Potent synthetic inhibitors of tyrosyl tRNA synthetase derived from C-pyranosyl analogues of SB-219383

Article abstract of DOI:10.1016/S0960-894X(01)00040-3

Novel pyranosyl analogues of SB-219383 have been synthesised to elucidate the structure-activity relationships around the pyran ring. Analogues with highly potent stereoselective and bacterioselective inhibition of bacterial tyrosyl tRNA synthetase have been identified. A major reduction in the overall polarity of the molecule can be tolerated without loss of the nanomolar level of inhibition.

Full text of DOI:10.1016/S0960-894X(01)00040-3

Bioorganic & Medicinal ChemistryLetters 11 (2001) 715±718
Potent Synthetic Inhibitors of Tyrosyl tRNA Synthetase Derived
from C-Pyranosyl Analogues of SB-219383
Richard L. Jarvest, John M. Berge, Pamela Brown,* Dieter W. Hamprecht,
David J. McNair, LucyMensah, Peter J. O'Hanlon and Andrew J. Pope
GlaxoSmithKline, New Frontiers Science Park, Third Avenue, Harlow, Essex, CM19 5AW, UK
Received 29 November 2000; accepted 12 January2001
AbstractÐNovel pyranosyl analogues of SB-219383 have been synthesised to elucidate the structure±activity relationships around
the pyran ring. Analogues with highly potent stereoselective and bacterioselective inhibition of bacterial tyrosyl tRNA synthetase
have been identi®ed. A major reduction in the overall polarityof the molecule can be tolerated without loss of the nanomolar level
of inhibition. # 2001 Elsevier Science Ltd. All rights reserved.
The increasing emergence of pathogenic bacteria resis-
tant to conventional antibiotics has made the develop-
ment of novel agents against bacterial targets ever more
urgent. In our search for novel inhibitors of bacterial
tRNA synthetases as potential antibacterial agents, the
natural product, SB-219383 1,1,2 was discovered as a
potent, selective inhibitor of bacterial tyrosyl tRNA
synthetase (YRS). The synthetic analogue 2 has been
found to be equipotent to 1, and has been used to infer
essential for activityand to give more scope for the
variation of polarity. Here we report the synthesis and
SARs of new pyranosyl ring analogues which resulted in
the identi®cation of a highlypotent sub-nanomolar
synthetic inhibitor of bacterial YRS.
In order to probe the requirements around the pyr-
anosyl ring, selected sugars were taken through the
Lewis acid catalysed C-glycosylation synthesis.4 These
included the l-fucose, d-galactose, d-glucose and l-lyx-
ose templates. From d-fucose, the ®nal diastereomers
4a and 4b were separated bypreparative HPLC and
isomers a and b 0inferred to have the (R) and (S)-stereo-
chemistryat C1 on the basis of the chemical shift of
H1.5 To further reduce polarity, the fucose derivative 4b
3
the absolute stereochemistryof the natural product. In
the previous paper,4 we showed that the pyranose ana-
logue 3 is a selective inhibitor of bacterial YRS,
although with an IC50 of 100 nM it was less potent than
2. We sought to simplifythe sca€old further, in order to
investigate which of the features of this molecule were
*Corresponding author. Tel.: +44-1279-627633; fax: +44-1279-627628; e-mail: pam_brown-1@sbphrd.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00040-3

716
R. L. Jarvest et al. / Bioorg. Med. Chem. Lett. 11 (2001) 715±718
was also converted into the butyl ester 5 bytreatment
with butanol in the presence of HCl. Starting from the
more polar d-galactose or d-glucose sugars, ®nal epi-
mer separation was not possible and the diastereomer
ratios of 6 and 7, respectively, were determined by
NMR.5 From l-lyxose, it was possible to reduce the
diastereomer corresponding to b with sodium borohy-
dride, which was then elaborated to 8 without epimer-
isation, as judged byNMR. In some cases, the bulky
pivaloyl protecting groups resulted in reduced yields in
the C-glycosylation step and attempts were made to
improve the scope of the reaction, for example byusing
alternative protecting groups. However, anyprotecting
group other than pivaloyl resulted in the formation
of a bicyclic acetal such as 10 during the attempted
C-glycosylation.
Due to the neighbouring group participation required
for the C-glycosylation, it is not possible to synthesise
analogues epimeric at the 2-position bythis method.
However, in order to probe further the requirements for
the polar hydroxy groups in this region, the 2-deoxy
Scheme 1. (i) NaOH, MeOH, H2O, 80 ꢀC; (ii) HCl; (iii) n-BuOH, HCl, 100 ꢀC, 16 h; (iv) F3CC(O)N(Me)TMS, N(iPr)2Et, pyridine, 0 ꢀC; (v) Z-
Tyr(Bn)OH, HOBt, EDAC, DMF, N(iPr)2Et, (vi) Me2C(OMe)2 pyridinium toluene-4-sulfonate; (vii) N,N0-thiocarbonyl diimidazole, benzene,
80 ꢀC, 38 h; (viii) Bu3SnH, benzene, 80 ꢀC, 18 h; (ix) DBU, 1,2-dichloroethane, 80 ꢀC, 18 h; (x) aq HCl, dioxan; (xi) H2, Pd/C, TFA.

R. L. Jarvest et al. / Bioorg. Med. Chem. Lett. 11 (2001) 715±718
717
compound 9 was prepared from the fucose analogue 11
using Barton methodology6 as shown in Scheme 1. Thus
the diasteromeric mixture 11a,b was completelydepro-
tected with sodium hydroxide in aqueous dioxan, and
then converted into the butyl ester 12. This was persily-
the butyl ester 5. The role of the hydroxyl groups
around the ring in recognition byYRS was also eluci-
dated. The axial stereochemistryat C-4 is crucial to potency
(compare 7 with 6) as is the equatorial stereochemistryat
the C-3 position (compare 8 with 3). However, removal
of the 2-hydroxyl group is well tolerated as shown by
the potent inhibition observed for compound 9.
3
lated prior to coupling to suitablyprotected tyrosine
followed byformation of the 3,4-acetonide. Conversion
of the unprotected 2-hydroxy into the thiocarbonylimi-
dazolide followed byradical de-oxygenation gave the 2-
deoxyanalogue 13. This was re-epimerised at the gly-
cine centre with DBU, followed bydeprotection to
a€ord the 2-deoxycompound 9 as a 2:1 mixture of
(R:S) isomers as determined byNMR.
By comparison with the potent hydroxylamine, 2, it has
been inferred that the conformation at the glycine centre
of 4b is (S). This has been further supported byX-ray
crystallography in the active site of YRS.8,9 The impor-
tance of the hydroxyl groups has also been noted in the
active site. The axial hydroxyl at C-4, which is crucial for
potency, makes two signi®cant interactions, one with the
Gly193NH in the protein, and one with an ordered water
molecule, while the two hydroxyl groups of the equator-
ial diol appear to interact with the Asp 195 side chain.
The compounds were evaluated as inhibitors of Staphy-
lococcus aureus tyrosyl tRNA synthetase (YRS), in an
aminoacylation assay,7 and the results are given in
Table 1. At C-5, the hydroxymethyl group of 6 results in
a roughlysimilar level of potencyto the unsubstituted
compound 3. However, introducing a methyl group at
C-5, 4b, gives rise to a 20-fold improvement, taking the
inhibition to 4 nM, the level of SB-219383 itself. The
potent inhibition of compound 4b is both stereoselective
(no inhibition for the epimer 4a) and bacterioselective (no
inhibition of mammalian YRS up to 3 mM). The potency
of 4b is further improved to 0.8 nM on the formation of
Despite the greatlyreduced polarityof 9 compared to
SB-219383,10 the compound did not show anyanti-
bacterial activityagainst a varietyof bacteria up to
32 mg mLÀ1. Compound 5, however, had weak anti-
bacterial activityagainst the pathogens Moraxella cat-
arrhalis and Streptococcus pyogenes, with MICs of 8 mg
mLÀ1
.
Table 1. Inhibition of S. aureus tyrosyl tRNA synthetase
5
Compound
Sugar ring
Stereochemistryat C1 0
IC50 (nM)
100
3
l-Arabinose3
(S)
6
d-Galactose
l-Fucose
3:1 (R):(S)
190
4a
4b
(R)
(S)
NIa
4
5
7
l-Fucose butyl ester
(S)
0.8
NI
d-Glucose
2:3 (R):(S)
8
9
l-Lyxose
(S)
71%@ 3 mM
2-Deoxy-fucose
2:1 (R):(S)
3.6
aNI=no inhibition at 3 mM.

718
R. L. Jarvest et al. / Bioorg. Med. Chem. Lett. 11 (2001) 715±718
In conclusion, we have identi®ed keyinteractions of the
pyranosyl analogues of SB-219383 and synthesised
compounds with highlypotent and selective inhibition
of bacterial tyrosyl tRNA synthetase. We have also
shown that a major reduction in the overall polarityof
the molecule can be tolerated without loss of the nano-
molar level of inhibition.
eomers. Byanalogyto the l(+)-arabinose derived series 3,4
the diastereomer with the higher-®eld chemical shift of H1,
isomer a, is inferred to have the (R) stereochemistryat C1 .
0
Examples of chemical shifts (ppm) and coupling constants (Hz) in
C-pyranosyl dipeptides (D2O)
Compound Isomer H-1
H-2
3
4
6
7
a
b
a
b
a
b
a
b
2.68, dd, J 9.9, 1.5 Hz
3.87, dd, J 2.0,10.0 Hz
2.67, dd, J 10.1, 1.9 Hz 3.83, t, J 9.6 Hz
3.00, t, J 9.5 Hz
3.62, t, J 9.7 Hz
Acknowledgements
under m, 3.74±3.89Hz
2.67, d, J 9.9 Hz
3.55, t, J 9.6 Hz
The authors would like to thank Xiayang Qiu for X-ray
crystallography of the S. aureus YRS-inhibitor complex,
D. O'Brien and C. Seaman for 2D-NMR experiments,
and S. Rittenhouse for antibacterial evaluation.
under m, 3.87±3.95Hz
2.87 d J 10.0 Hz
3.88, d, J 9.9 Hz
3.22, t, J 9.7 Hz
3.59. t, J 9.1 Hz
6. Barton, D. H. R.; Motherwell, R. S. H.; Motherwell, W. B.
J. Chem. Soc., Perkin Trans. 1 1981, 2363.
7. Brown, P.; Mensah, L. M.; O'Hanlon, P. J.; Osborne, N. F.;
Pope, A. J.; Richardson, C. M.; Walker, G. Bioorg. Med.
Chem. 1999, 7, 2473.
8. Qiu, X. et. al., manuscript in preparation. Brown, P.;
Berge, J. M.; Hamprecht, D. W.; Jarvest, R. L.; McNair, D.
J.; Mensah, L. M.; Pope, A. J.; Qiu, X.; Richardson, C. M.
Poster presentation at the 18th tRNA Workshop, Cambridge,
UK, April 18±21, 2000.
9. When a 5:1 mixture of 4a:4b was subjected to co-crystal-
lisation with S. aureus YRS, the crystals obtained contained
the (S)-isomer 4b only, supporting the hypothesis 0that the
active diastereomer has the (S)-stereochemistryat C1 .8
10. Calculated logP values for compounds 1 and 9 using
the Pomona clogP algorithm (1998) were À4.28 and 0.22,
respectively.
References and Notes
1. Stefanska, A. L.; Coates, N. J.; Mensah, L. M.; Pope, A. J.;
Ready, S. J.; Warr, S. R. J. Antibiot. 2000, 53, 345.
2. Houge-Frydrych, C. S. V.; Readshaw, S. A.; Bell, D. J. J.
Antibiot. 2000, 53, 351.
3. Berge, J. M.; Copley, R. C. B.; Eggleston, D. S.; Ham-
precht, D. W.; Jarvest, R. L.; Mensah, L. M.; O'Hanlon, P. J.;
Pope, A. J. Bioorg. Med. Chem. Lett. 2000, 10, 1811.
4. Brown, P.; Haltiwanger, R. C.; Jarvest, R. L.; Mensah, L.;
O'Hanlon, P. J.; Pope, A. J. Bioorg. Med. Chem. Lett. 2001,
11, 711.
5. It was found that in anypair of diastereomers at the glycine
centre, the chemical shift of the H1 proton di€ers by ꢁ1 ppm,
while the H1±H10 coupling constant is small in both diaster-