Stereoselective synthesis of optically active β-lactams, potential inhibitors of pilus assembly in pathogenic bacteria

Article abstract of DOI:10.1021/ol0059899

matrix presented Optically active β-lactams 3 are obtained in excellent yields (up to 93%) and with complete stereoselectivity from Meldrum's acid derivatives 1 and Δ²-thlazolines 2. A selective reduction to aldehydes 5 (R = Ar or CH₂

Full text of DOI:10.1021/ol0059899

ORGANIC
LETTERS
2000
Vol. 2, No. 14
2065-2067
Stereoselective Synthesis of Optically
Active â-Lactams, Potential Inhibitors of
Pilus Assembly in Pathogenic Bacteria
Hans Emtena1s,† Gabe Soto,‡ Scott J. Hultgren,‡ Garland R. Marshall,§ and
,†
Fredrik Almqvist*
Organic Chemistry, Umeå UniVersity, S-901 87 Umeå, Sweden, Department of
Molecular Microbiology, and Center of Molecular Design, Department of Molecular
Biology and Pharmacology, Washington UniVersity, 660 South Euclid,
Saint Louis, Missouri, 63130
fredrik.almqVist@chem.umu.se
Received April 26, 2000
ABSTRACT
Optically active â-lactams 3 are obtained in excellent yields (up to 93%) and with complete stereoselectivity from Meldrum’s acid derivatives
1 and ∆2-thiazolines 2. A selective reduction to aldehydes 5 (R ) Ar or CH Ar) was then accomplished by using DIBAL-H. This rigid framework,
2
with stereochemistry different than that of penicillin, is designed to be a suitable scaffold for the development of compounds inhibiting pilus
formation in uropathogenic Escherichia coli.
â-Lactams are known as a highly interesting class of
compounds, and research directed to their antibiotic proper-
ties is continuously attracting great interest in the scientific
community.1 More recently, the uses of â-lactams as reactive
intermediates and starting materials in a wide range of
synthetic settings have gained attention.2 One such example
is the exploration of the â-lactam synthon method (â-LSM)
by Ojima et al.3 Our interest in this field is directed to the
synthesis of a rigid framework to be used as scaffold for
development of potential novel antibiotics. Since an increas-
ing number of bacteria are developing resistance against the
drugs available today, such antibiotics should preferably be
directed to novel bacterial targets.4
Bacteria need to adhere to host tissue in order to cause
disease. Many pathogenic bacteria assemble pili, i.e., extra-
cellular protein organelles, to mediate attachment to host
epithelial cells. Pilus assembly is performed by periplasmic
chaperones, which bring subunits to the outer cell membrane
where they are incorporated in the growing pilus.5 A drug
inhibiting pilus formation, termed a pilicide, would therefore
have the potential of being an effective antibiotic.
† Umeå University.
‡ Department of Molecular Microbiology, Washington University.
§ Center of Molecular Design, Department of Molecular Biology and
Pharmacology, Washington University.
(1) For a review on the use of penicillin: see Nathwani, D.; Wood, M.
J. Drugs 1993, 45, 866. For other representative examples, see: (a)
Chemistry and Biology of â-lactam Antibiotics; Morin, R. B., Gorman, M.,
Eds.; Academic Press: New York, 1982; Vols. 1-3. (b) Brown, A. G.
Pure Appl. Chem. 1987, 59, 475.
(2) (a) The Chemistry of â-Lactams; Page, M. I., Ed.; Blackie Academic
and Professional: New York, 1996. (b) The Organic Chemistry of
â-Lactams; Georg, G. I., Ed.; VCH: New York, 1993.
(3) Ojima, I.; Lin, S. J. Org. Chem. 1998, 63, 224-225.
(4) (a) Niccolai, D.; Tarsi, L.; Thomas, R. J. Chem. Commun. 1997,
2333-2342. (b) Stephens, C.; Shapiro, L. Chem. Biol. 1997, 4, 637-641.
(5) Hultgren, S. J.; Abraham, S.; Caparon, M.; Falk, P.; St. Geme, J.
W., III; Normark, S. Cell 1993, 73, 887-901.
10.1021/ol0059899 CCC: $19.00 © 2000 American Chemical Society
Published on Web 06/20/2000

It has been shown, both by NMR spectroscopy6 and X-ray
crystallography,7 that synthetic peptides from the conserved
C-termini of the pilus proteins are bound by the PapD
chaperone (found in uropathogenic Escherichia coli, which
is the main cause of urinary tract infections). Furthermore,
some of the peptides were found to inhibit complex formation
between PapD and the adhesin PapG in an ELISA.6 It was
thus found that PapD binds polypeptides by anchoring of
the peptide carboxyl terminus to the side chains of Arg8 and
Lys112 (Figure 1), two residues that are invariant in all
in this report have different stereochemistry than the original
penicillin’s, thus having a chance to withstand enzymatic
degradation by penicillin-resistant bacteria. The overall
strategy has been to create small organic molecules with a
rigid framework, which would locate the pharmacophores
in the right position in space. In the crystal structures the
backbone atoms of the two C-terminal amino acids of the
peptides adopted a conformation which superimposed well
with a bicyclic â-lactam ring (Figure 1(a)). In addition, this
class of compounds allowed hydrophobic substituents (in-
dicated by R) to interact with the chaperone while maintain-
ing the important anchoring to Arg8 and Lys112. Moreover,
the crystal structures show that the C-terminal carboxyl group
is within such a distance from Lys112 and that replacing it
with an aldehyde would allow an imine to be formed with
Lys112. Although aldehydes may seem reactive, other alde-
hyde-containing compounds have previously been success-
fully employed as inhibitors and/or key substances in
inhibitor development.10
A ketene-imine cycloaddition is one possible way of
synthesizing the framework of interest.11 Bose et al. have
shown that penam derivatives were formed with the stereo-
chemistry that we desired if ∆2-thiazolines were condensed
with ketenes, generated in situ from acid chlorides and a
base.12 Unfortunately, the reported yields were as low as 11%
when ∆2-thiazolines such as 2 were used, and the compounds
were not optically active. An effort to increase the yield in
the cycloaddition step via a selenium-∆2-thiazoline was
reported later.13 Although the yields in the cycloadditions
reached 92%, four steps, with an overall yield of < 20%,
were required to prepare the selenium-containing thiazoline
and a reductive demethylselenation step had to be performed
after the cycloaddition. Furthermore, acid chlorides are not
suitable for the preparation of acyl ketenes 4, which are the
ketenes of choice for our purposes. Other reports14 have
suggested Meldrum’s acid derivatives (e.g., 1) as acyl ketene
precursors. The ease with which ketenes are formed from
derivatives 1 was attractive since a method tolerant of
different functional groups was required in order to give
optically active â-lactams. To the best of our knowledge there
are no previous reports on cycloadditions between ∆2-
thiazolines and ketenes generated in situ from Meldrum’s
acid derivatives. We now report that these two intermediates
react with each other under anhydrous and acidic conditions
to give the desired optically active â-lactams with complete
stereoselectivity15 (Scheme 1).
Figure 1. Careful examination of the crystal structures of peptide-
PapD complexes has resulted in the design of a rigid â-lactam
framework. These compounds superimpose well with the two
C-terminal amino acids (a), which allows hydrophobic interactions
via the R-group and mediates a possibility of forming an imine
with a lysine residue in the chaperone.
periplasmic chaperones and required for pilus assembly.
These data have recently been further confirmed when the
crystal structures of the complex between the chaperones
PapD and FimC together with the pilus subunits PapK and
FimH were solved.8
Direct use of peptides as drugs has several severe
drawbacks, i.e. peptides are poorly absorbed on oral admin-
istration, and they undergo rapid enzymatic degradation and
are usually quickly excreted.9 However, peptides may serve
as starting points for development of drugs as demonstrated
in the recent development of inhibitors of HIV protease
which slow the progress of AIDS.
On the basis of the crystal structures of peptide-PapD
complexes,7 â-lactams were selected as potential chaperone
inhibitors. It should be stressed that the â-lactams of interest
The ∆2-thiazoline 2 can be conveniently prepared in two
steps from commercially available L-cysteine methyl ester
(6) Flemmer Karlsson, K.; Walse, B.; Drakenberg, T.; Roy, S.; Bergquist,
K.-E.; Pinkner, J. S.; Hultgren, S. J.; Kihlberg, J. Bioorg. Med. Chem. 1998,
6, 2085-2101.
(10) (a) Higaki, J. N.; Chakravarty, S.; Bryant, C. M.; Cowart, L. R.;
Harden, P.; Scardina, J. M.; Mavunkel, B.; Luedtke, G. R.; Cordell, B. J.
Med. Chem. 1999, 42, 3889-3898. (b) Siev, D. V.; Semple, J. E. Org.
Lett. 2000, 2, 19-22.
(7) (a) Kuehn, M. J.; Ogg, D. J.; Kihlberg, J.; Slonim, L. N.; Flemmer,
K.; Bergfors, T.; Hultgren, S. J. Science 1993, 262, 1234-1241. (b) Soto,
G. E.; Dodson, K. W.; Ogg, D.; Liu, C.; Heuser, J.; Knight, S.; Kihlberg,
J.; Jones, C. H.; Hultgren, S. J. EMBO J. 1998, 17, 6155-6167.
(8) (a) Sauer, F. G.; Fu¨tterer, K.; Pinkner, J. S.; Dodson, K. W.; Hultgren,
S. J.; Waksman, G. Science 1999, 285, 1058-1061. (b) Choudhury, D.;
Thompson, A.; Stojanoff, V.; Langermann, S.; Pinkner, J.; Hultgren, S. J.;
Knight, S. D. Science 1999, 285, 1062-1066.
(11) (a) Staudinger, H. Liebigs Ann. Chem. 1907, 356, 51-123. (b) The
Organic Chemistry of â-Lactams; Georg, G. I., Ed.; VCH: New York, 1993;
Chapter 6.
(12) Bose, A. K.; Manhas, M. S.; Chib, J. S.; Chawla, H. P. S.; Dayal,
B. J. Org. Chem. 1974, 39, 2877-2884.
(13) Nagao, Y.; Kumagai, T.; Abe, S. T. T.; Ochiai, M.; Inoue, Y.; Taga,
T.; Fujita, E. J. Org. Chem. 1986, 51, 4739-4741.
(9) Plattner, J. J.; Norbeck, D. W. Obstacles to drug deVelopment from
peptide leads; Clark, C. R., Moos, W. R., Eds.; Ellis Harwood Ltd.-Halstead
Press: Chichester, England, 1989; pp 92-126.
(14) (a) Yamamoto, Y.; Watanabe, Y.; Ohnishi, S. Chem. Pharm. Bull.
1987, 35, 1860-1870. (b) Yamamoto, Y.; Watanabe, Y. Chem. Pharm.
Bull. 1987, 35, 1871-1879.
2066
Org. Lett., Vol. 2, No. 14, 2000

to the ketone in our ring-fused â-lactams 3 (∼164 ppm
instead of ∼200 ppm, as usually found for ketones), we
anticipated that a selective reduction of the ester functionality,
which had chemical shifts of approximately 170 ppm, could
succeed. Such a direct approach would thus avoid tedious
protection and deprotection steps in the synthesis of the target
â-lactam aldehydes 5. Indeed, this turned out to be correct
and a selective reduction of the aromatic esters (3a-3d) to
the corresponding aldehydes (5a-5d) could be achieved
(Scheme 2).
Scheme 1
Scheme 2
hydrochloride16 with an overall yield of >70%. Also the
derivatives 1 are easily synthesized from Meldrum’s acid
and acid chlorides in >80% yields.14a All derivatives
prepared thus far are storable in the freezer for months
without any detectable decomposition.
A series of penam derivatives were then synthesized with
aryl-, n-hexyl-, and cyclohexyl substituents; excellent yields
of the corresponding â-lactams 3a, 3b, 3e, and 3f were
accomplished (72-93%). The yields were somewhat lower
when a methylene group was introduced between the
carbonyl group and the aryl moiety, and 3c and 3d were
obtained in 62% and 65% yields, respectively. The low yield
for methyl ketone 3g, having a molecular weight of only
229 g/mol, was explained by losses during evaporation of
the solvent after chromatography.
Unfortunately, the isolated aldehydes in the aliphatic series
were unstable and decomposed.17 It is worth noting that the
reduction could only be accomplished using an electrophilic
reducing agent such as DIBAL-H. Use of nucleophilic
reducing agents (e.g., NaCNBH3, NaBH4, and LiBH4)
resulted in unreacted starting material or, at higher temper-
atures, decomposition. The aryl-substituted derivatives are
stable after freeze-drying and can be stored without decom-
position. These compounds will be evaluated as chaperone
inhibitors in the near future. The excellent and reproducible
yields in the synthesis of 3 and the convenience with which
Meldrum’s acid derivatives 1 are prepared constitute a
platform for future development of statistically diverse
libraries of optically active â-lactams.
After proper assignment of the NMR spectra (HMBC),
we realized that the carbonyl group in the ketone functional-
ity of compounds 3 was more shielded than the carbonyl
group in the ester functionality. Normally, one would expect
the ketone moiety to be reduced easier than the ester. Because
of the unusual chemical shift of the carbonyl corresponding
Acknowledgment. We warmly thank Professor Jan
Kihlberg for stimulating discussions and helpful comments
on this manuscript. We are grateful to the Swedish Natural
Science Research Council, the Knut and Alice Wallenberg
Foundation, and the foundation for technology transfer in
Umeå for financial support.
(15) General procedure for the synthesis of â-lactams: Meldrum’s
acid derivative 1b (630 mg, 2.11 mmol) and thiazoline 2 (185 mg, 1.27
mmol) were dissolved in dry benzene (28 mL) and cooled to 5 °C. HCl
gas was bubbled through the mixture for 10 min. The resulting turbid
mixture was heated for 1.5 h at 79 °C and then cooled to room temperature.
The resulting mixture was diluted with ethyl acetate and washed with ice-
cooled water and brine. The water phase was extracted twice with CH2Cl2,
and the combined organic extracts were dried with Na2SO4, filtered, and
concentrated. Flash chromatography (heptane:ethyl acetate 1:1 f 3:7) gave
3b (403 mg, 93%) as a white foam: [R]20 2.7° (c 1.77, CHCl3); IR λ
1739, 1664, 1508, 1388, 1211, 977, 773 cm-1D; 1H NMR (400 MHz, CDCl3)
δ 8.14 (d, 1H, J ) 8.23 Hz), 7.96 (d, 1H, J ) 8.23 Hz), 7.89 (m, 1H), 7.66
(dd, 1H, J1 ) 7.23 Hz, J2 ) 1.19 Hz), 7.46-7.59 (m, 3H), 6.93 (s, 1H),
5.89 (s, 1H), 5.43 (d, 1H, J ) 6.31 Hz), 3.84 (s, 3H), 3.60 (dd, 1H, J1 )
11.25 Hz, J2 ) 6.40 Hz), 3.32 (d, 1H, J ) 11.25); 13C NMR (100 MHz,
CDCl3) δ 169.7, 165.7, 161.2, 133.7, 131.7, 130.6, 129.7, 128.2, 127.3,
126.4, 124.9, 124.8, 103.4, 94.4, 60.9, 53.2, 32.1; HRMS (FAB+) calcd
for C18H16NO4S 342.0800; observed 342.0803.
Supporting Information Available: Experimental pro-
cedures and characterization data for all products. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL0059899
(17) Some of the ∆2-thiazoline 2 was re-formed during the reaction,
suggesting that a retro opening of the â-lactam ring, catalyzed by the
electrophilic reducing agent, was a competing side reaction.
(16) Almqvist, F.; Guillaume, D.; Hultgren S. J.; Marshall, G. R.
Tetrahedron Lett. 1998, 39, 2293-2294.
Org. Lett., Vol. 2, No. 14, 2000
2067