Design, synthesis and bioactivity evaluation of tribactam beta lactamase inhibitors.

Article abstract of DOI:10.1016/S0960-894X(02)00061-6

Known carbapenem compounds with inhibitory effect towards beta-lactamase enzymes are formed from bicyclical beta lactam structural scaffolds. On the basis of results from theoretical computational methods and molecular modelling we have designed and devel

Full text of DOI:10.1016/S0960-894X(02)00061-6

Bioorganic & Medicinal Chemistry Letters 12 (2002) 971–975
Design, Synthesis and Bioactivity Evaluation of
Tribactam ꢀ Lactamase Inhibitors
Anton Copar,a Tadeja Prevec,a Borut Anzic,a Tomaz Mesar,a Lovro Selic,a
Mateja Vilara and Tom Solmajera,b,
*
aLek d.d., Research and Development, Celovsˇka 135, 1526 Ljubljana, Slovenia
bNational Institute of Chemistry, POB 660, Hajdrihova 19, 1001 Ljubljana, Slovenia
Received 27 June 2001; revised 13 December 2001; accepted 22 January 2002
Abstract—Known carbapenem compounds with inhibitory effect towards b-lactamase enzymes are formed from bicyclical beta
lactam structural scaffolds. On the basis of results from theoretical computational methods and molecular modelling we have
designed and developed a synthetic route towards novel, biologically active tricyclic derivatives of carbapenems. # 2002 Elsevier
Science Ltd. All rights reserved.
Introduction
from various bacterial sources and stability against
renal dihydropeptidase (Fig. 1).
Due to their high efficiency and high specificity, b-lac-
tam antibiotics are currently the most widely used anti-
bacterial agents. However, this effectiveness has been
seriously compromised by the ability of bacteria to
produce b-lactamases, enzymes which hydrolyse the
b-lactam ring, thus rendering the drug biologically
inactive.
In order to address the increasing threat of bacterial
infections a novel class of tricyclic carbapenem anti-
biotics (trinems) were developed and an extremely good
biological profile was observed by the best compound in
the series, sanfetrinem.5 On the other hand, compounds
with 6-substituted methylene penem scaffold, for exam-
ple, BRL 42715 have shown excellent inhibitory activity
towards various classes of b-lactamases.6
Clinical experience has established that co-administra-
tion of a classical b-lactamase-sensitive compound with
a specific b-lactamase inactivator represents an efficient
strategy in fighting the b-lactamases. Although the few
inactivators in clinical use (e.g., clavulanate,1 sulbac-
tam2 and tazobactam3) readily inactivate most class A
enzymes, they are rather inefficient against the class C
b-lactamases. It appears that there is no clinically avail-
able efficient inactivator of the latter class of enzymes,
which represents an increasingly worrying problem.4
Our goal was to design and synthesise a tricyclic system
in which two structural elements of compounds with
inhibitory activity towards the b-lactamase and/or d,d
peptidase would be combined. Introduction of an
ethylidene bond at position 6 was hypothesised to sta-
bilise the inhibitor–enzyme complex based on the con-
jugation of the carbonyl group of the beta lactam ring
with ethylidene b-orbitals at position 6; in addition to
this, a hydrophobic cyclohexane ring in the C3–C4
position of the penem should block the access of the
The clinically used inhibitors of b-lactamases are bicyc-
lic compounds with adjacent five-member ring on the
beta lactam core containing a heteroatomic ring atom
constituent. Primarily, compounds substituted at posi-
tions 3, 4 and 6 of the carbapenem bicycle have been
studied and evaluated for activity towards b-lactamases
Figure 1. Chemical structures of available inhibitors. (a) clavulanate,
X¼O, R¼CH2OH; (b) sulbactam X¼SO2; (c) tazobactam X¼SO2,
R¼CH2-triazole.
*Corresponding author. Tel.: +386-1-4760-277; fax: +386-1-4760-
300; e-mail: tom.solmajer@ki.si
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00061-6

972
A. Copar et al. / Bioorg. Med. Chem. Lett. 12 (2002) 971–975
water molecules to the tetrahedral acyl–enzyme complex
which is formed in the first step of the enzyme–inhibitor
interaction process thus facilitating the stability of the
tetrahedral intermediate and preventing the deacylation
step.7,8
3,30-thiopropionyl chloride (4) by substitution of chlor-
ine with allyloxy group and subsequent cyclisation.
Thiopyran 6 was then coupled with azetidone 1 in the
presence of sodium hydride. Allyloxycarbonyl group
was removed under standard protocols12 to give 8 as a
mixture of epimers which were separated with column
chromatograpy (ether/petroleum ether 3:1) (Scheme
2).
In this paper, we report the synthesis and biological
activity of unsubstituted trinems 14a and 14b and their
6-thia derivatives 14c and 14d.9
Compounds 2 and 8 were coupled with allyl glyoxylate
and consequently transformed into fully protected tri-
nems 10 with oxalimide cyclisation methodology13 in
moderate yields. The silyl protecting group was
removed with tetrabutyl ammonium fluoride and acetic
acid.14 Thus, prepared hydroxyethyl trinems 11 were
transformed into their 10-ethylidene derivatives 12 via
Mitsunobu reaction (diethylazodicarboxylate/triphenyl-
phosphine), 6a with the exception of compound 11d
which was first converted into mesylate ester 13d and
then treated with 1,8- diazabicyclo[5.4.0]undec-7-ene
(DBU)6a to give the 10-ethylidene derivative 12d. Only
(E)-trinems were isolated upon purification with column
chromatography. The structure of these compounds
were confirmed with COSY and NOESY NMR experi-
ments. Finally, the target compounds 14 were obtained
The syntheses of key intermediate 2 are depicted in
Scheme 1. Commercially available azetidone 1 reacted
with 1-trimethylsilyloxycyclohexene10 or 4-cyclohex-1-
enylmorpholine11 in the presence of ZnI2 to give the
mixture of 20-epimers. The least successful approach in
terms of yield and assay of the more desirable 20-S-epi-
mer was coupling of azetidone 1 with allyl 2-oxocyclo-
hexanoate in the presence of sodium hydride, and the
consequent removal of allyloxycarbonyl group.12 Epi-
mers were separated by the column chromatography
(ethylacetate/hexane 5:1).
The key intermediate for the 6-thia trinems was synthe-
sised as shown in Scheme 2. Allyl 4-hydroxy-5,6-dihy-
dro-2H-thiopyran-3-carboxylate (6) was prepared from
Scheme 1. (i) 1-Trimethylsilyoxycyclohexene, ZnI2, CH2Cl2; (ii) 4-cyclohex-1-enyl-morpholine, ZnI2, CH2Cl2; (iii) allyl 2-oxocyclohexanoate; NaH,
THF; (iv) Pd(PPh3)4, PPh3, NEt3, HCOOH, CH2Cl2.
Scheme 2. (i) AllylOH, Et3N; (ii) Na, AllylOH, Et2O; (iii) NaH, THF; (iv) Pd(PPh3)4, PPh3, Et3N, HCOOH, CH2Cl2.

A. Copar et al. / Bioorg. Med. Chem. Lett. 12 (2002) 971–975
973
Scheme 3. (i) Allyl oxalyl chloride, Et3N, CH2Cl2; (ii) P(OEt)3, hidrochinon, toluene; (iii) TBAF, AcOH, THF; (iv) PPh3, DEAD, CH2Cl2; (v) PPh3,
Pd(PPh3)4, sodium ethylhexanoate, CH2Cl2/THF; (vi) Et3N, MeSO2Cl, CH2Cl2; (vii) DBU, CH2Cl2; (viii) PPh3, Pd(PPh3)4, sodium ethylhexanoate,
CH2Cl2/THF.
Table 1. Inhibitory activitya at b-lactamase I from B. cereus
by removing the allyl ester group using Pd(Ph3P)4 in the
presence of sodium 2-ethylhexanoate (Scheme 3).15
Inhibitor
% Inhibition of b-lactam hydrolysis at
inhibitor concentrationb
Compounds 1416 were sufficiently stable and were initi-
ally screened for their inhibitory activity towards the
enzyme b-lactamase I from Bacillus cereus. This was
determined by spectroscopic monitoring of the reduction
in hydrolytic inactivation of a chromogenic substrate
(nitrocefin) by the enzyme in the presence of a b-lactamase
inhibitor. The data obtained for the series of synthesised
compounds were compared with results of potassium
clavulanate and sulbactam (Table 1).
100.0 mmol/L
10.0 mmol/L
1.0 mmol/L
Clavulanate
Sulbactam
14a
14b
14c
14d
15a
15b
78
77
61
23
60
38
3
33
40
14
9
2
ꢀ1
19
6
5
aFor compounds in which the inhibition at 100.0 mmol/L was less than
50% the inhibitory activity at lower concentrations was not deter-
mined.
b100 mmol of the enzyme lactamase I from B. cereus were added to
100 mmol of solution with inhibitor concentrations in range 1–
100 mmol/L and the mixture was incubated in the microtitre plate for
20 min at 25 ꢁC. The aliquot of nitrocefin was added and additional
incubation for 60 min followed. The % inhibition of the enzyme was
monitored at 490 nm.
In order to further test our hypothesis that ethylidene
bond at position 10 provides the necessary stability of
the acyl–enzyme complex, we have synthesised and tes-
ted the compounds 15a and 15b in which the ethylidene
group in compounds 14a and 14b is substituted to
hydroxyethyl.7 Both epimers 15a and 15b which were
obtained by deallylation from compounds 11a and 11b,

974
A. Copar et al. / Bioorg. Med. Chem. Lett. 12 (2002) 971–975
Table 2. Inhibitory activity of ethylidene trinems against Class A and
Class C b-lactamase
References and Notes
Inhibitor
IC50 mmol/L
E. cloacae 908Rb
1. Bush, K. C.; Macalintal, B. A.; Rasmussen, V.; Lee, J.;
Yang, Y. Antimicrob. Agents Chemother. 1993, 37, 851.
2. Sanders, C. C. Clin. Infect. Dis. 1992, 14, 1089.
3. Massova, I.; Mobashery, S. Acc. Chem. Res. 1997, 30, 162.
4. (a) Frere, J. M. Mol. Microbiol. 1995, 16, 385. (b) Frere,
J. M.; Dubus, A.; Galleni, M.; Matagne, A.; Amicosante, G.
Biochem. Soc. Trans. 1999, 27, 58. (c) Lakaye, B.; Dubus, A.;
Lepage, S.; Groslambert, S.; Frere, J. M. Mol. Microbiol.
1999, 31, 89.
E. coli TEM1a
Clavulanate
25
39
120
325.0
100
100
3.0
1.8
14a
14b
14c
14d
4.2
aTEM 1 concentration was 0.3 mmol. Experimental details as in foot-
note b of Table 1.
b908R concentration was 0.8 mmol.
5. (a) Perboni, A.; Tamburini, B.; Rossi, T.; Donati, D.; Tar-
zia, G.; Gaviraghi, G. In Recent Advances in the Chemistry of
Anti-infective Agents; Bentley, P. H., Ponsford, R., Eds.; The
Royal Society of Chemistry: Graham House, Cambridge,
1992; p 21. (b) Di Modugno, E.; Erbetti, I.; Ferrari, L.;
Galassi, G.; Hammond, S. M.; Xerri, L. Antimicrob. Agents
Chemother. 1994, 38, 2362. (c) Di Fabio, R.; Feriani, A.;
Gaviraghi, G.; Rossi, T. Bioorg. Med. Chem. Lett. 1995, 5,
1235. (d) Padova, A.; Roberts, S. M.; Donati, D.; Marchioro,
C.; Perboni, A. Tetrahedron 1996, 52, 263. (e) Hanessian, S.;
Rozema, M. J.; Reddy, G. B.; Braganza, J. F. Bioorg. Med.
Chem. Lett. 1995, 5, 2535. (f) Babini, G. S.; Yuan, M.; Liver-
more, D. M. Antimicrob. Agents Chemother. 1998, 42, 1168.
(g) Sendai, M.; Miwa, T. US Patent 5,459,260, 1995.
6. (a) Basker, M. J.; Osborne, N. F. J. Antibiot. 1990, 3, 470.
(b) Bennet, I.; Broom, N. J. P.; Bruton, G.; Calvert, S.; Clarke,
B. P.; Coleman, K.; Edmondson, R.; Edwards, P.; Jones, D.;
Osborne, N. F.; Walker, G. J. Antibiot. 1991, 44, 70. (c)
Coleman, K.; Griffin, D. R. J.; Page, J. W. J.; Upshon, P. A.
Antimicrob. Agents Chemother. 1989, 33, 1580. (d) Coleman,
K. Exp. Opin. Invest. Drugs 1995, 4, 693. (d) Chaeuk, I.; Maiti,
S. N.; Micetich, R. G.; Daneshtalab, M.; Atchison, K.; Phi-
lips, O. A.; Kunugita, C. J. Antibiot. 1994, 47, 1031. (e) Phi-
lips, O. A.; Czajkowski, D. P.; Spevak, P.; Singh, M. P.;
Hanehara-Kunugita, C.; Hyodo, A.; Micetich, R. G.; Maiti,
S. N. J. Antibiot. 1997, 50, 350.
respectively (Scheme 3), were showing very limited
inhibitory activity (Table 1) thus pointing to the pivotal
role of the double bond in position 10. This result is in
accord with excellent stability of trinem antibiotics
towards b-lactamase degradation.5f
In Table 2, the results of inhibition by ethylidene tri-
nems 14a–14d against representative Class C b-lacta-
mase from Enterobacter cloacae 908R are listed along
with data for a benchmark Class A enzyme (Escherichia
coli TEM1).
Compounds 14a–d were shown to competitively inhibit
this Class C b-lactamase in micromolar range. We note
that the Class C b-lactamase enzymes such as E. cloacae
908R represent a phenomenon which compromises fur-
ther therapy by the existing of antibiotic/inhibitor com-
binations. Cephalosporins were found to be extremely
good substrates for Class C b-lactamases encoded by
the AmpC gene products.3 Our design strategy of
blocking the water access to the deacylation site in the
inhibitor–lactamase enzyme complex by addition of a
hydrophobic moiety onto penem scaffold resulted in
promising lead compounds which are being further
derivatised to increase their potency.9b
7. Solmajer, T.; Copar, A. Seventh ꢀ-Lactamase Workshop,
Holy Island (UK), 5–9 April, 1998.
8. Heinze-Krauss, I.; Angehrn, P.; Charnas, R. L.; Guberna-
tor, K.; Gutknecht, E. M.; Hubschwerlen, C.; Kania, M.;
Oefner, C.; Page, M. G. P.; Sogabe, S.; Specklin, J. L.; Wink-
ler, F. J. Med. Chem. 1998, 41, 3961.
9. (a) Copar, A.; Solmajer, T.; Anzie, B.; Kuzman, T.; Mesar,
T.; Kocjan, D. WO/98/27094: The Patent Cooperation Treaty
(PCT), 1998, Geneva, Switzerland, p 1. (b) Vilar, M.; Galeni,
M.; Solmajer, T.; Turk, B.; Frere, J.-M.; Matagne, A. Anti-
microb. Agents Chemother. 2001, 45, 2215.
10. Taniguchi, Y.; Inanaga, J.; Yamaguchi, M. Bull. Chem.
Soc. Jpn. 1981, 54, 3229.
11. Taguchi, K.; Westheimer, F. H. J. Org. Chem. 1971, 36,
1570.
12. Choi, W.-B.; Churchill, H. R. O.; Lynch, J. E.; Thomp-
son, A. S.; Humprey, G. R.; Volante, R. P.; Reider, P. J.;
Shinkai, I. Tetrahedron Lett. 1994, 35, 2275.
13. (a) Ernst, I.; Gosteli, J.; Woodward, R. B. J. Am. Chem.
Soc. 1979, 101, 6301. (b) Perrone, E.; Alpegiani, M.; Bedeschi,
A.; Guidici, F.; Franceschi, G. Tetrahedron Lett. 1984, 25,
2399. (c) Yoshida, A.; Tajima, Y.; Takeda, N.; Oida, S. Tet-
rahedron Lett. 1984, 25, 2793.
Conclusions
Sodium(8S,9R,10E)-1-aza-10-ethylidene-11-oxo-tricyclo-
(7.2.0.03,8)undec-2-en-2-carboxylate (14a) and sodium
(8S,9R,10E)-1-aza-10-ethylidene-11-oxo-6-thia-tricyclo-
(7.2.0.03,8)undec-2-en-2-carboxylate (14c) have been
shown to have activity at the enzyme b-lactamase I from
B. cereus and another Class A b-lactamase (E. coli
TEM1) comparable to the two reference compounds.
The potential shown by compounds 14a–d to inhibit
Class C b-lactamase E. cloacae 908R as well could be
useful in efforts to retain the b-lactam antibiotics activ-
ity in spite of the presence of Class C b-lactamases in
bacteria.
14. Hanessian, S.; Desilets, D.; Bennani, Y. L. J. Org. Chem.
1990, 55, 1235.
15. Jeffrey, P. D.; McCombie, S. W. J. Org. Chem. 1982, 47,
587.
Acknowledgements
16. All the compounds have been characterised by routine
analytical techniques:
Compound 14a: mp 222–225 ꢁC dec. (DMF/ether). 1H
NMR 300 MHz (d, ppm, DMSO-d6): 1.10–2.05 (m, 7H), 1.77
The authors would like to thank Professor J.-M. Frere
(Liege) and members of his group for continuing sup-
port and encouragement.

A. Copar et al. / Bioorg. Med. Chem. Lett. 12 (2002) 971–975
975
(d, J=7.1 Hz, 3H), 2.65–2.80 (dt, J=10.3, 5.1 Hz, 1H), 3.54–
3.64 (m, 1H), 4.53–4.61 (dd, J=10.3, 1.3 Hz, 1H), 6.21–6.30
(dq, J=7.1, 1.3 Hz, 1H). IR (nmax, cmꢀ1, KBr): 2933, 2858,
1740, 1592. MS: m/z 232 (Mꢀ23)+.
300 MHz (d, ppm, DMSO-d6): 1.80 (dd, J=7.1, 1.0 Hz,
3H), 2.34–2.64 (m, 5H), 3.04 (ddt, J=5.2, 10.6, 1.0 Hz,
1H), 3.98 (dt, J=12.6, 2.6 Hz, 1H), 4.66 (d, J=10.0 Hz,
1H), 6.34 (dq, J=7.1, 1.8 Hz, 1H). IR (nmax, cmꢀ1, KBr):
3375, 1757, 1606.
Compound 14b: mp 215–219 ꢁC (ether). H NMR 300 MHz
1
(d, ppm, DMSO-d6): 1.10–2.05 (m, 7H), 1.75 (d, J=7.1 Hz,
3H), 2.65–2.80 (dt, J=10.3, 5.1 Hz, 1H), 3.37–3.43 (m, 1H),
4.53–4.61 (dd, J=10.3, 1.3 Hz, 1H), 6.21–6.30 (dq, J=7.1,
1.3 Hz, 1H). IR (nmax, cmꢀ1, KBr): 2930, 2865, 1745, 1594.
MS: m/z 232 (Mꢀ23)+.
Compound 14d: mp 223–233 ꢁC (DMF/ether). 1H NMR
300 MHz (d, ppm, DMSO-d6): 1.76 (d, J=7.4, 1H), 1.86–1.98
(m, 1H), 2.41 (dt, J=12.9, 2.9 Hz, 1H), 2.48–2.66 (m, 2H),
2.70–3.20 (m, 1H), 3.15–3.25 (m, 1H), 3.96 (dd, J=14.7,
2.8 Hz), 4.08–4.12 (m, 1H), 6.18 (dq, J=7.0, 1.5 Hz). IR (nmax
,
Compound 14c: mp 195–225 ꢁC (DMF/ether). 1H NMR
cmꢀ1, KBr): 3427, 1752, 1594.