Synthesis and evaluation of tetrahedral intermediate mimic inhibitors of 3-deoxy-D-manno-octulosonate 8-phosphate synthase

Article abstract of DOI:10.1016/j.bmcl.2011.12.025

3-Deoxy-D-manno-octulosonate 8-phosphate (KDO8P) synthase catalyses the first committed step in the biosynthesis of 3-deoxy-D-manno-octulosonate (KDO), an important component of the lipopolysaccharide of Gram-negative bacteria. The pathway for KDO biosynt

Full text of DOI:10.1016/j.bmcl.2011.12.025

Bioorganic & Medicinal Chemistry Letters 22 (2012) 907–911
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of tetrahedral intermediate mimic inhibitors
of 3-deoxy-D-manno-octulosonate 8-phosphate synthase
⇑
Aidan N. Harrison, Sebastian Reichau, Emily J. Parker
Biomolecular Interaction Centre and Department of Chemistry, University of Canterbury, Christchurch 8040, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history:
3-Deoxy-
D-manno-octulosonate 8-phosphate (KDO8P) synthase catalyses the first committed step in the
Received 10 November 2011
Revised 2 December 2011
Accepted 5 December 2011
Available online 10 December 2011
biosynthesis of 3-deoxy-
D
-manno-octulosonate (KDO), an important component of the lipopolysaccha-
ride of Gram-negative bacteria. The pathway for KDO biosynthesis has been identified as a potential tar-
get of antibacterial drug design. The reaction catalysed by KDO8P synthase is an aldol-like condensation
between phosphoenolpyruvate (PEP) and D-arabinose 5-phosphate (A5P) and proceeds through a bis-
phosphorylated tetrahedral intermediate. In this study a bisphosphate analogue of the tetrahedral inter-
mediate was synthesised and was found to inhibit the metal-dependent KDO8P synthase from Neisseria
meningitidis and the metal-dependent KDO8P synthase from Acidithiobacillus ferrooxidans with inhibition
constants in the low micromolar range. Additionally, monophosphorylated inhibitors were synthesised to
determine the relative importance of the two phosphate groups of this bisphosphate analogue for
enzyme inhibition. The removal of either of these two phosphate groups gave less potent inhibitors for
both enzymes.
Keywords:
KDO Biosynthesis
KDO8P synthase
Intermediate mimic
Mechanism-based inhibitor design
Lipopolysaccharide biosynthesis
Ó 2011 Elsevier Ltd. All rights reserved.
3-Deoxy-
D
-manno-octulosonate phosphate (KDO8P) synthase
amino acid substitution in the metal binding site, where a metal
binding Cys in the metal-dependent enzyme is substituted for a con-
served Asn in the metal-independent form.10
(Fig. 1) catalyses the aldol condensation between phosphoenolpyr-
uvate (PEP, 1) and arabinose 5-phosphate (A5P, 2) to give the eight
carbon phosphorylated sugar, 3-deoxy-
(KDO8P, 3). This reaction is the first committed step in the biosyn-
thesis of 3-deoxy-
-manno-octulosonate (KDO),1 an eight carbon
sugar that is an important component of lipopolysaccharide (LPS)
which is found in the outer membrane of Gram-negative bacteria.
Previous studies have shown that disruption of the KDO pathway
leads to cell wall disruption and eventual death of the bacterium,
and therefore KDO8P synthase has been identified as a potential
target for antibacterial drug design.2
D
-manno-octulosonate
Many of the mechanistic aspects of the KDO8P synthase reaction
have been determined. It is known that the si face of PEP approaches
the re face of A5P to form the new carbon–carbon bond between the
C3 carbon of PEP and the C1 carbon of A5P.12 The phosphate elimina-
tion is known to occur via breakage of the carbon–oxygen bond of
the C2 carbon from PEP and the bridging oxygen of the phosphate es-
ter.13 Computational studies have suggested that this gives an
oxocarbenium ion 4 (Fig. 1), of which the electrophilic C2 carbon is
attacked by a nucleophilic water molecule.14 This gives the tetrahe-
dral intermediate 5 which has been directly observed by mass spec-
trometry (Fig. 1).15 The same computational studies have suggested
that the divalent metal ion is not directly involved in the reaction
mechanism, as is thought to occur in the related aldolase 3-deoxy-
D
There are two groups of KDO8P synthases that differ in their
requirement for a divalent metal ion for catalytic activity. The
KDO8P synthases from Escherichia coli3 and Neisseria meningitidis4
do not require a metal ion for activity, whereas a divalent metal
ion is necessary for catalysis by the KDO8P synthases from Aquifex
aeolicus, Aquifex pyrophilus, Helicobacter pylori, Acidithiobacillus fer-
rooxidans and Chlamydia psittaci.5–8
D
-arabino-heptulosonate 7-phosphate synthase,16 but plays a struc-
tural role in orientating the substrates to lower the activation energy
of the enzyme reaction, a role that is easily mimicked by amino acid
side chains in the metal-independent KDO8P synthases, consistent
with relatively facile interconversion between the metal-dependent
and metal-independent groups.5,10,17,18
The structures of both metal-independent9,10 and metal-depen-
dent11 KDO8P synthases have been determined by X-ray crystallog-
raphy. The functional forms of these enzymes are homotetramers of
(b
a
)
barrel subunits. The active sites are almost completely con-
Several inhibitors for KDO8PS have been synthesized and
tested.19–22 The most effective of these inhibitors is aminophosph-
onate 7 (Fig. 2) which was designed to mimic the oxocarbenium
ion intermediate (4, Fig. 1),19 with an inhibition constant, Ki, of
0.37
Compounds resembling the tetrahedral intermediate or substrate
8
served between the two groups of the enzyme, apart from a single
lM
for the metal-independent KDO8PS from E. coli.23
⇑
Corresponding author. Tel.: +64 3 364 2871; fax: +64 3 364 2110.
E-mail address: Emily.parker@canterbury.ac.nz (E.J. Parker).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.12.025

908
A. N. Harrison et al. / Bioorg. Med. Chem. Lett. 22 (2012) 907–911
-2O3PO
HO
A5P
2
HO
PEP
2-
OPO3
HO
1
HO
KDO8P Synthase
phosphate
OH
-2
PO3
O
O
-
O
HO
CO2
OH
-
CO2
KDO8P
H
3
O
H
-2O3PO
-2O3PO
HO
-2O3PO
HO
HO
HO
HO
OH
OH
O
HO
HO
HO
HO
OH
-2
-2
PO3
PO3
O
O
H
O
CO2
- H
O
H
-
-
CO2
CO2
Oxocarbenium
open chain KDO8P
phosphohemiketal
ion
6
intermediate
4
5
Figure 1. The reaction catalysed by KDO8P synthase. The transformation of PEP (1) and A5P (2) to KDO8P (3) is shown as well as the proposed oxocarbenium ion (4) and
known tetrahedral intermediate (5) as intermediates of the reaction mechanism.
this tetrahedral intermediate to act as inhibitors for both metal-
dependent and metal-independent forms of this enzyme.
O
P
O-
O
P
O-
O-
O-
O
P
OH OH
An initial examination of the potential for mimics of the tetra-
hedral intermediate to act as inhibitors for this enzyme was carried
out by testing the phospholactates 9 and 10 (Fig. 2).26 Both enan-
tiomers were found to show relatively poor, but competitive inhi-
bition with respect to PEP, of both the metal-dependent KDO8P
synthase from A. ferrooxidans and the metal-independent KDO8P
synthase from N. meningitidis (Table 1). There was little discrimina-
tion between the enantiomeric phospholactates 9 and 10, with
-O
-O
H2N
NH
O
-
-
CO2
CO2
OH OH
8
7
O-
O-
O
P
O
P
O
O
O-
O-
O-
O-
P
-O
-O
O
O
O
(2R)-2-phospholactate 9 giving Ki values of 870 90
lM and
P
-
-
-
O
CO2
CO2
CO2
390 70 M for the KDO8P synthase from N. meningitidis and A.
l
ferrooxidans respectively, whereas the (2S)-2-phospholactate 10
gave slightly poorer inhibition of both enzymes. The configuration
of the stereogenic centre at C2 of the tetrahedral intermediate 5
(Fig. 1) is unclear27,28 and it is possible that the hydroxyl group
is required to be present to create a preference for a specific config-
uration. Intriguingly the ‘PEP head’ of the aminophosphate 7, gly-
9
10
11
13
O-
O
O
P
O-
O
P
OH
-O
-O
-
-
O
CO2
HO
CO2
phosate
relatively poor inhibitor of KDO8PS with Ki values of 260 40
and 330 40
8
(Fig. 2), was also examined and found to be
a
M
12
l
l
M for the enzymes from N. meningitidis and A. ferro-
Figure 2. Aminophosphonate 7 and molecules tested as inhibitors of KDO8PS in
this study (8–13).
oxidans respectively. This finding indicates the importance of the
A5P-mimicking component of animophosphonate 7. Based on
these inhibition results three inhibitors were proposed. The bis-
phosphate tetrahedral intermediate mimic 11 (Fig. 2) contains
both charged PEP functional groups and a primary phosphate
group linked through a hydrocarbon tail. A chain length of eight
was selected in order to allow appropriate spacing of the primary
phosphate group to mimic the phosphate of the tetrahedral inter-
mediate derived from substrate A5P. The other two compounds 12
and 13 (Fig. 2) lack either the primary or secondary phosphate
group and were designed to investigate the relative importance
of the two phosphate groups for the inhibition of KDO8P synthase.
Following our recently published procedure,25 the syntheses
for all three compounds proceeded via a common precursor, the
complex, but lacking the A5P-like phosphate moiety have also
been tested as inhibitors against KDO8P synthase from E. coli and
H. pylori showing complex, but generally poor inhibition. The best
inhibitor of this series has an IC50 value of 132 lM against the
KDO8PS from H. pylori.21,24
Recently we described a tetrahedral intermediate mimic as the
first potent inhibitor of 3-deoxy- -arabinose-heptulosonate 7-phos-
D
phate synthase from Mycobacterium tuberculosis,25 an enzyme
which shares many elements of its structure and reaction mecha-
nism with KDO8P synthase. Here we show that this strategy is trans-
ferable to the synthesis of simple tetrahedral intermediate mimics
for the inhibition of KDO8PS. These compounds incorporate, with
appropriate spacing, the key charged carboxylate and phosphate
functionalities derived from PEP and the phosphate group of A5P.
These findings highlight the potential for compounds mimicking
a-hydroxy ester 18, which was synthesised in four steps from
1,6-hexanediol (Scheme 1). 1,6-Hexanediol 14, was monoprotected
with tert-butyldimethylsilyl chloride, and this alcohol 15 was then
oxidised using Dess–Martin periodinane29 to give the aldehyde 16.

A. N. Harrison et al. / Bioorg. Med. Chem. Lett. 22 (2012) 907–911
909
Table 1
a
-hydroxy ester 18 which was purified by flash chromatography.
Inhibition constants of the inhibitors tested in this studya
The reaction sequence from 15 to 18 did not require purification
of the intermediates and could be conveniently carried out in
one day.
Inhibitor
Ki Nme (
lM)
Ki Afe (lM)
O
O-
The synthesis of bisphosphate 11 proceeded by deprotection of
O-
P
the silyl ether of
a-hydroxy ester 18 with tert-butylammonium
fluoride33 to give the diol 19. This diol was then diphosphorylated
in two steps via the phosphite, which was subsequently oxidised
with meta-chloroperbenzoic acid to give the protected bisphos-
phate 20 (Scheme 2).34Bisphosphate 20 was then subjected to final
deprotection by removal of the phosphate benzyl ether groups by
hydrogenolysis, followed by the hydrolysis of the isopropyl ester
group under basic conditions to give the bisphosphate 11.
260 40
330 40
H2N
-
CO2
8
O
P
O-
O-
O
9
870 90
390 70
-
CO2
The synthesis of monophosphate inhibitor 12 proceeded from
the a-hydroxy ester 18 by monophosphorylation of the C2 hydroxyl
group, prior to deprotection of the silyl ether group of ester 21 to give
the benzyl protected monophosphate 22 (Scheme 3). The phosphate
benzyl ether groups were deprotected by hydrogenolysis, however
purification by anion exchange chromatography proved to be diffi-
cult due to the poor resolution of the compound from the column.
Passage of the crude product through a DOWEX-H+ column proved
O
P
O-
O
O-
1200 50
1000 220
-
CO2
10
to be a more efficient alternative purification method, with 1H, 13
C
O-
O-
O
O
P
and 31P NMR analysis indicating no impurities. The monophosphate
inhibitor 12 dissolved in an aqueous solution was lyophilised and
the compound obtained as a white powder.
-O
O
P
7.9 1.6
20
3
-O
-
O
CO2
A similar route was followed for the synthesis of monophos-
phate inhibitor 13 (Scheme 4). The C2 hydroxyl group of a-hydroxy
ester 18 was benzylated before removal of the silyl ether. Alcohol
24 was then phosphorylated to give the protected monophosphate
25. Deprotection and purification by anion exchange chromatogra-
phy gave the product 13 as a white powder.
11
O-
O-
O
O
P
1000 80
540 50
-
HO
CO2
The ability of the three compounds (11, 12, 13) to inhibit the
metal-independent KDO8P synthase from N. meningitidis and the
metal-dependent KDO8P synthase from A. ferrooxidans was deter-
mined by kinetic assay (Table 1). The bisphosphate inhibitor 11
was found to give competitive inhibition with respect to PEP with
12
OH
CO2
O
P
-O
-O
-
270 40
190 40
O
13
an inhibition constant of 7.9 1.6
lM for the KDO8P synthase from
aInhibitors were assayed on KDO8P synthase from N. meningitidis (Nme) and
A. ferrooxidans (Afe) KDO8P synthases using a direct assay system following the loss
of PEP at 232 nm. All compounds were found to be competitive inhibitors with
respect to PEP.
N. meningitidis and 20 M against KDO8P synthase from A. ferr-
3
l
oxidans. These Ki values are approximately 2–3 fold higher than the
Km values for PEP for each enzyme, making the tetrahedral inter-
mediate mimic 11 the second most potent inhibitor for KDO8P
synthase, behind the aminophosphonate 7. The importance of both
the A5P and PEP phosphate moieties for the inhibition of KDO8P
synthase is highlighted by the results for the inhibition of both
KDO8P synthase enzymes by the monophosphate inhibitors.
Aldehyde 16 was then used to produce the chloroepoxide 17 using
a modified Darzen’s condensation.30,31 The crude epoxide 17 was
then reduced using sodium cyanoborohydride32 to give the
a
OH
OH
HO
TBDMSO
14
15
b
Cl
CO2iPr
c
O
TBDMSO
TBDMSO
TBDMSO
16
O
17
18
d
OH
CO2iPr
Scheme 1. Synthesis of hydroxy ester 18. (a) TBDMSCl, imidazole, THF (72%), (b) Dess–Martin periodinane, CH2Cl2, (c) Cl2CHCO2iPr, iPrOK, iPrOH, Et2O, (d) NaCNBH3, iPrOH
(26% from 15).

910
A. N. Harrison et al. / Bioorg. Med. Chem. Lett. 22 (2012) 907–911
OH
CO2iPr
OH
a
HO
TBDMSO
CO2iPr
19
18
b,c
O
O-
O
O
OBn
OBn
P
P
O-
O
d,e
O
P
-O
O
BnO
BnO
-O
P
O
CO2iPr
-
O
CO2
20
11
Scheme 2. Synthesis of bisphosphate inhibitor 11. (a) TBAF, THF (83%), (b) 1H-tetrazole, (iPr)2NP (OBn)2, CH2Cl2, (c) m-CPBA (65% from 19), (d) H2, Pd/C, EtOAc, (e) KOH/H2O,
DOWEX-H+ (2% from 20).
O
O
OBn
OBn
P
OH
CO2iPr
a,b
TBDMSO
TBDMSO
CO2iPr
18
12
21
c
O-
O-
O
O
O
O
OBn
OBn
P
P
d,e
-
HO
CO2
HO
CO2iPr
22
Scheme 3. Synthesis of monophosphate inhibitor 12. (a) 1H-tetrazole, (iPr)2NP(OBn)2, CH2Cl2, (b) m-CPBA (80% from 18), (c) TBAF, THF (56%), (d) H2, Pd/C, EtOAc, (e) KOH/
H2O, DOWEX-H+ (41% from 22).
OH
CO2iPr
OBn
CO2iPr
a
TBDMSO
TBDMSO
23
24
18
b
OBn
CO2iPr
OBn
CO2iPr
O
P
c,d
BnO
BnO
HO
O
O
25
13
e,f
OH
CO2
-O
O
P
-O
-
Scheme 4. Synthesis of monophosphate inhibitor 13. (a) Ag2O, KI, BnBr, CH2Cl2 (66%), (b) TBAF, THF (83%), (c) 1H-tetrazole, (iPr)2NP(OBn)2, CH2Cl2, (d) m-CPBA (84% from 24),
(e) H2, Pd/C, EtOAc, (f) KOH/H2O, DOWEX-H+ (12% from 25).
Monophosphate inhibitor 12, lacking the A5P-like phosphate moi-
ety inhibited the KDO8P synthase from N. meningitidis with an
itor to bind in both substrate-binding sites at either end of the
enzyme active site through charged groups provides more effective
inhibition of KDO8P synthase.
inhibition constant of 1000 80
A. ferrooxidans with an inhibition constant of 540 50
l
M and the KDO8P synthase from
M. These
l
In conclusion an inhibitor designed as a tetrahedral intermedi-
ate mimic, incorporating only key charged moieties but without
the stereochemically and synthetically complex hydroxyl function-
ality of A5P was able to effectively inhibit both the metal-
independent KDO8P synthase from N. meningitidis and the
metal-dependent KDO8P synthase from A. ferrooxidans. This com-
pound was considerably more effective than inhibitors that lacked
either of the phosphate groups, indicating that these charged
groups both play an important role in the inhibition of KDO8P
values are similar to the Ki values found for the average of the Ki
values for phospholactates 9 and 10, which is unsurprising since
the hydrocarbon tail of 12 is unlikely to form contacts with KDO8P
synthase active site residues, and therefore the presence of this tail
does not incur any inhibitory advantage. Interestingly monophos-
phate inhibitor 13, lacking the PEP phosphate moiety, inhibited
both KDO8P synthase enzymes more effectively than monophos-
phate inhibitor 12. This result suggests that the ability of the inhib-

A. N. Harrison et al. / Bioorg. Med. Chem. Lett. 22 (2012) 907–911
911
11. Duewel, H. S.; Radaev, S.; Wang, J.; Woodard, R. W.; Gatti, D. L. J. Biol. Chem.
2001, 276, 8393.
12. Dotson, G. D.; Nanjappan, P.; Reily, M. D.; Woodard, R. W. Biochemistry 1993,
32, 12392.
13. Hedstrom, L.; Abeles, R. Biochem. Biophys. Res. Commun. 1988, 157, 816.
14. Tao, P.; Schlegel, H. B.; Gatti, D. L. J. Inorg. Biochem. 2010, 104, 1267.
15. Li, Z.; Sau, A. K.; Shen, S.; Whitehouse, C.; Baasov, T.; Anderson, K. S. J. Am.
Chem. Soc. 2003, 125, 9938.
synthases from both bacterial sources. These findings suggest that
the charged functionalities of both substrates are important for
inhibitor binding, and that further elaboration of the tetrahedral
bisphosphate inhibitor 11 to include the A5P-derived hydroxyl
moieties may show increased inhibition of the enzyme.
16. Ahn, M.; Pietersma, A. L.; Schofield, L. R.; Parker, E. J. Org. Biomol. Chem. 2005, 3,
4046.
Acknowledgments
17. Shulami, S.; Furdui, C.; Adir, N.; Shoham, Y.; Anderson, K. S.; Baasov, T. J. Biol.
Chem. 2004, 279, 45110.
18. Oliynyk, Z.; Briseno-Roa, L.; Janowitz, T.; Sondergeld, P.; Fersht, A. R. Protein
Eng., Des. Sel. 2004, 17, 383.
19. Du, S.; Tsipori, H.; Baasov, T. Bioorg. Med. Chem. Lett. 1997, 7, 2469.
20. Sheffer-Dee-Noor, S.; Belakhov, V.; Baasov, T. Bioorg. Med. Chem. Lett. 1993, 3,
1583.
21. Xu, X.; Wang, J.; Grison, C.; Petek, S.; Coutrot, P.; Birck, M. R.; Woodard, R. W.;
Gatti, D. L. Drug Design and Discovery 2003, 18, 91.
22. Belakhov, V.; Dovgolevsky, E.; Rabkin, E.; Shulami, S.; Shoham, Y.; Baasov, T.
Carbohydr. Res. 2004, 339, 385.
23. Du, S.; Faiger, H.; Belakhov, V.; Baasov, T. Bioorg. Med. Chem. 1999, 7, 2671.
24. Grison, C.; Petek, S.; Finance, C.; Coutrot, P. Carbohydr. Res. 2005, 340, 529.
25. Reichau, S.; Jiao, W.; Walker, S. R.; Hutton, R. D.; Baker, E. N.; Parker, E. J. J. Biol.
Chem. 2011, 286, 16197.
A.N.H. was supported by a Maurice Wilkins Centre for Molecu-
lar Biodiscovery Scholarship and S.R. was supported by a New Zea-
land International Doctoral Scholarship.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.12.025.
References and notes
1. Levin, D. H.; Racker, E. J. Biol. Chem. 1959, 234, 2532.
2. Rick, P. D.; Fung, L. W. M.; Ho, C.; Osborn, M. J. J. Biol. Chem. 1977, 252, 4904.
3. Ray, P. H.; Benedict, C. D. J. Bacteriol. 1980, 142, 60.
4. Ahn, M.; Cochrane, F. C.; Patchett, M. L.; Parker, E. J. Bioorg. Med. Chem. 2008, 16,
9830.
5. Duewel, H. S.; Woodard, R. W. J. Biol. Chem. 2000, 275, 22824.
6. Shulami, S.; Yaniv, O.; Rabkin, E.; Shoham, Y.; Baasov, T. Extremophiles 2003, 7,
471.
7. Birck, M. R.; Woodard, R. W. J. Mol. Evol. 2001, 52, 205.
8. Allison, T. M.; Yeoman, J. A.; Hutton, R. D.; Cochrane, F. C.; Jameson, G. B.;
Parker, E. J. Biochim. Biophys. Acta, Proteins Proteomics 2010, 1804, 1526.
9. Radaev, S.; Dastidar, P.; Patel, M.; Woodard, R. W.; Gatti, D. L. J. Biol. Chem. 2000,
275, 9476.
26. Walker, S. R.; Cumming, H.; Parker, E. J. Org. Biomol. Chem. 2009, 7, 3031.
27. Kona, F.; Xu, X.; Martin, P.; Kuzmic, P.; Gatti, D. L. Biochemistry 2007, 46, 4532.
28. Allison, T. M.; Hutton, R. D.; Cochrane, F. C.; Yeoman, J. A.; Jameson, G. B.;
Parker, E. J. Biochemistry 2011, 50, 3686.
29. Dess, P. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
30. Coutrot, P.; Grison, C.; Tabyaoui, M.; Czernecki, S.; Valery, J. M. J. Chem. Soc.,
Chem. Commun. 1988, 1515.
31. Castro, B.; Villieras, J.; Ferracutti, N. C. R. Acad. Sci., Paris, Ser. C 1969, 268, 1403.
32. Grison, C.; Coutrot, F.; Comoy, C.; Lemilbeau, C.; Coutrot, P. Tetrahedron Lett.
2000, 41, 6571.
33. Corey, E. J.; Venkateswarlu, A. J. Am. Chem. Soc. 1972, 94, 6190.
34. Nurminen, E. J.; Mattinen, J. K.; Lonnberg, H. J. Chem. Soc. Perkin Trans. 2 1998,
1621.
10. Cochrane, F. C.; Cookson, T. V. M.; Jameson, G. B.; Parker, E. J. J. Mol. Biol. 2009,
390, 646.