Inhibitors of the kinase IspE: Structure-activity relationships and co-crystal structure analysis

Article abstract of DOI:10.1039/b804375b

Enzymes of the non-mevalonate pathway for isoprenoid biosynthesis are therapeutic targets for the treatment of important infectious diseases. Whereas this pathway is absent in humans, it is used by plants, many eubacteria and apicomplexan protozoa, includ

Full text of DOI:10.1039/b804375b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Inhibitors of the kinase IspE: structure–activity relationships and co-crystal
structure analysis†
Anna K. H. Hirsch,a Magnus S. Alphey,b Susan Lauw,c Michael Seet,a Luzi Barandun,a Wolfgang Eisenreich,c
Felix Rohdich,*c William N. Hunter,*b Adelbert Bacherc and Franc¸ois Diederich*a
Received 13th March 2008, Accepted 18th April 2008
First published as an Advance Article on the web 2nd June 2008
DOI: 10.1039/b804375b
Enzymes of the non-mevalonate pathway for isoprenoid biosynthesis are therapeutic targets for the
treatment of important infectious diseases. Whereas this pathway is absent in humans, it is used by
plants, many eubacteria and apicomplexan protozoa, including major human pathogens such as
Plasmodium falciparum and Mycobacterium tuberculosis. Herein, we report on the design, preparation
and biological evaluation of a new series of ligands for IspE protein, a kinase from this pathway. These
inhibitors were developed for the inhibition of IspE from Escherichia coli, using structure-based design
approaches. Structure–activity relationships (SARs) and a co-crystal structure of Aquifex aeolicus IspE
bound to a representative inhibitor validate the proposed binding mode. The crystal structure shows
that the ligand binds in the substrate–rather than the adenosine 5ꢀ-triphosphate (ATP)-binding pocket.
As predicted, a cyclopropyl substituent occupies a small cavity not used by the substrate. The optimal
volume occupancy of this cavity is explored in detail. In the co-crystal structure, a diphosphate anion
binds to the Gly-rich loop, which normally accepts the triphosphate moiety of ATP. This structure
provides useful insights for future structure-based developments of inhibitors for the parasite enzymes.
targets by Jomaa et al., who showed that fosmidomycin [OHC–
Introduction
N(OH)–(CH2)3–PO3H2] cures malaria in rodents.6 The compound
The development of drugs with a novel mode of action for the
is a potent inhibitor of the second enzyme of the pathway, 1-deoxy-
treatment of malaria and tuberculosis is of urgent necessity in
D-xylulose 5-phosphate reductoisomerase (IspC, EC 1.1.1.267),
light of the rapid emergence of multi-drug-resistant parasites.1,2
For this purpose, inhibition of the enzymes of the non-mevalonate
pathway for the biosynthesis of the essential isoprenoid precur-
sor molecules, isopentenyl diphosphate (IPP) and dimethylallyl
diphosphate (DMAPP),3 has been recognised as a highly attractive
approach (Scheme 1). This biosynthetic pathway starts with
the condensation of pyruvate and glyceraldehyde 3-phosphate
and is exclusively used by a number of pathogens, such as the
tuberculosis-causing M. tuberculosis4 and the malaria-causing
P. falciparum.5,6 Both diseases still pose a major health concern
as they are responsible for an estimated 300–500 million new
infections and 3 million deaths annually.1,7 Importantly, the non-
mevalonate pathway is completely distinct to that used by humans,
making its component enzymes highly attractive drug targets. In
fact, the enzymes of the pathway have been validated as drug
and is currently being tested in clinical trials.8 In addition, other
series of phosphate and phosphonate-based inhibitors of IspC are
under investigation.9
The growing number of published X-ray crystal structures of
the constituent enzymes of the non-mevalonate pathway10 has
opened up the opportunity for lead generation by structure-
based design.11 In recent years, we applied this approach to
the development of the first families of ligands for two central
enzymes of the pathway, IspE (4-diphosphocytidyl-2C-methyl-
D-erythritol kinase, EC 2.7.1.148)12,13 and IspF (2C-methyl-D-
erythritol 2,4-cyclodiphosphate synthase, EC 4.6.1.12).14 The
kinase IspE requires ATP and Mg2+ cations to catalyse the
phosphorylation of the 2-OH group of 4-diphosphocytidyl-2C-
methyl-D-erythritol (CDP-ME), producing 4-diphosphocytidyl-
2C-methyl-D-erythritol-2-phosphate (CDP-ME2P) (Scheme 1).15
Using the structure-based design approach, we developed first-
generation inhibitors of IspE, which target the substrate–rather
than the ATP-binding pocket and feature competitive inhibition
constants (Ki) for the enzyme from E. coli down to the upper
nanomolar range.12 To extend this work further, a new series
of derivatives was designed to (i) improve their water solubility,
(ii) validate the proposed binding mode through SARs and (iii)
explore the optimal filling of a small, hydrophobic cavity that
is not occupied by the substrate.16,17 The synthesis and in vitro
evaluation of these new inhibitors are reported here. Gratifyingly,
enhancement of ligand solubility enabled co-crystallisation with
IspE from A. aeolicus to give definitive proof of the postulated
binding mode.
aLaboratorium fu¨r Organische Chemie, ETH Zu¨rich, HCI, CH-8093,
Zu¨rich, Switzerland. E-mail: diederich@org.chem.ethz.ch; Fax: +41 44
6321109
bDivision of Biological Chemistry and Drug Discovery, College of Life
Sciences, MSI/WTB Complex, University of Dundee, Dow Street, Dundee,
DD1 5EH, United Kingdom. E-mail: w.n.hunter@dundee.ac.uk; Fax: +44
138 232 2558
cTechnische Universita¨t Mu¨nchen, Lehrstuhl fu¨r Organische Chemie und
Biochemie, Center for Integrated Protein Research, Lichtenbergstraße 4,
D-85748, Garching, Germany. E-mail: felix.rohdich@ch.tum.de; Fax: +49
89 289 1336
† Electronic supplementary information (ESI) available: Synthesis of
precursors to the new inhibitors, biological data and evaluation. See DOI:
10.1039/b804375b
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2719–2730 | 2719
©

View Article Online
Scheme 1 The non-mevalonate pathway for the biosynthesis of the C5 precursors to isoprenoids, IPP and DMAPP.3 ADP = adenosine 5ꢀ-diphosphate,
DXS = 1-deoxy-D-xylulose 5-phosphate synthase (EC 2.2.1.7), IspD = 4-diphosphocytidyl-2C-methyl-D-erythritol synthase (EC 2.7.7.60), IspG =
2C-methyl-D-erythritol 2,4-cyclodiphosphate reductase (EC 1.17.4.3), IspH = 1-hydroxy-2-methyl-2-(E)-butenyl-4-diphosphate reductase (EC 1.17.1.2).
2V2Q, respectively).13 The active site of IspE can be divided into
three main binding pockets: the adenosine-binding, the cytidine-
binding and the ME/phosphate-binding pockets (Fig. S1).
Results and discussion
Inhibitor design
At the onset of this project, two X-ray crystal structures of IspE
We designed our first-generation inhibitors for E. coli IspE,
such as ( )-1–( )-4 (Fig. 1),12 using the modelling programme
MOLOC20 and the crystal structure of the ternary complex men-
tioned above.19 They target the cytidine-binding pocket as shown
in Fig. 2 for the cyclopropyl derivative ( )-1. This compound was
found to be a competitive inhibitor of the kinase with a Ki value
were available, one of the apo-enzyme of Thermus thermophilus
18
˚
(1.7 A resolution, RCSB Protein Data Bank (PDB) code: 1UEK)
and the other of the ternary complex of E. coli IspE bound to
CDP-ME and 5ꢀ-adenyl-b,c-amidotriphosphate (AppNp), a non-
˚
hydrolysable ATP analogue (2.0 A resolution, PDB code: 1OJ4,
Fig. S1).19 In the course of this project, two co-crystal structures
of 290
100 nM.12 The central cytosine scaffold was predicted
of A. aeolicus IspE and synthetic, cytidine-based ligands were
solved (at 2.3 A and 2.4 A resolution, PDB codes: 2V2V and
to sandwich between Tyr25 and Phe185 and form three hydrogen
bonds to both backbone and side chain of His26, in analogy to
˚
˚
Fig. 1 Inhibitors of IspE investigated in this study.
2720 | Org. Biomol. Chem., 2008, 6, 2719–2730 This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Validation and quantification of these postulated individual
enzyme–ligand interactions clearly required more extensive SARs
and X-ray analysis of co-crystal structures. Thus, compounds
( )-5–( )-11 (Fig. 1) were newly prepared to explore the optimal
filling of the hydrophobic pocket occupied by the cyclopropyl
group of bound ( )-1. Replacement of the tetrahydrothiophenyl
ring in 12 and 13 by alicyclic residues was expected to yield
information on the postulated S–aromatic interaction. The N-
methylated sulfonamide ( )-14 was designed to evaluate the
energetics of the predicted hydrogen bond to the side chain of
Asp141. Compounds ( )-15–( )-17 lack the sulfonamide group
entirely.
IspE is difficult to crystallise and does not even tolerate
traces of organic solvents in the crystallisation buffer. This
has prevented the analysis of co-crystal structures of the first-
generation ligands,12 even though these compounds are sufficiently
soluble in the presence of small amounts of organic co-solvents for
conducting in vitro assays. Small cytosine derivatives possess re-
markably low water-solubility, despite low partition/distribution
coefficients (e.g. compound ( )-1: clogP = clogD = 0.3; Table 1).
Therefore, compounds 18–22 were specifically prepared with the
aim to obtain improved aqueous solubility. Particularly successful
in this respect was the introduction of an oxetane derivative as
a ribose substitute.23 The proposed, modelling-supported binding
mode of the resulting ligand 22 is depicted in Fig. S2.
Syntheses of the inhibitors
The syntheses of the representative target molecules 22 and
( )-18 are shown in Schemes 2 and 3. The remaining ligands
were constructed by an analogous route or following the published
synthetic strategy (see the experimental section of the main text
for the synthesis and characterisation of the inhibitors, and the
ESI for the synthesis and characterisation of the precursors).12
5-Iodocytosine (23)24 was reacted with the Michael acceptor 2423
to afford the alkylated cytosine derivative 25. The choice of both
the right base and temperature turned out to be crucial for this
reaction. The alkyne cross-coupling partner 26 was obtained in a
reaction of propargyl amine with cyclopropanesulfonyl chloride.12
The cytosine derivative 25 and the alkyne 2612 smoothly underwent
Sonogashira cross-coupling, affording inhibitor 22 in good yield
(71%, Scheme 2).
For the synthesis of ligand ( )-18, morpholine was converted
into morpholine-4-sulfonyl chloride (27) in 84% yield, following
a literature procedure.25 This intermediate was reacted with
propargyl amine to obtain alkyne 28. Sonogashira cross-coupling
of 28 with cytosine derivative ( )-2912 afforded inhibitor ( )-18
in good yield (75%, Scheme 3).
Fig. 2 MOLOC-generated molecular model of inhibitor ( )-112 in the
active site of E. coli IspE (PDB code: 1OJ4).19 (a) Overview of the
cytidine-binding pocket. (b) Close-up of the small, hydrophobic pocket
showing the van der Waals surfaces of the receptor and the cyclopropyl ring
of the ligand. Colour code: protein skeleton: C: grey; inhibitor skeleton:
˚
C: green; O: red; N: blue; P: orange; S: yellow. Distances are given in A.
The units for the indicated distances and the colour code are maintained
throughout the article, if not otherwise stated.
the binding mode observed for the cytosine ring of CDP-ME. A
tetrahydrothiophenyl ring was selected as ribose substitute and
postulated to occupy the centre of a pseudo-p sandwich made
of Tyr25 and Pro182. According to the modelling, favourable
S–aromatic interactions21 with the phenolic ring of Tyr25 are
established in both diastereoisomeric complexes formed by
( )-1, which are expected to be of similar stability. A propargylic
sulfonamide vector was chosen to direct an alkyl group, such as
the cyclopropyl substituent in ( )-1, into a newly discovered, small
and lipophilic pocket lined by the side chains of Phe185, Leu15 and
Leu28. This pocket is not utilised by the substrate CDP-ME. The
sulfonamide moiety in its favourable staggered conformation, with
the N-lone pair bisecting the O–S–O angle,12,22 was expected to
form ionic hydrogen bonds to the side chains of Lys10 and Asp141.
Enzyme assay
The IC50 (IC50 = concentration of inhibitor at which 50%
maximum initial velocity is observed) and Ki values for all new
inhibitors were determined in a photometric assay, employing a
complex set of auxiliary enzymes and E. coli IspE (see Table 1 and
ESI).26 The inhibition was verified for selected ligands with a direct
NMR spectroscopic assay.26 Using the programme Dynafit, the
mode of inhibition was assigned (see Fig. S3 and S4 for selected
curves to determine IC50 and Ki values).27 While a competitive
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2719–2730 | 2721
©

View Article Online
Table 1 Inhibitory activities and calculated physicochemical properties of inhibitors for E. coli IspE
Ligand
Kic/lMa
Kiu/lMb
IC50/lM
Modec
clogPd
clogDd
(
(
(
(
(
(
(
(
(
(
(
12
13
(
(
(
)-112
)-212
)-312
)-412
)-5
0.29 0.1
2.6 0.1
0.64 0.1
8.2 1.7
8.0 0.1
2.0 0.3
1.8 0.3
0.52 0.1
8
22
6
0
2
0
Compet.
Compet.
Compet.
Mixed
Compet.
Compet.
Compet.
Compet.
Compet.
Mixed
0.3
−0.03
0.5
1.0
1.6
2.6
1.4
0.9
0.9
1.5
2.0
1.0
0.9
1.1
0.9
2.3
1.7
0.3
−0.06
0.5
1.0
1.5
2.6
1.4
0.8
0.9
1.4
2.0
0.9
0.9
1.1
0.9
2.2
1.2
27.3 11
48 20
51
47
8
1
1
0
0
1
1
1
0
6
1
)-6
)-7
)-8
8
)-9
0.56
0
17
55
290
39
44
13
1300 30
190 10
395 20
110
597 15
117
144
590 10
)-10
)-11
0.89 0.1
2.5 0.4
1.5 0.2
1.5 0.2
2.5 0.2
41.0 3.0
4.7 0.3
11.8 0.8
13.1 1.2
19.0 9.2
67.7 35
Mixed
Compet.
Compet.
Compet.
Compet.
Compet.
Compet.
Compet.
)-14
)-15
)-16
)-17
)-18
(
(
19
20
21
22
2
−1.0
−0.6
0.4
2.3
−0.1
−1.0
−4.0
0.4
1.6
−0.2
3
1
28.7 1.7
Compet.
b
a Kic = competitive inhibition constant. Kiu = uncompetitive inhibition constant. c Compet.: competitive inhibition; mixed: mixed competitive–
uncompetitive inhibition. d The logarithmic partition coefficients clogP (octanol/water) and clogD (octanol/water at pH 7.4) values were calculated
with the programmes ACD/LogP and ACD/LogD (ACD-Labs) with an uncertainty of 0.7 to 0.9.
Scheme 2 Synthesis of inhibitor 22. Reagents and conditions: (i) Cs2CO3,
Scheme 3 Synthesis of inhibitor ( )-18. Reagents and conditions: (i) Et3N,
DMF, 25 ◦C, 100 h, 33%; (ii) 25, Et3N, [PdCl2(PPh3)2], CuI, DMF, 25 ◦C,
[Pd(PPh3)4], CuI, DMF, 50 ◦C, 3.5 h, 75%.
16 h, 71%. DMF = N,N-dimethylformamide.
Ki value increases by nearly an order of magnitude upon changing
mechanism with respect to CDP-ME binding was attributed to
most ligands, a mixed competitive (Kic)–uncompetitive (Kiu) mode
of inhibition applied to some exceptions. The kinetics provide a
first indication that the ligands occupy the substrate-binding site,
as predicted.
The previously described cyclopropyl derivative ( )-1 remains
the best of the inhibitors of E. coli IspE reported to date.12
The three-membered ring seems to provide an optimal filling of
the lipophilic pocket lined by the side chains of Phe185, Leu15
and Leu28. The binding affinity of ligands with a smaller alkyl
substituent is reduced as the pocket is not properly filled. Thus, the
from ( )-1 to methyl-substituted ( )-2 and by a factor of two
upon changing from ( )-1 to ethyl-substituted ( )-3 (see Fig. S5
for its proposed binding mode). The affinity of isopropyl- (( )-8)
and cyclobutyl- (( )-9) substituted ligands is very similar to that of
ethyl-substituted ( )-3; in these cases, however, the alkyl residues
seem slightly too large for optimal volume occupancy.
Mecozzi and Rebek investigated the ideal volume occupancy
of apolar, confined spaces in capsular synthetic receptors. They
found that the most stable inclusion complex forms if 55 9%
of the cavity space is occupied by the guest.16 This so-called
“55% rule” holds for numerous complexation events in synthetic
2722 | Org. Biomol. Chem., 2008, 6, 2719–2730
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
supramolecular chemistry.28,29 Recently, it was applied to the filling
of a hydrophobic pocket in the active site of an enzyme, the
antimalarial target plasmepsin II.17 At smaller packing coefficients
(PCs), the van der Waals interactions in the cavity are less than
optimal, while at higher PCs, the mobility of the binding partners
becomes reduced, resulting in large entropic losses that counteract
enthalpic gains.
˚
˚
the side chain of Asp141 (d(OAsp · · · N) = 2.9 A and 3.1 A, see
Fig. S7).
Finally, efforts to obtain fully water soluble inhibitors for
X-ray crystallography were rewarded. Among the five com-
pounds specifically prepared for this purpose, two [( )-18 (Ki =
13.1 lM) and 22 (Ki = 28.7 lM)] did not require addition of Me2SO
to perform the in vitro binding assay. Although the former still
requires ethanol as a co-solvent, the latter is fully water soluble,
enabling X-ray crystallographic studies.
By filling the cavity with a hydrocarbon network using the
modelling package Spartan 06,16,17,30 the free space in the small,
3
˚
lipophilic pocket in E. coli IspE was determined to be about 100 A .
Binding affinity indeed nicely correlates with volume occupancy:
the cyclopropyl ring in ( )-1 shows a PC of 56% and inhibition
is most effective (Ki = 0.29 lM). Lower PCs (28% for methyl in
( )-2 and 45% for ethyl in ( )-3) or higher PCs (62% for isopropyl
in ( )-8 and 69% for cyclobutyl in ( )-9) correlate with weaker
inhibition.
X-Ray crystallography
Building on previous experience,13 A. aeolicus IspE was chosen
to obtain co-crystals with the water-soluble inhibitor 22. The
structure of A. aeolicus IspE in complex with 22 was determined to
˚
2.2 A resolution (PDB code: 2VF3, Fig. S8). The statistics indicate
When moving to slightly larger sulfonamide substituents, bind-
ing affinity decreases strongly. Thus, the Ki values of n-propyl-
(( )-4) and n-butyl- (( )-5) substituted ligands show a nearly
30-fold increase (Ki ≈ 8 lM) with respect to ( )-1. Modelling
suggests that the n-propyl residue might still be accommodated in
the cavity, although in an energetically less favourable, contorted
gauche conformation (PC 61%).31 Even when folded, the n-butyl
substituent no longer fits well into the pocket. Hence, a facile
conformational change of the propargylic sulfonamide linker
could equally well direct this substituent out of the pocket into
solvent-exposed space. Larger substituents, such as n-hexyl (in
( )-6), cyclopentyl (in ( )-10) or cyclohexyl (in ( )-11), most
certainly do not bind in the lipophilic pocket but rather reach
into the opposite direction, towards the solvent (see Fig. S6).
An increase in binding affinity measured for these ligands largely
results from their increased lipophilicity (see the clogP and clogD
values in Table 1), and the resulting more favourable partitioning
from the aqueous buffer into the less polar protein environment.
The tetrahydrothiophenyl ring remains the best ribose substitute
in the investigated family of ligands (for additional ribose replace-
ments, see ref. 12). Similar inhibitory activities are measured for
the cyclopentyl- and the cyclobutylmethyl-substituted ligands 12
and 13. Substitution of the tetrahydrothiophenyl ring in ( )-1 by
a cyclopentyl ring in 12 thus lowers the binding affinity by a factor
that an acceptable medium resolution model was produced
(Table S1). For example, Rwork and Rfree values are 22.7 and
27.6%, respectively. The structure is isomorphous with previously
determined complexes.13 Although adenosine 5ꢀ-monophosphate
(AMP) was present in the crystallisation conditions, no electron
density corresponding to this ligand could be observed.
A. aeolicus IspE displays the typical galactose, homoserine,
mevalonate and phosphomevalonate (GHMP) kinase fold that
comprises two domains.19 The N-terminal domain (residues 1–
155) consists of an elongated six-stranded b sheet, the concave
side of which is flanked by three a helices and two segments of
a 310 helix. The C-terminal domain comprises an anti-parallel
four-stranded b sheet, bordering the N-terminal-domain b sheet,
four a helices and two 310 helices. The overall distribution of
secondary-structure elements is well-conserved in this enzyme
family. Although the crystal structure presents two molecules in
the asymmetric unit, featuring “active sites A and B”, the enzyme
is a monomer as indicated by size exclusion gel chromatography
and analytical ultracentrifugation (data not shown). The enzyme
active site is a deep, polar cavity located between the N- and C-
terminal domains.
The co-crystal structure clearly shows that our proposed binding
mode is indeed adopted by the water-soluble inhibitor 22, i.e.
it binds at the cytidine-binding pocket (Fig. 3). As discussed
previously, the amino acid sequences and structures of the active
sites of A. aeolicus and E. coli IspE show minor differences.13
As a result of this, the nucleobase portion and the sulfonamide
moiety form slightly different hydrogen-bonding patterns to those
predicted by modelling for E. coli IspE (see Fig. S2). In particular,
the cytosine moiety of 22 is engaged in two additional hydrogen
bonds (Fig. 3b). In active site B, the cytosine moiety forms four
of five, corresponding to a free enthalpy increment of DDG300K
≈
1 kcal mol−1. The two rings are sandwiched between the side chain
of Tyr25 and the pyrrolidine ring of Pro182 (see Fig. 2). We explain
the increased affinity with favourable interactions between the S
atom of the sandwiched heterocyclic ring and the phenolic side
chain. The energetics of the proposed S–aromatic interaction are
in good agreement with literature values.21
=
Important contributions of the sulfonamide moiety to the
observed binding affinity were clearly confirmed. N-methylation
in ( )-14 (Ki = 2.5 lM) reduces the inhibitory potency nearly by a
factor of ten as compared to ( )-1. Therefore, the postulated ionic
hydrogen bond between the sulfonamide NH and the side chain of
Asp141 (Fig. 2) could contribute as much as DDG300K = 1.3 kcal
mol−1 to the overall binding free enthalpy. Complete substitution
of the sulfonamide moiety for heterocyclic amines in ( )-15–
( )-17 further reduces the inhibitory potency. The piperidine
derivative ( )-16 maintains the largest affinity (Ki = 4.7 lM). This
finding can be explained with the formation of a favourable ionic
hydrogen bond between the protonated piperidinium residue and
hydrogen bonds to the side chain and backbone NH and C O
of His25. An additional hydrogen bond is observed between the
=
NH2 group of the ligand and the backbone C O of Lys145. The
cytosine ring is positioned at the centre of a p sandwich made of
Tyr24 and Tyr175 (= Phe185 in E. coli IspE). The sulfonamide
functionality, at the other end of the inhibitor, is stabilised by
hydrogen bonds to the side chains of Asn11, Tyr31 and Asp130
(Fig. 3b).
The sulfonamide moiety of 22 is bound in a similar way in
active site A, whereas the nucleobase shows a modified hydrogen-
bonding pattern, which could arise from a different conformation
of the oxetanyl substituent. As a result, the hydrogen bonds
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2719–2730 | 2723
©

View Article Online
Fig. 3 X-Ray crystal structure of A. aeolicus IspE co-crystallised with 22 (PDB code: 2VF3). Shown is active site B (for a view of active site A, see Fig. S8).
(a) Overview of the binding mode of 22 in the cytidine-binding pocket. (b) View onto the cytosine and sulfonamide moieties and their hydrogen-bonding
interactions with the protein. (c) View onto the cyclopropyl residue approaching the OH group of Tyr175 in the small, lipophilic pocket. The distance
˚
between the cyclopropyl substituent and the OH group of Tyr175 is given in A.
between cytosine and the backbone amide NH of His25 cannot
be formed (Fig. S8b). A superposition of 22 in active sites A and
B is shown in Fig. S8c. Although the protein backbone is almost
perfectly superposed throughout the active sites, some flexibility
can be seen at the level of the side chains. Most importantly, the
oxetanyl substituent clearly adopts a different conformation to
that of 22 bound in active site B.
Electron density was observed in the Gly-rich loop of the
adenosine-binding pocket but this was not compatible with AMP
that was present in the crystallisation conditions. The density could
be modelled, and refined satisfactorily as diphosphate. Indeed, this
overlays well with the diphosphate moiety of AppNp in the high-
resolution crystal structure of the E. coli enzyme complex (data
not shown). The diphosphate forms similar hydrogen bonds to the
backbone amide NH groups of the Gly-rich loop, a typical motif
for the recognition of phosphate groups.32
well conserved (Fig. S9b). This result consolidates the choice
of targeting the inhibitors to the cytidine-binding pocket and
establishes A. aeolicus IspE as a suitable model system for future
rounds of design.
Of particular interest for our future efforts to design inhibitors
for IspE from important pathogens such as M. tuberculosis
or P. falciparum is the fact that the cyclopropyl ring of 22
is accommodated in the small, hydrophobic pocket lined by
Leu14, Ile27, Tyr175 and Leu208 (Fig. 3c). The residence of the
cyclopropyl ring was not necessarily to be expected since the
change from Phe185 (in E. coli IspE) to Tyr175 (in A. Aeolicus
IspE) makes the pocket much more hydrophilic. Indeed, the
cyclopropyl ring prevents any solvation of the phenolic OH group
of Tyr175 (Fig. 3c). Because this should be rather unfavourable,
it could well explain the reduced inhibitory activity of 22 against
A. aeolicus IspE. With the propargylic sulfonamide residue, our
ligands obviously provide an excellent vector for precisely and
favourably addressing this OH group in future studies. This
may be important since the enzymes from M. tuberculosis and
P. falciparum possess a conserved tyrosine residue here and
according to homology modelling (not shown), the phenolic OH
group is directed into the small pocket.
A superposition of the co-crystal structure of A. aeolicus IspE
with 22 and diphosphate (PDB code: 2VF3) and the previously
obtained co-crystal structure of a cytidine-based ligand (PDB
code: 2V2V)13 confirmed our hypothesis that the Gly-rich loop is
extremely well conserved and preorganised (Fig. S9a). In addition,
it can be observed that the cytidine-binding pocket is equally
2724 | Org. Biomol. Chem., 2008, 6, 2719–2730
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
to the published protocol.15b NADH and phosphoenolpyruvate
potassium salt were purchased from Biomol, ATP and pyruvate
kinase/lactate dehydrogenase from Sigma-Aldrich.
Conclusion
We described the design, synthesis, in vitro evaluation and X-ray
co-crystal structure analysis of a series of inhibitors of the kinase
IspE. These were prepared as a follow-up to the first-generation
inhibitors to validate a proposed binding mode. Through inter-
esting SARs, we managed to gain additional evidence for our
hypothesis. By designing water-soluble inhibitors and solving the
co-crystal structure of one ligand with A. aeolicus IspE, the
suggested binding mode was validated. The inhibitor occupies
the cytidine-binding pocket and exhibits a Ki value in the lower
micromolar activity range. This proof of concept opens the way
for further modification and optimisation of the inhibitors. The
ultimate goal of this research project is the development of ligands
with even better activity against IspE from medically important
species to make progress on the way to anti-infectives with a novel
mode of action.
Enzyme-coupled photometric assay for IC50 determination.
Assay mixtures were prepared as described with some
modifications:26 60 mm3 of a solution containing 100 mM Tris-
HCl (pH 8.0), 10 mM MgCl2, 2 mM dithiothreitol, 2.5 mM phos-
phoenolpyruvate potassium salt, 2 mM ATP, 0.46 mM NADH,
1 U of lactate dehydrogenase, 1 U of pyruvate kinase and IspE
protein were added to 60 mm3 of the inhibitor solutions (final
concentration varied from 8–1000 lM). The reaction was started
by addition of 60 mm3 of CDP-ME (final concentration 1 mM)
and monitored at 340 nm.
Enzyme-coupled photometric assay for Ki determination. Assay
mixtures were prepared as follows: 50 mm3 of a solution containing
100 mM Tris-HCl (pH 8.0), 10 mM MgCl2, 2 mM dithiothreitol,
2.5 mM phosphoenolpyruvate potassium salt, 2 mM ATP, 0.46 mM
NADH, 1 U of lactate dehydrogenase, 1 U of pyruvate kinase
and 2.4 mU IspE protein were added to 50 mm3 each of the
inhibitor solution. The reaction was started by addition of 50 mm3
CDP-ME. For the determination of each single KM value in the
presenceofthe inhibitor, the concentration of CDP-ME was varied
from 35–400 lM, while keeping the concentration of the inhibitor
fixed. The mixtures were incubated at 27 ◦C, and the reaction
was monitored photometrically at 340 nm. The Ki values of the
inhibitors were obtained by observing the behaviour of KM values
at different inhibitor concentrations (0–400 lM) and processing
the data using the programme Dynafit.27
Experimental
General details
Compounds 1–4,12 23,24 24,23 26,12 27,25 ( )-29,12 30,33 31,34 32–
3335 and 3412 were prepared according to literature procedures.
The synthesis and characterisation of the precursors 35–45 is
described in the ESI. Reagents and solvents were purchased
reagent grade and used without further purification. THF and
CH2Cl2 were freshly distilled from sodium benzophenone ketyl
and CaH2. All reactions were performed in oven-dried glassware
under an Ar atmosphere, unless otherwise stated. All products
were dried under high vacuum (10−2 Torr) before analytical
characterisation. TLC: Aluminium sheets coated with SiO2-60
UV254 from Macherey-Nagel, visualisation by UV light at 245 nm
and staining with a solution of KMnO4 (1.5 g), K2CO3 (10 g),
5% NaOH (2.5 cm3) in H2O (150 cm3) or a solution of ninhydrin
(0.3 g) in butanol (100 cm3) and glacial acetic acid (3 cm3). Column
chromatography (CC): SiO2-60 (230–400 mesh, 0.040–0.063 mm)
from Fluka. Uncorrected melting points (mp) were determined
in an open capillary using a Bu¨chi-510 apparatus. IR spectra
were recorded on a Perkin-Elmer BX FT-IR spectrophotometer
(ATR-unit, Attenuated Total Reflection, Golden Gate) as a film
or neat. NMR spectra were recorded at 25 ◦C on a Varian
Gemini-300 spectrometer, using the solvent peak as an internal
reference. Coupling constants (J) are given in Hz. The resonance
multiplicity is described as s (singlet), br s (broad singlet), d
(doublet), t (triplet), q (quartet), quint (quintet), sext (sextet) and m
(multiplet). High-resolution mass spectra (HR-MS) were recorded
on an IonSpec Ultima FT-ICR with 3-hydroxypicolinic acid (3-
HPA) as matrix (MALDI), Micromass AutoSpec-Ultima (EI)
and Finnigan TSQ 7000 (electronspray ionisation). The detected
masses are given as m/z with M+ representing the molecular
ion. Elemental analyses were performed by the Mikrolabor at the
Laboratorium fu¨r Organische Chemie, ETH Zu¨rich. The nomen-
clature was generated with the computer programme ACD/Name
(ACD/Labs).
13C NMR assay. Assay mixtures contained 100 mM Tris-HCl
(pH 8.0), 10 mM MgCl2, 5 mM ATP, 10% (v/v) D2O, 2 mM
dithiothreitol, 1.5 mM [1,3,4–13C3]-CDP-ME, and 13 lg IspE
protein in a volume of 500 mm3. Inhibitory substances were added
at final concentrations ranging from 0–3 mM. The mixtures were
incubated at 37 ◦C and terminated by the addition of EDTA
to a final concentration of 20 mM. The solution was analysed
by 13C NMR spectroscopy on a Bruker DRX 500 (125 MHz)
spectrometer and referenced to an internal standard of [1-13C1]-
glucose (0.9 mM).
General procedures
General procedure A for the Sonogashira cross-coupling of
5-iodocytosine derivatives. To an Ar-degassed Schlenk flask
charged with the iodocytosine (1.0 eq.), the acetylene (1.4–2.3 eq.)
and Et3N (3.0–7.0 eq.) in dry DMF, [PdCl2(PPh3)2] or [Pd(PPh3)4]
(0.10 eq.) and CuI (0.20 eq.) were added at 25 ◦C. The mixture
was left to stir at 25–50 ◦C in the dark, the solvent evaporated
in vacuo, and the residue purified by CC.
If purification by CC yielded the triethylammonium salt, the
solid was dissolved in CH2Cl2–i-PrOH (3 : 1), washed with
saturated aqueous NaCl solution (3×), dried over Na2SO4, filtered
and concentrated in vacuo.
Preparation of the inhibitors
Enzyme assays
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}butane-1-sulfonamide (( )-5). Gen-
eral procedure A, starting from ( )-2912 (100 mg, 0.30 mmol), 3033
Materials. [1,3,4-13C3]-CDP-ME was prepared according to a
literature procedure.26 E. coli IspE was also obtained according
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2719–2730 | 2725
©

View Article Online
(110 mg, 0.60 mmol), Et3N (0.12 cm3, 0.90 mmol), [PdCl2(PPh3)2]
(21 mg, 0.030 mmol) and CuI (11 mg, 0.060 mmol) in dry DMF
(8.0 cm3). The mixture was left to stir at 25 ◦C for 2.5 h. Purification
by CC (SiO2; CH2Cl2–MeOH 99 : 1→94 : 6) afforded ( )-5 (80 mg,
72%) as an off-white solid (Found: C 48.7, H 6.1, N 14.8. Calcd
for C15H22N4O3S2: C 48.6, H 6.0, N 15.1%); mp 149–151 ◦C;
mmax(neat)/cm−1 3194, 2959, 2870, 1636, 1496, 1403, 1301, 1228,
1136, 1073, 972, 918, 833, 780, 726, 668; dH(300 MHz, (CD3)2SO)
0.84 (t, J = 7.4, 3 H), 1.35 (sext, J = 7.4, 2 H), 1.60–1.71 (m,
2 H), 1.91–2.07 (m, 3 H), 2.15–2.25 (m, 1 H), 2.81–2.89 (m, 1 H),
3.11 (t, J = 8.0, 2 H), 3.18–3.26 (m, 1 H), 4.02 (d, J = 5.7, 2 H),
6.13–6.16 (m, 1 H), 6.94 (br s, 1 H), 7.61 (t, J = 5.7, 1 H), 7.63 (br s,
1 H), 8.12 (s, 1 H); dC(75 MHz, (CD3)2SO) 13.5, 20.9, 25.2, 28.9,
32.7, 32.8, 36.9, 51.1, 64.1, 75.2, 89.1, 91.3, 145.0, 153.5, 163.8;
MALDI-HR-MS: calcd for C15H23N4O3S2+ ([M + H]+): 371.1206;
found: 371.1207.
(80 mg, 0.50 mmol), Et3N (0.30 cm3, 2.1 mmol), [PdCl2(PPh3)2]
(21 mg, 0.030 mmol) and CuI (11 mg, 0.060 mmol) in dry
DMF (8.0 cm3). The mixture was left to stir at 25 ◦C for 3 h.
Purification by CC (SiO2; CH2Cl2–MeOH 99 : 1→94 : 6) afforded
( )-8 (99 mg, 92%) as an off-white solid; mp 197–200 ◦C;
mmax(neat)/cm−1 3387, 3068, 2974, 1668, 1636, 1506, 1489, 1457,
1439, 1403, 1312, 1232, 1136, 1074, 1057, 973, 889, 778, 692;
dH(300 MHz, (CD3)2SO) 1.25 (d, J = 6.9, 6 H), 1.92–2.07 (m,
3 H), 2.16–2.23 (m, 1 H), 2.81–2.89 (m, 1 H), 3.19–3.32 (m, 2 H),
4.03 (d, J = 5.7, 2 H), 6.12–6.16 (m, 1 H), 6.86 (br s, 1 H), 7.55 (t,
J = 5.7, 1 H), 7.83 (br s, 1 H), 8.11 (s, 1 H); dC(75 MHz, (CD3)2SO)
16.2 (2 C), 28.9, 32.8 (2 C), 36.8, 51.5, 64.2, 75.1, 89.1, 91.3, 144.9,
+
153.5, 163.8; MALDI-HR-MS: calcd for C14H21N4O3S2 ([M +
H]+): 357.1050; found: 357.1049.
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}cyclobutanesulfonamide
(( )-9).
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}hexane-1-sulfonamide (( )-6). Gen-
eral procedure A, starting from ( )-2912 (100 mg, 0.30 mmol), 35
(100 mg, 0.60 mmol), Et3N (0.12 cm3, 0.90 mmol), [PdCl2(PPh3)2]
(21 mg, 0.030 mmol) and CuI (11 mg, 0.060 mmol) in dry DMF
(8.0 cm3). The mixture was left to stir at 25 ◦C for 3 h. Purification
by CC (SiO2; CH2Cl2–MeOH 99 : 1→94 : 6) afforded ( )-6 (49 mg,
41%) as an off-white solid (Found: C 51.4, H 6.5, N 13.8. Calcd
for C17H26N4O3S2: C 51.2, H 6.6, N 14.1%); mp 179–182 ◦C;
mmax(neat)/cm−1 3325, 3061, 2929, 2858, 1640, 1496, 1402, 1304,
1228, 1134, 1076, 921, 858, 783; dH(300 MHz, (CD3)2SO) 0.78 (t,
J = 6.8, 3 H), 1.16–1.36 (m, 6 H), 1.61–1.71 (m, 2 H), 1.91–2.06
(m, 3 H), 2.16–2.25 (m, 1 H), 2.82–2.89 (m, 1 H), 3.11 (t, J = 8.0,
2 H), 3.19–3.26 (m, 1 H), 4.02 (d, J = 5.7, 2 H), 6.13–6.17 (m,
1 H), 6.93 (br s, 1 H), 7.56 (t, J = 5.7, 1 H), 7.84 (br s, 1 H), 8.13
(s, 1 H); dC(75 MHz, (CD3)2SO) 13.8, 21.9, 23.2, 27.3, 28.9, 30.8,
32.7, 32.8, 36.9, 51.3, 64.1, 75.2, 89.1, 91.3, 145.0, 153.4, 163.8;
MALDI-HR-MS: calcd for C17H27N4O3S2+ ([M + H]+): 399.1519;
found: 399.1514.
General procedure A, starting from ( )-2912 (84 mg, 0.26 mmol),
38 (70 mg, 0.40 mmol), Et3N (0.11 cm3, 0.80 mmol), [PdCl2(PPh3)2]
(19 mg, 0.026 mmol) and CuI (10 mg, 0.050 mmol) in dry DMF
(7.0 cm3). The mixture was left to stir at 25 ◦C for 2.5 h.
Purification by CC (SiO2; CH2Cl2–MeOH 99 : 1→94 : 6) afforded
( )-9 (51 mg, 54%) as an off-white solid (Found: C 48.6, H
5.5, N 14.9. Calcd for C15H20N4O3S2: C 48.9, H 5.5, N 15.2%);
mp 184–186 ◦C; mmax(neat)/cm−1 3382, 3299, 3208, 3066, 2945,
2857, 2707, 1640, 1601, 1504, 1438, 1404, 1343, 1298, 1275, 1227,
1182, 1135, 1081, 1010, 971, 912, 881, 854, 780, 722, 675, 616;
dH(300 MHz, (CD3)2SO) 1.83–2.09 (m, 5 H), 2.14–2.37 (m, 5 H),
2.81–2.89 (m, 1 H), 3.19–3.27 (m, 1 H), 4.00 (d, J = 5.7, 2 H),
4.02 (quint, J = 8.1, 1 H), 6.13–6.16 (m, 2 H), 6.88 (br s, 1 H),
7.54 (t, J = 5.7, 1 H), 7.83 (br s, 1 H), 8.13 (s, 1 H); dC(75 MHz,
(CD3)2SO) 16.2, 23.1 (2 C), 28.9, 32.7, 32.8, 36.8, 53.0, 64.2,
75.0, 89.1, 91.5, 145.0, 153.4, 163.7; MALDI-HR-MS: calcd for
C15H21N4O3S2 ([M + H]+): 369.1050; found: 369.1048.
+
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}cyclopentanesulfonamide (( )-10).
General procedure A, starting from ( )-2912 (84 mg, 0.26 mmol),
39 (73 mg, 0.40 mmol), Et3N (0.11 cm3, 0.80 mmol), [PdCl2(PPh3)2]
(19 mg, 0.026 mmol) and CuI (10 mg, 0.050 mmol) in dry DMF
(7.0 cm3). The mixture was left to stir at 25 ◦C for 2.5 h. Purification
by CC (SiO2; CH2Cl2–MeOH 99 : 1→94 : 6) afforded ( )-10
(75 mg, 75%) as an off-white solid; mp 200–203 ◦C; mmax(neat)/cm−1
3382, 3290, 3207, 3063, 2948, 2869, 2707, 1640, 1603, 1504, 1472,
1404, 1297, 1250, 1227, 1182, 1121, 1081, 971, 916, 881, 850, 779,
756, 722, 677; dH(300 MHz, (CD3)2SO) 1.48–1.73 (m, 4 H), 1.80–
2.08 (m, 7 H), 2.14–2.25 (m, 1 H), 2.81–2.89 (m, 1 H), 3.18–3.26
(m, 1 H), 3.68 (quint, J = 7.7, 1 H), 4.04 (d, J = 5.7, 2 H), 6.12–6.16
(m, 1 H), 6.88 (br s, 1 H), 7.56 (t, J = 5.7, 1 H), 7.83 (br s, 1 H),
8.12 (s, 1 H); dC(75 MHz, (CD3)2SO) 25.5 (2 C), 27.4 (2 C), 28.9,
32.7 (2 C), 36.9, 60.0, 64.2, 75.1, 89.1, 91.4, 145.0, 153.4, 163.7;
MALDI-HR-MS: calcd for C16H23N4O3S2+ ([M + H]+): 383.1206;
found: 383.1198.
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}butane-2-sulfonamide (( )-7). Gen-
eral procedure A, starting from ( )-2912 (120 mg, 0.37 mmol),
( )-36 (94 mg, 0.53 mmol), Et3N (0.15 cm3, 1.1 mmol),
[PdCl2(PPh3)2] (26 mg, 0.037 mmol) and CuI (14 mg, 0.074 mmol)
◦
in dry DMF (7.0 cm3). The mixture was left to stir at 25 C for
2.5 h. Purification by CC (SiO2; CH2Cl2–MeOH 99 : 1→94 : 6)
afforded ( )-7 (99 mg, 72%) as an off-white solid (Found: C 48.7,
H 6.1, N 14.8. Calcd for C15H22N4O3S2: C 48.6, H 6.0, N 15.1%);
mp 184–186 ◦C; mmax(neat)/cm−1 3382, 3297, 3207, 3073, 2938,
2861, 1640, 1495, 1405, 1300, 1234, 1122, 1088, 969, 847, 775,
721, 660; dH(300 MHz, (CD3)2SO) 0.93 (t, J = 7.5, 3 H), 1.24 (d,
J = 6.9, 3 H), 1.33–1.49 (m, 1 H), 1.88–2.08 (m, 4 H), 2.14–2.25
(m, 1 H), 2.81–2.89 (m, 1 H), 3.03–3.16 (m, 1 H), 3.18–3.26 (m,
1 H), 4.03 (d, J = 5.4, 2 H), 6.12–6.16 (m, 1 H), 6.88 (br s, 1 H),
7.57 (t, J = 5.4, 1 H), 7.83 (br s, 1 H), 8.11 (s, 1 H); dC(75 MHz,
(CD3)2SO) 10.9, 12.9, 22.9, 28.9, 32.8 (2 C), 36.9, 57.1, 64.2,
75.1, 89.1, 91.4, 144.9, 153.4, 163.8; MALDI-HR-MS: calcd for
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}cyclohexanesulfonamide (( )-11).
General procedure A, starting from ( )-2912 (100 mg, 0.30 mmol),
40 (120 mg, 0.60 mmol), Et3N (0.30 cm3, 2.1 mmol), PdCl2(PPh3)2]
(21 mg, 0.030 mmol) and CuI (11 mg, 0.060 mmol) in dry DMF
(8.0 cm3). The mixture was left to stir at 25 ◦C for 3 h.
C15H23N4O3S2 ([M + H]+): 371.1206; found: 371.1206.
+
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}propane-2-sulfonamide (( )-8). Gen-
eral procedure A, starting from ( )-2912 (97 mg, 0.30 mmol), 37
2726 | Org. Biomol. Chem., 2008, 6, 2719–2730
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Purification by CC (SiO2; CH2Cl2–MeOH 99 : 1→94 : 6) afforded
( )-11 (78 mg, 66%) as an off-white solid; mp 203–205 ◦C;
mmax(neat)/cm−1 3389, 3300, 3212, 3065, 2935, 2854, 2705, 1644,
1605, 1505, 1451, 1403, 1309, 1271, 1248, 1226, 1179, 1136, 1081,
972, 921, 878, 857, 779, 705; dH(300 MHz, (CD3)2SO) 1.08–1.40
(m, 5 H), 1.58–1.61 (m, 1 H), 1.75–1.79 (m, 2 H), 1.91–2.23 (m,
6 H), 2.82–2.89 (m, 1 H), 3.04–3.14 (m, 1 H), 3.18–3.25 (m,
1 H), 4.01 (d, J = 5.7, 2 H), 6.13–6.17 (m, 1 H), 6.93 (br s, 1 H),
7.56 (t, J = 5.7, 1 H), 7.83 (br s, 1 H), 8.13 (s, 1 H); dC(75 MHz,
(CD3)2SO) 24.5 (2 C), 24.8, 25.9 (2 C), 28.8, 32.8 (2 C), 36.9, 59.1,
64.2, 75.1, 89.1, 91.4, 145.0, 153.5, 163.8; MALDI-HR-MS: calcd
1106, 1066, 983, 916, 886, 753, 691; dH(300 MHz, CD3OD) 1.02–
1.06 (m, 2 H), 1.08–1.14 (m, 2 H), 1.90–2.15 (m, 3 H), 2.28–2.33
(m, 1 H), 2.58–2.63 (m, 1 H), 2.91–2.96 (m, 1 H), 3.00 (s, 3 H),
3.21–3.27 (m, 1 H), 4.32 (s, 2 H), 6.21–6.24 (m, 1 H), 8.30 (s, 1 H);
dC(75 MHz, CD3OD) 5.5 (2 C), 27.9, 29.8, 34.2, 35.4, 39.2, 41.4,
67.1, 77.4, 90.8, 91.8, 147.2, 156.9, 166.0; MALDI-HR-MS: calcd
for C15H21N4O3S2 ([M + H]+): 369.1050; found: 369.1050.
+
( )-4-Amino-5-(3-morpholin-4-ylprop-1-yn-1-yl)-1-(tetrahydro-
2-thienyl)pyrimidin-2(1H)-one (( )-15). General procedure A,
starting from ( )-2912 (68 mg, 0.21 mmol), 3134 (53 mg,
0.42 mmol), Et3N (0.90 mm3, 0.63 mmol), [PdCl2(PPh3)2] (15 mg,
0.033 mmol) and CuI (8.0 mg, 0.06◦6 mmol) in dry DMF (4.8 cm3).
The mixture was left to stir at 25 C for 2 h. Purification by CC
(SiO2; CH2Cl2–MeOH 95 : 5) afforded ( )-15 (64 mg, 95%) as
an off-white solid; mp 194–196 ◦C; mmax(neat)/cm−1 3466, 2919,
2853, 2323, 1979, 1656, 1643, 1561, 1555, 1510, 1502, 1477, 1401,
1323, 1300, 1229, 1108, 1072, 1004, 976, 919, 889, 859, 774, 728,
700, 668, 632; dH(300 MHz, CDCl3) 1.81–1.94 (m, 1 H), 2.02–2.18
(m, 2 H), 2.28–2.40 (m, 1 H), 2.60 (t, J = 4.4, 4 H), 2.91–2.99
(m, 1 H), 3.15–3.23 (m, 1 H), 3.49 (s, 2 H), 3.76 (t, J = 4.4, 4
H), 5.70 (br s, 1 H), 6.31–6.34 (m, 1 H), 7.56 (br s, 1 H), 8.18 (s,
1 H); dC(75 MHz, (CD3)2SO) 29.1, 32.9, 36.7, 47.5, 52.0 (2 C), 64.1,
66.1 (2 C), 76.5, 89.9, 90.8, 145.3, 153.8, 164.1; MALDI-HR-MS:
calcd for C15H21N4O2S+ ([M + H]+): 321.1380; found: 321.1384.
+
for C17H25N4O3S2 ([M + H]+): 397.1363; found: 397.1364.
N-[3-(4-Amino-1-cyclopentyl-2-oxo-1,2-dihydropyrimidin-5-yl)-
prop-2-yn-1-yl]cyclopropanesulfonamide (12). General procedure
A, starting from 41 (100 mg, 0.33 mmol), 2612 (100 mg, 0.66 mmol),
Et3N (0.14 cm3, 0.98 mmol), [PdCl2(PPh3)2] (23 mg, 0.033 mmol)
and CuI (13 mg, 0.066 mmol) in dry DMF (7.6 cm3). The mixture
was left to stir at 25 ◦C for 2.5 h. Purification by CC (SiO2;
CH2Cl2–MeOH 96 : 4) afforded 12 (100 mg, 90%) as an off-white
solid (Found C 53.4, H 5.9, N 16.6. Calcd for C15H20N4O3S:
C 53.6, H 6.0, N 16.7%); mp > 197 ◦C (decomposition);
mmax(neat)/cm−1 3382, 3055, 1669, 1653, 1635, 1617, 1506, 1495,
1456, 1403, 1352, 1238, 1084, 1039, 988, 972, 930, 914, 885, 832,
809, 779, 752, 734, 697, 668, 659; dH(300 MHz, (CD3)2SO) 0.94
(d, J = 6.3, 4 H), 1.53–1.91 (m, 8 H), 2.63 (quint, J = 6.3, 1 H),
4.04 (d, J = 5.7, 2 H), 4.72 (quint, J = 7.8, 1 H), 6.69 (br s, 1 H),
7.56 (t, J = 5.7, 1 H), 7.64 (br s, 1 H), 7.89 (s, 1 H); dC(75 MHz,
(CD3)2SO) 5.0 (2 C), 23.6 (2 C), 29.5, 30.7 (2 C), 33.0, 57.7,
75.6, 88.8, 91.4, 146.0, 154.1, 164.0; MALDI-HR-MS: calcd for
C15H21N4O3S+ ([M + H]+): 337.1329; found: 337.1328.
( )-4-Amino-5-(piperidin-1-ylprop-1-yn-1-yl)-1-(tetrahydro-
2-thienyl)pyrimidin-2(1H)-one (( )-16). General procedure A,
starting from ( )-2912 (68 mg, 0.21 mmol), 3235 (52 mg,
0.42 mmol), Et3N (0.90 mm3, 0.63 mmol), [PdCl2(PPh3)2] (15 mg,
0.021 mmol) and CuI (8.0 mg, 0.04◦2 mmol) in dry DMF (4.8 cm3).
The mixture was left to stir at 25 C for 2 h. Purification by CC
(SiO2; CH2Cl2–MeOH 91 : 9) afforded ( )-16 (29 mg, 43%) as
a brown oil; mmax(thin film)/cm−1 3441, 3288, 3081, 2934, 2854,
1653, 1634, 1558, 1539, 1521, 1505, 1496, 1399, 1363, 1308, 1263,
1229, 1149, 1125, 1036, 993, 939, 889, 803, 780, 732, 697, 653;
dH(300 MHz, CDCl3) 1.45–1.47 (m, 2 H), 1.63–1.69 (m, 4 H),
1.80–1.93 (m, 1 H), 2.00–2.18 (m, 2 H), 2.30–2.42 (m, 1 H), 2.55
(m, 4 H), 2.91–2.99 (m, 1 H), 3.15–3.22 (m, 1 H), 3.47 (s, 2 H),
5.86 (br s, 1 H), 6.31–6.34 (m, 1 H), 7.28 (br s, 1 H), 8.16 (s, 1 H);
dC(75 MHz, CDCl3) 23.7, 25.7 (2 C), 28.3, 33.4, 38.9, 48.5, 53.7,
66.0 (2 C), 75.7, 90.9, 91.6, 145.3, 155.1, 164.4; MALDI-HR-MS:
calcd for C16H23N4OS+ ([M + H]+): 319.1587; found: 319.1586.
N-{3-[4-Amino-1-(cyclobutylmethyl)-2-oxo-1,2-dihydropyrimi-
din-5-yl]prop-2-yn-1-yl}cyclopropanesulfonamide (13). General
procedure A, starting from 42 (100 mg, 0.33 mmol), 2612 (100 mg,
0.66 mmol), Et3N (0.14 cm3, 0.98 mmol), [PdCl2(PPh3)2] (23 mg,
0.033 mmol) and CuI (13 mg, 0.066 mmol) in dry DMF (7.6 cm3).
The mixture was left to stir at 25 ◦C for 2.5 h. Purification by CC
(SiO2; CH2Cl2–MeOH 96 : 4) afforded 13 (82 mg, 74%) as an off-
white solid (Found C 53.8, H 6.0, N 16.7. Calcd for C15H20N4O3S:
◦
C 53.6, H 6.0, N 16.7%); mp 152–153 C; mmax(neat)/cm−1 3438,
3056, 2963, 2854, 2323, 1662, 1652, 1622, 1505, 1456, 1444, 1394,
1367, 1326, 1306, 1233, 1192, 1150, 1037, 933, 889, 825, 760, 694,
683, 659; dH(300 MHz, (CD3)2SO) 0.95 (d, J = 6.3, 4 H), 1.65–
1.91 (m, 6 H), 2.58–2.94 (m, 2 H), 3.69 (d, J = 7.4, 2 H), 4.03 (d,
J = 5.6, 2 H), 6.69 (br s, 1 H), 7.57 (t, J = 5.6, 1 H), 7.61 (br s,
1 H), 7.97 (s, 1 H); dC(75 MHz, (CD3)2SO) 5.0 (2 C), 17.7, 24.8
(2 C), 29.6, 33.0, 34.0, 53.4, 75.4, 88.0, 91.5, 149.1, 154.2, 164.6;
MALDI-HR-MS: calcd for C15H21N4O3S+ ([M + H]+): 337.1329;
found: 337.1330.
( )-4-Amino-5-(pyrrolidin-1-ylprop-1-yn-1-yl)-1-(tetrahydro-2-
thienyl)pyrimidin-2(1H)-one (( )-17). General procedure A,
starting from ( )-2912 (68 mg, 0.21 mmol), 3335 (46 mg,
0.42 mmol), Et3N (0.90 mm3, 0.63 mmol), [PdCl2(PPh3)2] (15 mg,
0.021 mmol) and CuI (8.0 mg, 0.04◦2 mmol) in dry DMF (4.8 cm3).
The mixture was left to stir at 25 C for 2 h. Purification by CC
(SiO2; CH2Cl2–MeOH 91 : 9) afforded ( )-17 (34 mg, 53%) as
a brown oil; mmax(thin film)/cm−1 3289, 3143, 1662, 1646, 1653,
1635, 1506, 1457, 1404, 1363, 1317, 1229, 1127, 1034, 1008, 903,
804, 783, 668, 649; dH(300 MHz, CDCl3): 1.83–2.03 (m, 5 H),
2.04–2.16 (m, 2 H), 2.28–2.38 (m, 1 H), 2.66 (m, 4 H), 2.91–2.99
(m, 1 H), 3.15–3.23 (m, 1 H), 3.61 (s, 2 H), 5.76 (br s, 1 H),
6.31–6.35 (m, 1 H), 7.16 (br s, 1 H), 8.16 (s, 1 H); dC(75 MHz,
(CD3)2SO) 23.1 (2 C), 28.9, 32.7, 36.6, 43.2, 52.1 (2 C), 64.0,
76.3, 89.5, 90.2, 145.3, 153.6, 163.9; MALDI-HR-MS: calcd for
C15H21N4OS+ ([M + H]+): 305.1431; found: 305.1432.
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}-N-methylcyclopropanesulfonamide
(( )-14). General procedure A, starting from ( )-2912 (120 mg,
0.36 mmol), 43 (87 mg, 0.50 mmol), Et3N (0.15 cm3, 1.1 mmol),
[Pd(PPh3)4] (42 mg, 0.036 mmol) and CuI (15 mg, 0.072 mmol)
◦
in dry DMF (7.0 cm3). The mixture was left to stir at 50 C for
26 h. Purification by CC (SiO2; CH2Cl2–MeOH–Et3N 95 : 5 : 1)
afforded ( )-14 (41 mg, 31%) as an off-white solid; mp 83–85 ◦C;
mmax(neat)/cm−1 3403, 3209, 1644, 1417, 1325, 1272, 1245, 1150,
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2719–2730 | 2727
©

View Article Online
off-white solid; mp 198–200 ◦C; mmax(neat)/cm−1 3403, 3301, 3006,
2563, 1725, 1639, 1603, 1506, 1437, 1389, 1355, 1338, 1318, 1273,
1245, 1198, 1186, 1152, 1132, 1108, 1067, 1020, 986, 934, 889, 846,
799, 781, 752, 690, 668; dH(300 MHz, (CD3)2SO) 3.83 (s, 3 H), 4.10
(d, J = 5.1, 2 H), 4.49 (q, J = 9.9, 2 H), 4.95 (s, 2 H), 6.86 (br s,
1 H), 7.38 (d, J = 8.1, 2 H), 7.78 (br s, 1 H), 7.92 (d, J = 8.1,
2 H), 8.19 (s, 1 H), 8.33 (t, J = 5.1, 1 H); dC(75 MHz, (CD3)2SO)
32.6, 51.4, 52.0, 52.8 (q, J = 30.3), 75.3, 88.7, 90.6, 122.2 (q, J =
276.4), 127.5 (2 C), 128.5, 129.2 (2 C), 142.7, 149.3, 153.9, 164.6,
165.8; MALDI-HR-MS: calcd for C18H18F3N4O5S+ ([M + H]+):
459.0945; found: 459.0947.
( )-N-{3-[4-Amino-2-oxo-1-(tetrahydro-2-thienyl)-1,2-dihydro-
pyrimidin-5-yl]prop-2-yn-1-yl}morpholine-4-sulfonamide (( )-18).
General procedure A, starting from ( )-2912 (120 mg, 0.37 mmol),
28 (110 mg, 0.54 mmol), Et3N (0.15 cm3, 1.1 mmol), [Pd(PPh3)4]
(43 mg, 0.037 mmol) and CuI (14 mg, 0.054 mmol) in dry
◦
DMF (7.0 cm3). The mixture was left to stir at 50 C for 3.5 h.
Purification by CC (SiO2; CH2Cl2–MeOH 92 : 8) afforded
( )-18 (110 mg, 75%) as an off-white solid; mp 197–199 ◦C;
mmax(neat)/cm−1 3327, 2846, 2551, 2401, 2149, 1639, 1489, 1402,
1348, 1322, 1261, 1229, 1175, 1137, 1111, 1068, 1023, 993, 924,
837, 779, 728, 688, 627; dH(300 MHz, (CD3)2SO) 1.92–2.05 (m,
3 H), 2.15–2.22 (m, 1 H), 2.80–2.88 (m, 1 H), 3.06 (t, J = 4.5, 4
H), 3.18–3.22 (m, 1 H), 3.30 (s, 1 H), 3.62 (t, J = 4.5, 4 H), 3.98
(d, J = 3.6, 2 H), 6.12–6.15 (m, 1 H), 6.83 (br s, 1 H), 7.84 (t, J =
3.6, 1 H), 8.12 (s, 1 H); dC(75 MHz, (CD3)2SO) 28.9, 32.8, 33.2,
37.0, 45.7 (2 C), 64.4, 65.5 (2 C), 75.3, 89.3, 91.7, 145.2, 153.8,
Ethyl {3-[4-amino-5-{3-[(cyclopropylsulfonyl)amino]prop-1-yn-
1-yl}-2-oxopyrimidin-1(2H)-yl]oxetan-3-yl}acetate (22). Gen-
eral procedure A, starting from 25 (46 mg, 0.12 mmol), 2612 (44 mg,
0.28 mmol), Et3N (50 mm3, 0.36 mmol), [PdCl2(PPh3)2] (7.0 mg,
0.01 mmol) and CuI (4 mg, 0.02 mmol) in dry DMF (3 cm3). The
mixture was left to stir at 25 ◦C for 16 h. Purification by CC (SiO2;
CH2Cl2–MeOH 95 : 5) afforded 22 (35 mg, 71%) as an off-white
solid; mp 146–148 ◦C; mmax(neat)/cm−1 3103, 2930, 2388, 2284,
1712, 1645, 1542, 1495, 1410, 1372, 1326, 1302, 1245, 1203, 1148,
1086, 1069, 1021, 980, 906, 887, 836, 785, 732, 708; dH(300 MHz,
CDCl3) 0.98–1.04 (m, 2 H), 1.17–1.26 (m, 5 H), 2.51–2.60 (m, 1
H), 3.40 (s, 2 H), 4.08 (q, J = 7.2, 2 H), 4.12 (d, J = 4.2, 2 H),
4.62 (d, J = 7.5, 2 H), 4.88 (d, J = 7.5, 2 H), 6.54 (br s, 1 H),
6.68 (br s, 1 H), 6.99 (br s, 1 H), 7.56 (s, 1 H); dC(75 MHz, CDCl3)
5.2 (2 C), 13.7, 30.0, 32.9 (2 C), 38.5, 60.6, 60.9, 74.1, 78.4 (2
C), 91.8, 146.2, 153.4, 164.5, 170.1; MALDI-HR-MS: calcd for
C17H23N4O6S+ ([M + H]+): 411.1338; found: 411.1333.
164.0; MALDI-HR-MS: calcd for C15H22N5O4S2 ([M + H]+):
+
400.1108; found: 400.1113.
Triethylammonium [4-amino-5-{3-[(cyclopropylsulfonyl)amino]-
prop-1-yn-1-yl}-2-oxo-1,2-dihydropyrimidin-1(2H)-yl]acetate (19).
A solution of 20 (70 mg, 0.20 mmol) in aq. Et3N solution (1 M,
9.0 cm3) was heated to reflux for 1 h. The resulting mixture was
concentrated in vacuo, co-evaporated with ethanol (3 × 10 cm3)
and dried to afford 19 (60 mg, 90%) as a white solid; mp > 220 ◦C
(decomposition); mmax(neat)/cm−1 3286, 3202, 3060, 2449, 1728,
1660, 1609, 1495, 1427, 1393, 1309, 1241, 1187, 1142, 1050, 949,
892, 834, 799, 784, 760, 716, 624; dH(300 MHz, (CD3)2SO) 0.95 (d,
J = 5.7, 4 H), 2.59–2.63 (m, 1 H), 4.04 (d, J = 5.7, 2 H), 4.35 (s,
2 H), 6.81 (br s, 1 H), 7.59 (t, J = 5.7, 1 H), 7.73 (br s, 1 H), 7.92
(s, 1 H); dC(75 MHz, (CD3)2SO) 5.1 (2C), 29.6, 33.0, 50.1, 75.1,
88.5, 91.6, 149.8, 154.2, 165.0, 169.8; MALDI-HR-MS: calcd for
C12H15N4O5S+ ([M + H]+): 327.0758; found: 327.0763.
Ethyl
[3-(4-amino-5-iodo-2-oxopyrimidin-1(2H)-yl)oxetan-3-
yl]acetate (25). A suspension of 2324 (190 mg, 0.80 mmol) and
Cs2CO3 (144 mg, 0.44 mmol) in dry DMF (9.0 cm3) was left to
stir at 25 C for 1 h. Compound 2423 (0.042 cm3, 0.40 mmol)
◦
Ethyl [4-amino-5-{3-[(cyclopropylsulfonyl)amino]prop-1-yn-1-
yl}-2-oxo-1,2-dihydropyrimidin-1(2H)-yl]acetate (20). General
procedure A, starting from 44 (323 mg, 1.0 mmol), 2612 (318 mg,
2.0 mmol), Et3N (0.42 cm3, 3.0 mmol), [Pd(PPh3)4] (120 mg,
0.10 mmol) and CuI (38 mg, 0.20 mmol) in dry DMF (20 cm3).
The mixture was left to stir at 50 ◦C for 3.5 h. Purification by
CC (SiO2; CH2Cl2–MeOH ◦90 : 10) afforded 20 (267 mg, 75%) as
a white solid; mp 221–223 C; mmax(neat)/cm−1 3421, 3053, 2875,
2233, 1750, 1668, 1644, 1508, 1489, 1440, 1422, 1397, 1382, 1356,
1327, 1308, 1205, 1166, 1144, 1061, 1035, 936, 888, 838, 785, 739,
701, 675; dH(300 MHz, (CD3)2SO) 0.95 (d, J = 6.3, 4 H), 1.18 (t,
J = 7.1, 3 H), 2.62 (quint, J = 6.3, 1 H), 4.04 (s, 2 H), 4.11 (q,
J = 7.1, 2 H), 4.46 (s, 2 H), 6.88 (br s, 1 H), 7.58 (br s, 1 H), 7.80
(br s, 1 H), 7.96 (s, 1 H); dC(75 MHz, (CD3)2SO) 5.1 (2 C), 14.1,
29.6, 32.9, 49.9, 61.0, 74.8, 88.7, 91.7, 149.3, 153.8, 164.8, 168.1;
MALDI-HR-MS: calcd for C14H19N4O5S+ ([M + H]+): 355.1071;
found: 355.1067.
in dry DMF (2.0 cm3) was added, and the mixture was left to
stir at 25 ◦C for 100 h. The resulting mixture was concentrated
in vacuo. Purification by CC (SiO2; CH2Cl2–MeOH 95 : 5)
afforded 25 (50 mg, 33%) as a yellow solid (Found C 35.2, H 4.05,
N 10.8. Calcd for C11H14IN3O4: C 34.9, H 3.75, N 11.1%); mp >
175 ◦C (decomposition); mmax(neat)/cm−1 3403, 3212, 1712, 1644,
1418, 1353, 1318, 1272, 1244, 1151, 1132, 1151, 1105, 1083, 1066,
982, 917, 886, 863, 752, 691, 614; dH(300 MHz, CD3OD) 1.21 (t,
J = 7.1, 3 H), 3.34 (s, 2 H), 4.12 (q, J = 7.1, 2 H), 4.60 (d, J = 7.8,
2 H), 4.92 (d, J = 7.8, 2 H), 7.81 (s, 1 H); dC(75 MHz, (CD3)2SO)
14.0, 56.0, 60.2, 60.4, 62.7, 77.7 (2 C), 148.9, 153.1, 164.0, 169.5;
MALDI-HR-MS: calcd for C11H15IN3O4 ([M + H]+): 380.0102;
+
found: 380.0098.
N-Prop-2-yn-1-ylmorpholine-4-sulfonamide (28). To a solu-
tion of propargyl amine (0.10 cm3, 1.5 mmol) and Et3N (0.23 cm3,
1.7 mmol) in dry CH2Cl2 (10 cm3), 2725 (280 mg, 1.5 mmol) was
slowly added at 0 ◦C. The mixture was left to stir at 40 ◦C for 18 h
and concentrated in vacuo. The residue was taken up in CH2Cl2
(50 cm3) and washed with saturated aqueous NaCl solution (3 ×
30 cm3). The organic phase was dried over Na2SO4, filtered and
concentrated in vacuo and afforded 28 (140 mg, 46%) as a yellow
solid; mp 70–72 ◦C; mmax(neat)/cm−1 3276, 3257, 2963, 2929, 2857,
1622, 1458, 1430, 1314, 1304, 1261, 1213, 1140, 1125, 1110, 1088,
1075, 1018, 995, 936, 924, 840, 722, 711, 659, 628; dH(300 MHz,
CDCl3) 2.35 (t, J = 2.4, 1 H), 3.25 (t, J = 4.2, 4 H), 3.76 (t, J = 4.2,
Methyl 4-{[4-amino-2-oxo-5-(3-{[(2,2,2-trifluoroethyl)sulfonyl]-
amino}prop-1-yn-1-yl)pyrimidin-1(2H)-yl]methyl}benzoate (21).
General procedure A, starting from 45 (310 mg, 0.81 mmol), 2,2,2-
trifluoro-N-prop-2-yn-1-ylethanesulfonamide (34)12 (260 mg,
1.3 mmol), Et3N (0.34 cm3, 2.5 mmol), [PdCl2(PPh3)2] (58 mg,
0.081 mmol) and CuI (32 mg, 0.16 mmol) in dry DMF (17 cm3).
The mixture was left to stir at 25 ◦C for 20 h. Purification by CC
(SiO2; CH2Cl2–MeOH 90 : 10) afforded 21 (270 mg, 73%) as an
2728 | Org. Biomol. Chem., 2008, 6, 2719–2730
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
4 H), 3.87 (dd, J = 2.4, 6.3, 2 H), 4.71 (br s, 1 H); dC(75 MHz,
CDCl3) 33.0, 45.9 (2 C), 66.0 (2 C), 72.7, 79.4; electrospray-
ionisation-HR-MS: calcd for C7H13N2O3S+ ([M + H]+): 205.0641;
found: 205.0658.
thank Prof. Erick M. Carreira and Georg Wuitschik, ETH Zu¨rich,
for kindly providing compound 24.
References
X-Ray crystal structure of the complex with 22
1 (a) U. Weiss, Nature, 2002, 415, 669; (b) J. K. Baird, New
Engl. J. Med., 2005, 352, 1565–1577; (c) http://www.cdc.gov/malaria/
impact/index.htm; (d) M. Schlitzer, ChemMedChem, 2007, 2, 944–986.
2 (a) A. L. Kritski, L. S. Rodrigues de Jesus, M. K. Andrade, E. Werneck-
Barroso, M. A. Vieira, A. Haffner and L. W. Riley, Chest, 1997, 111,
1162–1167; (b) J. E. Olle´-Goig, Trop. Med. Int. Health, 2006, 11, 1625–
1628.
3 (a) M. Rohmer, M. Knani, P. Simonin, B. Sutter and H. Sahm,
Biochem. J., 1993, 295, 517–524; (b) M. K. Schwarz, ETH Zurich, Ph.
D. dissertation, No. 10951, 1994; (c) S. T. J. Broers, PhD dissertation,
No. 10978, ETH Zurich, 1994; (d) W. Eisenreich, B. Menhard, P. J.
Hylands, M. H. Zenk and A. Bacher, Proc. Natl. Acad. Sci. U. S. A.,
1996, 93, 6431–6436; (e) W. Eisenreich, M. Schwarz, A. Cartayrade, D.
Arigoni, M. H. Zenk and A. Bacher, Chem. Biol., 1998, 5, R221–R233.
4 Y. Boucher and W. F. Doolittle, Mol. Microbiol., 2000, 37, 703–716.
5 (a) F. Rohdich, J. Wungsintaweekul, M. Fellermeier, S. Sagner, S. Herz,
K. Kis, W. Eisenreich, A. Bacher and M. H. Zenk, Proc. Natl. Acad. Sci.
U. S. A., 1999, 96, 11758–11763; (b) F. Rohdich, S. Hecht, A. Bacher
and W. Eisenreich, Pure Appl. Chem., 2003, 75, 393–405.
Sample preparation. A plasmid encoding A. aeolicus IspE with
an N-terminal histidine tag with a protease cleavage site was used
to produce recombinant enzyme. Purification by affinity and ion
exchange chromatography gave a yield of 15 mg dm−3 of enzyme.36
Crystallisation, data collection and processing. A. aeolicus
IspE at 30 mg dm−3 in 10 mM Tris-HCl, 20 mM NaCl, 1 mM
dithriothreitol, pH 8.5, was incubated with AMP and 22 each at
4 mM for 1 h on ice, filtered using a 0.1 lm spin filter (Milli-
pore) then crystallised using hanging drop vapour diffusion at
18 ◦C. Crystals grew in 2 days in 2 mm3 drops consisting of 1 mm3
protein complex and 1 mm3 reservoir solution (1.6 M (NH4)2SO4,
0.1 M Na cacodylate, pH 6.5, 0.1 M NaBr). The pyramidal crystals
typically grew to dimensions of 0.4 × 0.4 × 0.4 mm. Crystals
were cryo-protected in artificial mother liquor containing 20%
glycerol and flash cooled for data collection on beam-line BM14
at the European Synchrotron Radiation Facility, Grenoble, using
6 H. Jomaa, J. Wiesner, S. Sanderbrand, B. Altincicek, C. Weidemeyer,
M. Hintz, I. Tu¨rbachova, M. Eberl, J. Zeidler, H. K. Lichtenthaler, D.
Soldati and E. Beck, Science, 1999, 285, 1573–1576.
7 http://www.cdc.gov/tb/WorldTBDay/resources_global.htm.
8 (a) For some selected publications on fosmidomycin and derivatives: T.
Kuzuyama, T. Shimizu, S. Takahashi and H. Seto, Tetrahedron Lett.,
1998, 39, 7913–7916; (b) M. A. Missinou, S. Borrmann, A. Schindler, S.
Issifou, A. A. Adegnika, P.-B. Matsiegui, R. Binder, B. Lell, J. Wiesner,
T. Baranek, J. Jomaa and P. G. Kremsner, Lancet, 2002, 360, 1941–1942;
(c) J. Wiesner, R. Ortmann, H. Jomaa and M. Schlitzer, Angew. Chem.,
Int. Ed., 2003, 42, 5274–5293; (d) B. Lell, R. Ruangweerayut, J. Wiesner,
M. A. Missinou, A. Schindler, T. Baranek, M. Hintz, D. Hutchinson,
H. Jomaa and P. G. Kremsner, Antimicrob. Agents Chemother., 2003,
47, 735–738; (e) S. Oyakhirome, S. Issifou, P. Pongratz, F. Barondi, M.
Ramharter, J. F. Kun, M. A. Missinou, B. Lell and P. G. Kremsner,
Antimicrob. Agents Chemother., 2007, 51, 1869–1871; (f) N. Singh, G.
Cheve, M. A. Avery and C. R. McCurdy, Curr. Pharm. Des., 2007, 13,
1161–1177.
˚
a MAR CCD 225 detector. Data to 2.2 A were processed and
scaled using MOSFLM37 and SCALA,38 respectively (Table S1).
The crystals are cubic in space group P213 with two molecules in
˚
the asymmetric unit and a unit cell length a = 137.3 A and are
isomorphous with the previously solved A. aeolicus IspE structure
(to be published elsewhere).
Structure determination. The CCP4 suite of programmes was
used for the analysis.39 Rigid body and rounds of restrained
refinement,40 interspersed with model building in COOT,41 al-
lowed two molecules of 22, nine halide ions and 156 water
molecules to be included in the model. The head group of 22
was well-defined in the electron density and given full occupancy,
but parts of the tail appeared to be more flexible and occupancy
of 0.75 was assigned to this region of the inhibitor. AMP was in
the crystallisation conditions but there was no electron density
corresponding to adenine. A diphosphate and water molecules
could be modelled in the ATP binding site. The use of NaCl
and NaBr in the crystallisation conditions resulted in six Cl− and
three Br− ions being assigned. The complex was refined to an
Rfactor and Rfree of 22.7% and 27.6%, respectively. Model geometry
was assessed using PROCHECK,42 and all residues were found
in the most favourable or allowed regions of the Ramachandran
plot. Crystallographic statistics are presented in Table S1, and
coordinates and structure factors have been deposited in the PDB
under accession code 2VF3.
9 (a) Y.-H. Woo, R. P. M. Fernandes and P. J. Proteau, Bioorg. Med.
Chem., 2006, 14, 2375–2385; (b) V. Devreux, J. Wiesner, J. L. Goeman,
J. Van der Eycken, H. Jomaa and S. Van Calenbergh, J. Med. Chem.,
2006, 49, 2656–2660; (c) T. Haemers, J. Wiesner, S. Van Poecke, J.
Goeman, D. Henschker, E. Beck, H. Jomaa and S. Van Calenbergh,
Bioorg. Med. Chem. Lett., 2006, 16, 1888–1891.
10 W. N. Hunter, J. Biol. Chem., 2007, 282, 21573–21577.
11 F. Hof and F. Diederich, Chem. Commun., 2004, 477–480.
12 A. K. H. Hirsch, S. Lauw, P. Gersbach, W. B. Schweizer, F. Rohdich,
W. Eisenreich, A. Bacher and F. Diederich, ChemMedChem, 2007, 2,
806–810.
13 C. M. Crane, A. K. H. Hirsch, M. S. Alphey, T. Sgraja, S. Lauw, V.
Illarionova, F. Rohdich, W. Eisenreich, W. N. Hunter, A. Bacher and F.
Diederich, ChemMedChem, 2008, 3, 91–101.
14 (a) C. M. Crane, J. Kaiser, N. L. Ramsden, S. Lauw, F. Rohdich,
W. Eisenreich, W. N. Hunter, A. Bacher and F. Diederich, Angew.
Chem., Int. Ed., 2006, 45, 1069–1074; (b) C. Baumgartner, C. Eberle,
F. Diederich, S. Lauw, F. Rohdich, W. Eisenreich and A. Bacher, Helv.
Chim. Acta, 2007, 90, 1043–1068.
15 (a) F. Rohdich, J. Wungsintaweekul, H. Lu¨ttgen, M. Fischer, W.
Eisenreich, C. A. Schuhr, M. Fellermeier, N. Schramek, M. H. Zenk
and A. Bacher, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 8251–8256;
(b) H. Lu¨ttgen, F. Rohdich, S. Herz, J. Wungsintaweekul, S. Hecht,
C. A. Schuhr, M. Fellermeier, S. Sagner, M. H. Zenk, A. Bacher and W.
Eisenreich, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 1062–1067; (c) T.
Kuzuyama, M. Takagi, K. Kaneda, H. Watanabe, T. Dairi and H. Seto,
Tetrahedron Lett., 2000, 41, 2925–2928.
Acknowledgements
This work was supported by the following agencies: ETH Research
council (to A.K.H.H. and F.D.), F. Hoffmann-La Roche Ltd,
Basel, Hans-Fischer Gesellschaft (to S.L., A.B., F.R. and W.E.),
The Wellcome Trust (to M.S.A and W.N.H.), Biotechnology
and Biological Sciences Research Council (UK) (to M.S.A and
W.N.H.) and the European Synchrotron Radiation Facility. We
16 S. Mecozzi and J. Rebek Jr., Chem.–Eur. J., 1998, 4, 1016–1022.
17 M. Zu¨rcher, T. Gottschalk, S. Meyer, D. Bur and F. Diederich,
ChemMedChem, 2008, 3, 237–240.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2719–2730 | 2729
©

View Article Online
18 T. Wada, T. Kuzuyama, S. Satoh, S. Kuramitsu, S. Yokoyama, S. Unzai,
J. R. H. Tame and S.-Y. Park, J. Biol. Chem., 2003, 278, 30022–30027.
19 L. Miallau, M. S. Alphey, L. E. Kemp, G. A. Leonard, S. M.
McSweeney, S. Hecht, A. Bacher, W. Eisenreich, F. Rohdich and W. N.
Hunter, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 9173–9178.
20 (a) P. R. Gerber and K. Mu¨ller, J. Comput. Aided Mol. Des., 1995, 9,
251–268; (b) Gerber Molecular Design (http://www.moloc.ch).
21 E. A. Meyer, R. K. Castellano and F. Diederich, Angew. Chem., Int.
Ed., 2003, 42, 1210–1250.
22 V. Petrov, V. Petrova, G. V. Girichev, H. Oberhammer, N. I. Giricheva
and S. Ivanov, J. Org. Chem., 2006, 71, 2952–2956.
23 G. Wuitschik, M. Rogers-Evans, K. Mu¨ller, H. Fischer, B. Wagner, F.
Schuler, L. Polonchuk and E. Carreira, Angew. Chem., Int. Ed., 2006,
45, 7736–7739.
24 K. A. Watanabe, T.-L. Su, R. S. Klein, C. K. Chu, A. Matsuda, M. W.
Chun, C. Lopez and J. J. Fox, J. Med. Chem., 1983, 26, 152–156.
25 J. U. Jeong, Int. Pat., WO 2005/013909, 2005.
26 V. Illarionova, J. Kaiser, E. Ostrozhenkova, A. Bacher, M. Fischer, W.
Eisenreich and F. Rohdich, J. Org. Chem., 2006, 71, 8824–8834.
27 P. Kuzmic, Anal. Biochem., 1996, 237, 260–273.
28 (a) O. Hayashida, L. Sebo and J. Rebek Jr., J. Org. Chem., 2002, 67,
8291–8298; (b) A. Scarso, L. Trembleau and J. Rebek Jr., Angew. Chem.,
Int. Ed., 2003, 42, 5499–5502; (c) L. Trembleau and J. Rebek Jr., Science,
2003, 301, 1219–1220; (d) A. Scarso, L. Trembleau and J. Rebek Jr.,
J. Am. Chem. Soc., 2004, 126, 13512–13518; (e) B. W. Purse and J.
Rebek Jr., Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 2530–2534; (f) R. J.
Hooley and J. Rebek Jr., Org. Lett., 2007, 9, 1179–1182.
Tetrahedron, 2006, 62, 2008–2015; (b) T. Gottschalk, B. Jaun and F.
Diederich, Angew. Chem., Int. Ed., 2007, 46, 260–264.
30 Spartan ‘06, Wavefunction Inc., Irvine, CA, USA, 2006.
31 For the binding of linear alkanes in confined spaces, see: J. Rebek Jr.,
Chem. Commun., 2007, 2777–2789.
32 A. K. H. Hirsch, F. R. Fischer and F. Diederich, Angew. Chem., Int.
Ed., 2007, 46, 338–352.
33 J. Deng, J. Tabei, M. Shiotsuki, F. Sanda and T. Masuda, Macro-
molecules, 2004, 37, 5538–5543.
34 P. Malherbe and R. Masciadri, Int. Pat., WO2005/094828, 2005.
35 (a) A. Viola, J. J. Collins, N. Filipp and J. S. Locke, J. Org. Chem., 1993,
58, 5067–5075; (b) J. H. Biel and F. DiPierro, J. Am. Chem. Soc., 1958,
80, 4609–4614.
36 T. Sgraja, M. S. Alphey, S. Ghilagaber, R. Marquez, M. N. Robertson,
J. L. Hemmings, S. Lauw, F. Rohdich, A. Bacher, W. Eisenre-
ich, V. Illarionova and W. N. Hunter, FEBS J., 2008, 275, 2779–
2794.
37 A. G. W. Leslie, Joint CCP4 and ESF-EAMCB Newsletter on Protein
Crystallography, 1992, No. 26.
38 P. Evans, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2006, 62, 72–
82.
39 Collaborative Computational Project, Number 4: Acta Crystallogr.,
Sect. D: Biol. Crystallogr., 1994, 50, 760–763.
40 G. N. Murshudov, A. A. Vagin and E. J. Dodson, Acta Crystallogr.,
Sect. D: Biol. Crystallogr., 1997, 53, 240–255.
41 P. Emsley and K. Cowtan, Acta Crystallogr., Sect. D: Biol. Crystallogr.,
2004, 60, 2126–2132.
29 (a) L. Pirondini, D. Bonifazi, B. Cantadori, P. Braiuca, M. Cam-
pagnoloa, R. De Zorzi, S. Geremia, F. Diederich and E. Dalcanale,
42 R. A. Laskowski, D. S. Moss and J. M. Thornton, J. Mol. Biol., 1993,
231, 1049–1067.
2730 | Org. Biomol. Chem., 2008, 6, 2719–2730
This journal is
The Royal Society of Chemistry 2008
©