Synthesis of N-acetylglucosamine thiazoline/lipid II hybrids

Article abstract of DOI:10.1016/S0040-4039(00)02020-7

Potential inhibitors of transglycosylases involved in bacterial cell wall biosynthesis were synthesized by combining N-acetylglucosamine thiazoline, a potent inhibitor of β-hexosaminidase, with functional groups present in lipid II, the natural substrate of the transglycosylases.

Full text of DOI:10.1016/S0040-4039(00)02020-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 615–618
Synthesis of N-acetylglucosamine thiazoline/lipid II hybrids
Thomas K. Ritter and Chi-Huey Wong*
Department of Chemistry, The Scripps Research Institute and the Skaggs Institute for Chemical Biology,
10550 North Torrey Pines Road, La Jolla, CA 92037, USA
Received 20 September 2000; revised 7 November 2000; accepted 8 November 2000
Abstract—Potential inhibitors of transglycosylases involved in bacterial cell wall biosynthesis were synthesized by combining
N-acetylglucosamine thiazoline, a potent inhibitor of b-hexosaminidase, with functional groups present in lipid II, the natural
substrate of the transglycosylases. © 2001 Elsevier Science Ltd. All rights reserved.
Due to the emerging problem of antibiotic resistance in
bacteria, there is increasing interest in the exploration
of new antibiotic targets. Among the promising targets,
the transglycosylase involved in bacterial cell wall
biosynthesis is of particular interest, as the polysaccha-
ride backbone of peptidoglycan appears to remain
intact during the course of antibiotic resistance develop-
ment, while the peptide moiety shows a high frequency
of change.1 The transition states of the reaction cata-
lyzed by transglycosylase and that by N-acetyl-b-
hexosaminidases have been considered to feature
similar geometries and charge distributions (Fig. 1),2
and this proposition has served as the basis for our
inhibitor design. Since N-acetylglucosamine (GlcNAc)
thiazoline 3 has been shown to be a potent competitive
inhibitor of N-acetyl-b-hexosaminidase,3,4 we have
decided to use 3 as a core structure. Herein, we report
the chemistry developed for the syntheses of several
small molecules that combine thiazoline 3 with some of
the structural features of lipid II, the natural substrate
of the bacterial cell wall transglycosylases.
Our synthesis began with the protection of the 4- and
6-hydroxyl groups of thiazoline 34 as a p-methoxyben-
zylidene (PMB) acetal (Scheme 1). This was followed
by attachment of a lactic acid residue to the 3-hydroxyl
group to afford muramic acid (MurNAc) derivative 4.
The yield of this reaction, albeit moderate, was com-
parable to other GlcNAc to MurNAc conversions.5 A
portion of the peptide side chain of lipid II was intro-
duced into the molecule by EDCI promoted coupling
with a suitably protected dipeptide to yield 5. Subse-
quent deprotection of the PMB acetal turned out to be
problematic. Treatment with aqueous acids led to faster
Figure 1. Proposed transition state of the transglycosylase reaction in bacterial peptidoglycan synthesis.
Keywords: carbohydrates; thiazolines; transglycosylase inhibitors.
* Corresponding author. Fax: (858) 784-2409; e-mail: wong@scripps.edu
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)02020-7

616
T. K. Ritter, C.-H. Wong / Tetrahedron Letters 42 (2001) 615–618
opening of the thiazoline ring than cleavage of the
acetal, and oxidative conditions oxidized the sulfur
atom of the thiazoline moiety. Eventually, it was found
that ethanedithiol in the presence of one equivalent
camphorsulfonic acid removed the benzylidene group
efficiently. Cleavage of the methyl esters then afforded
target molecule 6.6
core structure. Per-O-acetylated b-glucosamine 17 was
coupled to octanoic acid, followed by conversion of the
amide bond to a thioamide, which was cyclized under
the reaction conditions to thiazoline 8. Compound 8
was then subjected to the reaction sequence outlined
above, yielding target molecule 11.
Another consideration of the transglycosylation reac-
tion is the electrostatic and hydrogen bonding interac-
tions between the pyrophosphate moiety of the
In order to mimic the oligoisoprenyl membrane anchor
of lipid II, a long alkyl chain was incorporated into the
Scheme 1. (a) Octanoic acid, EDCl, DMAP, CH2Cl2, 95%; (b) Lawesson’s reagent, toluene, 65°C, 98% (2), 88% (7); (c) TEA,
MeOH/H2O, 93% (2), quant. (7); (d) PMB dimethyl acetal, CSA, CH3CN, 68% (3), 82% (8); (e) (S)-chloropropionic acid, NaH,
THF, 55% (3), 23% (8); (f) Ala-D-Glu(OMe)-OMe, EDCl, HOBt, NMM, CH2Cl2/DMF, 71% (4), 95% (9); (g) ethanedithiol, CSA,
CH3CN, 86% (5), 88% (10); (h) LiOH, MeOH/H2O, 80% (5), 90% (10).
Scheme 2. (a) EDCl, DMAP, CH2Cl2, 85%; (b) Lawesson’s reagent, toluene, 80°C, 71%; (c) N2H4–HOAc, DMF, 78%; (d) DAST,
CH2Cl2, −78°C, 35% (e) chloroacetic anhydride, pyridine, 69%; (f) P(OEt)3, NaI, CH3CN, 65°C, quant; (g) Lawesson’s reagent,
toluene, 65°C, 74%.

T. K. Ritter, C.-H. Wong / Tetrahedron Letters 42 (2001) 615–618
617
Scheme 3. (a) EDCl, DMAP, CH2Cl2, 91%; (b) H2, 10% Pd/C, MeOH, quant.; (c) NIS, PPh3 resin, DMF, 80%; (d) P(OMe)3,
80°C, 2 days, 89%; (e) Lawesson’s reagent, toluene, 65°C, 81%; (f) PhSH, TEA, THF, quant.; (g) n-octanol, DEAD, PPh3,
CH2Cl2, 73%; (h) PhSH, TEA, THF, quant.; (i) TEA, MeOH/H2O, 51% (23), 90% (26).
substrate and the enzyme.8 It is therefore desirable to
include a charged group in the inhibitor design. Toward
this end, 1 was coupled to TBS-protected hydroxyacetic
acid 12 to provide amide 13 (Scheme 2). Surprisingly,
reaction of 13 with Lawesson’s reagent gave the uncy-
clized thioamide 14 instead of the desired thiazoline.
Even activation of the anomeric position to fluoride 15
did not force ring closure. This lack of reactivity is
most likely due to a reduction in nucleophilicity of the
thioamide by the neighboring oxygen substituent of the
protected hydroxyl group.
corresponding thiazoline and monodeprotected with
thiophenol in the presence of triethyl amine to provide
23. This procedure was followed by hydrolysis of the
acetate esters to give 24.9 A membrane anchor was also
attached to phosphonate 23, by coupling to octanol in
a Mitsunobu reaction. Deprotection of 25 afforded
target molecule 27.
Investigation of the activities of molecules 6, 11, 24 and
27 as inhibitors of the bacterial cell wall transglycosy-
lases is under way.
In order to circumvent this problem, a less electronega-
tive heteroatom was introduced as substituent on the
References
thiazoline ring. Compound
1 was reacted with
chloroacetic anhydride. The chloride 16 was then con-
verted under Arbuzov conditions to phosphonate 17.
This substate was readily cyclized to thiazoline 18 when
submitted to Lawesson’s reagent. Unfortunately, 18
was found to be very unstable and deprotection of the
phosphonate moiety could not be achieved under Lewis
acidic or basic conditions.
1. Walsh, C. T.; Fisher, S. L.; Park, I. S.; Prahalad, M.; Wu,
Z. Chem. Biol. 1996, 3, 21.
2. Ferse, F.-T.; Floeder, K.; Hennig, L.; Findeisen, M.;
Welzel, P. Tetrahedron 1999, 55, 3749.
3. Bovin, N. V.; Zurabyan, S. E.; Khorlin, A. Ya. Izv. Akad.
Nauk. SSSR, Ser. Khim. 1981, 2, 441.
4. Knapp, S.; Vocdadlo, D.; Gao, Z.; Kirk, B.; Lou, J.;
Withers, S. G. J. Am. Chem. Soc. 1996, 118, 6804.
5. Hecker, S. J.; Minich, M. L. J. Org. Chem. 1980, 55, 6051.
At this point, it was decided to remove the heteroatom
substituent from the proximity of the thiazoline ring by
an additional methylene group spacer. Coupling of 1 to
19 was followed by deprotection of the benzyl ether and
conversion of the resulting hydroxyl group to halogen
(Cl, Br, I) substituents (Scheme 3). Of the three halo-
genides, only iodide 21 reacted in the Arbuzov reaction
to yield 22. Phosphonate 22 was then converted to the
1
6. Compound 6: H NMR (500 MHz, D2O) l 6.31 (1H, d,
J=7.0 Hz, H1), 4.60 (1H, m, GluHa), 4.44 (1H, q, AlaHa,
J=7.2), 4.34 (1H, q, LacHa, J=6.8), 4.16 (1H, dd, J=8.4,
4.4, H4), 4.12 (1H, dd, J=3.7, 2.2 Hz, H3), 3.80–3.77 (2H,
m, H2,6), 3.64 (1H, dd, J=12.3, 6.8 Hz, H6%), 3.31 (1H,
ddd, J=9.1, 6.9, 2.3 Hz, H5), 2.29 (3H, d, J=2.6 Hz, Me),
2.18 (2H, m, GluHg), 2.06 (1H, m, GluHb), 1.88 (1H, m,

618
T. K. Ritter, C.-H. Wong / Tetrahedron Letters 42 (2001) 615–618
GluHb%), 1.45 (3H, d, J=7.0 Hz, AlaHb), 1.43 (3H, d,
9. Compound 24: 1H NMR (500 MHz, D2O) l 6.34 (1H,
d, J=7.0 Hz, H1), 4.46 (1H, t, J=5.7 Hz, H4), 4.25
(1H, dd, J=4.6, 3.5 Hz, H3), 3.78 (1H, dd, J=12.5, 2.2
Hz, H6), 3.63 (1H, dd, J=12.5, 7.0 Hz, H6), 3.62 (1H, m,
H2), 3.55 (3H, d, J=10.3 Hz, OMe), 3.39 (1H, ddd,
J=9.2, 7.0, 2.4 Hz, H5), 2.79–2.73 (2H, m,
NHCOCH2CH2P), 1.98–1.89 (2H, m, NHCOCH2CH2P);
13C NMR (125 MHz, D2O) l 88.0, 78.3, 74.6, 72.1, 69.9,
62.1, 51.7; MS (ESI) pos. 328 [MH+], 350 [MNa+]; neg.
326 [M−H]−.
J=6.6 Hz, LacHb); 13C NMR (125 MHz, D2O) l 182.4,
178.8, 175.8, 174.2, 172.2, 88.3, 79.3, 76.8, 76.0, 73.9, 68.4,
62.1, 55.4, 49.9, 34.3, 29.0, 19.9, 18.9, 17.6; HR-MS
(MALDI) calcd for C19H29N3O10S [M+Na+] 514.1466,
found 514.1480.
7. Medgyes, A.; Farkas, E.; Lipta´k, A.; Pozsgay, V. Tetra-
hedron 1997, 53, 4159.
8. Sears, P.; Wong, C.-H. Angew. Chem., Int. Ed. 1999, 38,
2300.
.
.