Noviose mimics of the coumarin inhibitors of gyrase B

Article abstract of DOI:10.1016/j.tetlet.2003.10.076

The design, synthesis, and biological activity in vitro of modified coumarin inhibitors of gyrase B are presented. Noviose, the sugar portion of coumarin antibiotics, was replaced by simplified mimics, a 5′,5′- dimethylcyclohexane or piperidine.

Full text of DOI:10.1016/j.tetlet.2003.10.076

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 9259–9262
Noviose mimics of the coumarin inhibitors of gyrase B
Branislav Musicki,a,* Anne-Marie Periers,a Laurent Piombo,a Patrick Laurin,a Michel Klich,a
Claudine Dupuis-Hamelin,b Patrice Lassaigneb and Alain Bonnefoyb
aMedicinal Chemistry, Infectious Disease, Aventis Pharma, 102 route de Noisy, 93235 Romainville Cedex, France
bMedicinal Chemistry, Aventis Pharma, 102 route de Noisy, 93235 Romainville Cedex, France
Accepted 14 October 2003
Abstract—The design, synthesis, and biological activity in vitro of modified coumarin inhibitors of gyrase B are presented.
Noviose, the sugar portion of coumarin antibiotics, was replaced by simplified mimics, a 5%,5%-dimethylcyclohexane or piperidine.
© 2003 Elsevier Ltd. All rights reserved.
In previous reports from these laboratories1 we have
described the synthesis and structure–activity relation-
ships of different series of novobiocin-(1)- or cloro-
biocin-(2)-like coumarin inhibitors of DNA gyrase, a
bacterial enzyme that participates in the regulation of
DNA topology in cells.2 So far, we have been perform-
ing structural modifications on the coumarin part of the
analogues. In this report we describe attempts to
replace the noviosyl part of coumarin drugs with sim-
plified mimics. Our objective was to explore alternative,
simpler synthetic strategies, compared to the relatively
complex synthesis of noviose,1f trying at the same time
to improve pharmaco-chemical properties of the cou-
marin drugs, especially their bioavailability and
aqueous solubility.
the absence of the stabilising anomeric effect associated
with the more stable axial (a) orientation of the cou-
marin moiety in coumarin drugs.4 The calculation of
the relative energies of two possible conformations, the
‘active’ one 3 and the ‘inactive’ one 4, indicated an
energy difference of 3.1 kJ mol−1.5 Considering the
absence of solvation effects in the calculations and the
inherent limitations of the computational model, this
small energy difference in favour of the ‘inactive’ con-
formation did not discourage us from pursuing the
synthesis of the cyclohexyl target.
Our second choice for the noviosyl mimic was the
piperidine moiety 5, with a nitrogen atom replacing the
5%,5%-dimethyl group of noviose (Fig. 1). The presence
of the nitrogen atom could serve as a convenient handle
for diverse structural modifications of this part of the
molecule. As in the first target, the piperidine ring in 5
could not provide the necessary anomeric stabilisation
and the results of calculations of relative energies gave
the preference to the ‘inactive’ conformation 6 over the
‘active’ one 5 by 2.7 kJ mol−1.
In the search for simplified mimics of the noviosyl part
of coumarin antibiotics, our first choice was a cyclo-
hexyl mimic 3 bearing the 5%,5%-dimethyl group and
lacking the 4%-MeO group of noviose (Fig. 1).3 Obvi-
ously, replacement of the tetrahydropyran ring of the
sugar with a 5%,5%-dialkylcyclohexane is accompanied by
Keywords: antibiotics; coumarin; noviose; structure–activity.
* Corresponding author. Tel.: +33-1-4991-4007; fax: +33-1-4991-5087; e-mail: branislav.musicki@aventis.com
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.10.076

9260
B. Musicki et al. / Tetrahedron Letters 44 (2003) 9259–9262
( )-5,5-Dimethylcyclohex-2-en-1-ol (8)6 seemed to be an
ideal starting material for the preparation of struc-
turally simplified cyclohexyl inhibitors (Scheme 1).
Previously1d we described a short synthesis of a useful
coumarin building block: 4-benzyloxy-3-carbethoxy-
coumarin 7. As in the case of noviose, coupling the
cyclohexenol 8 with the coumarin 7 was effected in the
presence of PPh3 and DEAD (Mitsunobu conditions) in
dichloromethane to afford 9 in 51% yield after chro-
matographic separation. cis-Hydroxylation of the
olefinic bond in 9 was realised in the presence of
catalytic OsO4/4-methylmorpholine N-oxide to afford
exclusively the diol 10. Regioselective silylation of diol
10 at the 3%-position was carried out with Et3SiCl in the
presence of diisopropylethylamine and imidazole. Tet-
rahydropyranylation of the 2%-hydroxy group under
standard conditions provided the THP derivative 11
that was desilylated smoothly with Bu4NF to afford the
alcohol 12.
with TsOH in MeOH to give the corresponding cyclo-
hexyl target 15.
The synthesis of the racemic piperidine mimic was
performed similarly to the cyclohexyl target and is
outlined in Scheme 2. The starting piperid-4-en-3-ol 17
protected as its Cbz derivative was prepared accord-
ing to Ref. 7. The introduction of the 5-methylpyrrole-
carboxylate at the free 5%-hydroxyl group in 21 was
achieved with EDAC (1-ethyl-3-(3-dimethylaminoprop-
yl)carbodiimide-HCl) and DMAP in dichloromethane
to afford intermediate ester 22. Catalytic hydrogenation
of 22 in THF at room temperature provided an easily
separable mixture of 23a and 23b in 70% and 30%
yields, respectively. These were smoothly deprotected in
MeOH with TsOH to furnish the corresponding free
4%-alcohols, 24 and 25, respectively, 25 being isolated as
a p-toluenesulfonic acid salt. Further alkylation of the
secondary amino group in 25 under the usual condi-
tions provided the N-allyl and N-iPr analogues 26 and
27.
The free 4-OH group of the coumarin was recovered in
quantitative yield by hydrogenolysis in the presence of
a Pd catalyst. Introduction of the N-propargyloxycar-
bamate (a 5-methylpyrrole-2-carboxylate bioisostere) in
13 at C-3% was carried out using p-nitrophenyl chloro-
formate and O-propargylhydroxylamine as described in
Ref. 1d. The ester group (CO2Et) of 14 was converted
to an amide (CONH2) with gaseous NH3 in THF. In
the last step, the THP group was easily deprotected
Biological results: Table 1 shows the inhibition in the
supercoiling activity of E. coli/S. aureus DNA gyrase
by novobiocin, clorobiocin and coumarin inhibitors
with noviose mimics. In Table 1 also included are the
corresponding results of the reference compounds
incorporating sugar noviose: 161d (Scheme 1) and 28
(Scheme 2).1c
Figure 1. Calculated relative conformational energies of cyclohexyl mimics 3 and 4 and N-methylpiperidine mimics 5 and 6.
Scheme 1. Reagents and conditions: (a) PPh3, EtO2CNꢀNCO2Et, CH2Cl2, rt, 51%; (b) OsO4, 4-methylmorpholine N-oxide,
tBuOH, THF, H2O, rt, 91%; (c) Et3SiCl, DIPEA, Im, CH2Cl2, 0°C, 64%; (d) DHP, TsOH cat, CH2Cl2, rt, 66%; (e) Bu4NF, THF,
rt, 75%; (f) H2, Pd–C/10%, THF, rt, quant; (g) i. pNO2C6H4OCOCl, DMAP, CH2Cl2, 0°C; ii. HCꢁCCH2ONH2HCl, NaH,
DMAP, DMF, 0°C; (h) NH3, THF, rt; (i) TsOH cat, MeOH, CH2Cl2, rt, 54% overall for (g)–(i).

B. Musicki et al. / Tetrahedron Letters 44 (2003) 9259–9262
9261
Scheme 2. Reagents and conditions: (a) PPh3, EtO2CNꢀNCO2Et, CH2Cl2, rt, 51%; (b) OsO4, 4-methyl morpholine-N-oxide,
tBuOH, THF, H2O, rt, 92%; (c) Et3SiCl, DIPEA, Im, CH2Cl2, rt, 77%; (d) DHP, TsOH cat, CH2Cl2, rt, quant.; (e) Bu4NF, THF,
rt, 96%; (f) 5-methylpyrrole-2-carboxylic acid, EDAC, DMAP, CH2Cl2, 63%; (g) H2, Pd–C/10%, THF, rt.; (h) TsOH cat, MeOH,
CH2Cl2, rt, 39%.; (i) TsOH cat, MeOH, rt, 39%; (j) Allyl-Br, 27% or iPrI, 20%, K2CO3, DMF, rt.
Table 1. In vitro activity of coumarin inhibitors against E. coli/S. aureus DNA gyrase supercoiling activity (IC50),a,b and
selected in vitro antibacterial activity (MIC)c
Compound
Novob
15
16
24
25
26
27
28
IC50nova,b/IC50comp
MIC S. aureus 011HT3
1a,b
50.04
0.34b
2.5
1a
50.04
0.06b
>40
0.25b
>40
0.16b
>40
0.06b
>40
3.1b
0.3
a IC50 was determined for gyrase B of E. coli against novobiocin (0.25 mg/mL) as reference. For the details see Ref. 1c.
b IC50 was determined for gyrase B of S. aureus against novobiocin (0.5 mg/mL) as reference. For the details see Ref. 1d.
c MIC, minimal inhibitory concentrations (mg/mL) were measured by using a twofold broth microdilution after 24 h incubation.
these simplified targets showed only weak inhibitory
activity and moderate in vivo inhibition (MIC values).
We believe that the absence of the 4%-MeO substituent
of the noviose mimics is most probably responsible for
the diminution, by one order of magnitude, of the
inhibition of the supercoiling activity of DNA gyrase.
Proton NMR spectra of the final targets 15 and 24–27
confirmed the calculation results that predicted the
more stable conformations 4 and 6. However, the small
difference in inhibition in supercoiling of analogue 15
compared to 16, could be attributed to the absence of
the 4%-MeO group in these molecules rather than to
prevalence of the more favourable ‘inactive’ conforma-
tions for 15. As revealed by X-ray studies of complexes
of coumarin drugs and the 24 kDa N-terminal fragment
of gyrase B,8,9 the 4%-MeO group of noviose is involved
in hydrophobic interactions with the surrounding
amino acid residues of the gyrase B protein as well as in
hydrogen bonding to the side chain of Asn-46.10 Cyclo-
hexyl mimics with a 5%,5%-dimethyl group proved to be
better inhibitors in supercoiling activity of DNA gyrase
than the corresponding piperidine counterpart, proba-
bly due to the more effective interaction of the gem-
dimethyl group with the surrounding hydrophobic
pocket. Introduction of the N-alkyl groups in 25 did
not improve the inhibition of the series (26, 27),
whereas N-carbamate (24) substitution more or less
abolished the inhibition.
Acknowledgements
We are grateful to the Analytical department (Aventis,
Romainville) for performing the spectral analysis.
References
1. (a) Klich, M.; Laurin, P.; Musicki, B.; Schio, L. WO
9747634, 1998; Chem. Abstr. 1998, 128, 75634; (b)
Chartreaux, F.; Klich, M.; Schio, L. EP 894805, 1999;
Chem. Abstr. 1999, 130, 125344; (c) Laurin, P.; Ferroud,
D.; Klich, M.; Dupuis-Haelin, C.; Mauvais, P.; Las-
saigne, P.; Bonnefoy, A.; Musicki, B. Bioorg. Med. Chem.
Lett. 1999, 9, 2079–2084; (d) Haesslein, J.-L.; Klich, M.;
Laurin, P.; Musicki, B.; Periers, A.-M. WO 9935155,
1999; Chem. Abstr. 1999, 131, 454249; (e) Laurin, P.;
Ferroud, D.; Schio, L.; Klich, M.; Dupuis-Haelin, C.;
In conclusion, we successfully realised the synthesis of
cyclohexyl and piperidine mimics of noviose. However,

9262
B. Musicki et al. / Tetrahedron Letters 44 (2003) 9259–9262
Mauvais, P.; Lassaigne, P.; Bonnefoy, A.; Musicki, B.
Chemistry; Lipkowitz, K. B.; Boyd, D. B., Eds.;
Approaches to Empirical Force Fields; VCH: New York,
1991; Vol 2, Chapter 4.
Bioorg. Med. Chem. Lett. 1999, 19, 2875–2880; (f) Fer-
roud, D.; Collard, J.; Klich, M.; Dupuis-Haelin, C.;
Mauvais, P.; Lassaigne, P.; Bonnefoy, A.; Musicki, B.
Bioorg. Med. Chem. Lett. 1999, 19, 2881–2886.
6. Hartmann, D. J. Am. Chem. Soc. 1953, 75, 3723–3726.
7. Imanishi, T.; Miyashita, K.-I.; Nakai, A.; Inoue, M.;
Hanaoka, M. Chem. Pharm. Bull. 1983, 31, 1191–1198.
8. (a) Wigley, D. B.; Davies, G. J.; Dodson, E. J.; Maxwell,
A.; Dodson, G. Nature 1991, 351, 624–629; (b) Lewis, R.
J.; Singh, O. M. P.; Smith, C. V.; Maxwell, A.; Skarzyn-
sky, T.; Wonacott, A. J.; Wigley, D. B. J. Mol. Biol.
1994, 241, 128–130; (c) Lewis, R. J.; Singh, O. M. P.;
Smith, C. V.; Skarzynski, T.; Maxwell, A.; Wonacott, A.
J.; Wigley, D. B. EMBO J. 1996, 15, 1412–1420; (d) Tsai,
F. T. F.; Singh, O. M. P.; Skarzynski, T.; Wonacott, J.
A.; Weston, S.; Tucker, A.; Pauptit, R. A.; Breeze, A.;
Poyser, J. P.; O’Brien, R.; Ladbury, J. E.; Wigley, D. B.
Proteins: Struct. Funct. Genet. 1997, 28, 41–52.
2. For a review of DNA gyrase structure and function, see:
(a) Reece, R. J.; Maxwell, A. Crit. Rev. Biochem. Mol.
Biol. 1991, 26, 335–375; (b) Levine, C.; Hiasa, H.; Mari-
ans, K. J. Biochim. Biophys. Acta 1998, 1400, 29–43; (c)
Berger, J. M. Biochim. Biophys. Acta 1998, 1400, 3–18;
(d) Maxwell, A. Biochem. Soc. Trans. 1999, 27, 48–53.
3. Bell, W.; Block, M. H.; Cook, C.; Grant, A. J.; Timms,
D. J. Chem. Soc., Perkin Trans. 1 1997, 2789–2801. In
this publication, the authors used a somewhat similar
approach based on a series of spiro 5%,5%-dimethylcyclo-
hexane acetals bearing different biphenyl systems as a
coumarin replacement.
4. The thermodynamic equilibrium between the a-methyl
coumeroside and b-methyl coumeroside that favours the
a-anomer is largely determined by the anomeric effect.
See: Wick, A. E.; Blount, J. F.; Leimgruber, W. Tetra-
hedron 1976, 32, 2057–2065.
9. Lafitte, D.; Lamour, V.; Tsvetkov, P. O.; Makarov, A.
A.; Klich, M.; Deprez, P.; Moras, D.; Briand, C.; Gilli,
R. Biochemistry 2002, 41, 7217–7223.
10. Recently, we accomplished the synthesis of
a 4%-
desmethoxy analogue of RU79115 and the results of its
biological activity confirmed the importance of this group
in conferring a good inhibitory potency to the analogues.
These findings will be published in due course.
5. The relative energies of the conformers 3–6 were mini-
mized with the Insight/Discoverer software applying the
CVFF forcefield and conjugate gradient method. See:
Dinur, U.; Hagler, A. T. In Reviews of Computational