁ0.06
ꢁ0.06

1268
R. L. Jarvest et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1265–1268
The (1S,2R)-cyclopentylamine intermediate 16 was cou-
pled by the standard procedure to 3,5-dibromo and
2-ethoxy-3,5-dibromo benzaldehydes.3 The products 44
and 45 were highly potent antibacterial agents (Table
3). The antibacterial activity of 44 was about 4-fold
better than that of the corresponding propyl linker
analogue.
References and Notes
1. Public Health Laboratory Service Commun. Dis. Rep. 2001,
11, 7.
2. Low, D. E.; Keller, N.; Barth, A.; Jones, R. N. Clin. Infect.
Dis. 2001, 32 (Suppl. 2), S133.
3. Jarvest, R. L.; Berge, J. M.; Berry, V.; Boyd, H. F.; Brown,
M. J.; Elder, J. S.; Forrest, A. K.; Fosberry, A. P.; Gentry,
D. R.; Hibbs, M. J.; Jaworski, D. D.; O’Hanlon, P. J.; Pope,
A. J.; Rittenhouse, S.; Sheppard, R. J.; Slater-Radosti, C.;
Worby, A. J. Med. Chem. 2002, 45, 1959.
4. Jarvest, R. L.; Berge, J. M.; Brown M. J.; Brown, P.; Elder,
J. S.; Forrest, A. K.; Houge-Frydrych, C. S. V.; O’Hanlon, P.
J.; McNair, D. J.; Rittenhouse, S.; Sheppard, R. J. Bioorg.
Med. Chem. Lett. 2003, 13, 665.
5. Catalyst1 3.1 Accelrys Inc. www.accelrys.com.
6. Conformations of 52 representative molecules including the
cyclic linkers (6, 7 and 8) were generated within SYBYL 6.4,15
and used in hypothesis generation in Catalyst 3.1.5 Test mole-
cules under consideration were then subject to conformer
generation within SYBYL and scored against the hypothesis.
7. Davies, S. G.; Ichihara, O.; Lenoir, I.; Walters, A. S. J.
Chem. Soc., Perkin Trans. 1 1994, 1411.
8. Arnould, J. C.; Landier, F.; Pasquet, M. J. Tetrahedron
Lett. 1992, 33, 7133.
9. Abarghaz, M.; Kerbal, A.; Buorguignon, J.-J. Tetrahedron
Lett. 1995, 36, 6463.
The linker analogues retained selectivity for the bacter-
ial enzyme, none of the compounds giving significant
inhibition of mammalian (rat liver) MRS up to the
highest concentration tested (either 1 or 10 mM).
Compound 45 was tested against panels of clinical isolates
of S. aureus, Staphylococcus epidermidis, E. faecalis and
Enterococcus faecium that include a range of resistant
organisms.3 The MIC90 values (the concentration
required to inhibit 90% of the organisms) of the com-
pound were determined. Excellent activity was seen
against all the organisms, with MIC90 values at or
below 0.125 mg/mL: MIC90’s S. aureus, 0.06 mg/mL; S.
epidermidis, 0.125 mg/mL; E. faecalis, 0.03 mg/mL; and
E. faecium ꢁ0.016 mg/mL. The activity against the
panel of staphylococci was the best seen for this series,
again confirming the beneficial effect of the con-
strained linker.
10. TEAQ was prepared by reaction of 2-amino-4-ethoxy-
quinoline with Troc chloride.
11. Takeshita, M.; Chang, C. N.; Johnson, F.; Will, S.;
Grollman, A. P. J. Biol. Chem. 1987, 262, 10171.
12. Tao, T.; Parry, R. J. Organic Lett. 2001, 3, 3045.
13. Amice, P.; Conia, J. M. Tetrahedron Lett. 1974, 6, 479.
14. The stereochemistry of the isomers 42 and 43 could not be
unambiguously assigned by NMR. The assignment shown is
based on their biological activity.
In conclusion, we have identified the (1S,2R)-cyclo-
pentylmethyl moiety as the preferred conformationally
restricted linker for the MRS inhibitor series. Incor-
poration of this linker has resulted in a compound with
exceptionally good antibacterial activity against sta-
phylococci and enterococci, including antibiotic resis-
tant strains.
15. SYBYL1 6.4. Tripos Inc, St Louis, MO 63144, USA.