Isoprenoid Biosynthesis Inhibitors Targeting Bacterial Cell Growth

Article abstract of DOI:10.1002/cmdc.201600343

We synthesized potential inhibitors of farnesyl diphosphate synthase (FPPS), undecaprenyl diphosphate synthase (UPPS), or undecaprenyl diphosphate phosphatase (UPPP), and tested them in bacterial cell growth and enzyme inhibition assays. The most active compounds were found to be bisphosphonates with electron-withdrawing aryl-alkyl side chains which inhibited the growth of Gram-negative bacteria (Acinetobacter baumannii, Klebsiella pneumoniae, Escherichia coli, and Pseudomonas aeruginosa) at ～1–4 μg mL^?1levels. They were found to be potent inhibitors of FPPS; cell growth was partially “rescued” by the addition of farnesol or overexpression of FPPS, and there was synergistic activity with known isoprenoid biosynthesis pathway inhibitors. Lipophilic hydroxyalkyl phosphonic acids inhibited UPPS and UPPP at micromolar levels; they were active (～2–6 μg mL^?1) against Gram-positive but not Gram-negative organisms, and again exhibited synergistic activity with cell wall biosynthesis inhibitors, but only indifferent effects with other inhibitors. The results are of interest because they describe novel inhibitors of FPPS, UPPS, and UPPP with cell growth inhibitory activities as low as ～1–2 μg mL^?1.

Full text of DOI:10.1002/cmdc.201600343

DOI: 10.1002/cmdc.201600343
Full Papers
Isoprenoid Biosynthesis Inhibitors Targeting Bacterial Cell
Growth
Janish Desai+,[a] Yang Wang+,[b] Ke Wang,[b] Satish R. Malwal,[b] and Eric Oldfield*[a, b]
We synthesized potential inhibitors of farnesyl diphosphate
synthase (FPPS), undecaprenyl diphosphate synthase (UPPS), or
undecaprenyl diphosphate phosphatase (UPPP), and tested
them in bacterial cell growth and enzyme inhibition assays.
The most active compounds were found to be bisphospho-
nates with electron-withdrawing aryl-alkyl side chains which in-
hibited the growth of Gram-negative bacteria (Acinetobacter
baumannii, Klebsiella pneumoniae, Escherichia coli, and Pseudo-
monas aeruginosa) at ~1–4 mgmLꢀ1 levels. They were found to
be potent inhibitors of FPPS; cell growth was partially “res-
cued” by the addition of farnesol or overexpression of FPPS,
and there was synergistic activity with known isoprenoid bio-
synthesis pathway inhibitors. Lipophilic hydroxyalkyl phos-
phonic acids inhibited UPPS and UPPP at micromolar levels;
they were active (~2–6 mgmLꢀ1) against Gram-positive but not
Gram-negative organisms, and again exhibited synergistic ac-
tivity with cell wall biosynthesis inhibitors, but only indifferent
effects with other inhibitors. The results are of interest because
they describe novel inhibitors of FPPS, UPPS, and UPPP with
cell growth inhibitory activities as low as ~1–2 mgmLꢀ1.
Introduction
The emergence of resistance to antibiotics is a public health
threat,[1] so new drugs and new drug leads that act on new tar-
gets are of interest.[2] Many of the antibiotics which, over time,
have been of particular importance inhibit enzymes involved
in bacterial cell wall biosynthesis; therefore, this is a potentially
important area for drug discovery. Bacterial cell wall biosynthe-
sis involves many enzymes. Initially, dimethylallyl diphosphate
(DMAPP, 1) and isopentenyl diphosphate (IPP, 2) are produced
by the 2-C-methyl-d-erythritol-4-phosphate or mevalonate
pathways (Figure 1). In most bacteria, DMAPP then reacts se-
quentially with two IPP molecules to form trans-farnesyldiphos-
phate (FPP, 3) in a reaction catalyzed by FPP synthase (FPPS).
FPP then acts as the substrate for undecaprenyl diphosphate
synthase (UPPS) to form the C55 isoprenoid, undecaprenyl di-
phosphate (UPP, 4). UPP is hydrolyzed by UPP phosphatase
(UPPP) to form undecaprenyl phosphate (UP, 5), which is then
used (in most bacteria) for peptidoglycan, lipopolysaccharide,
and wall teichoic acid biosynthesis. FPPS, UPPS, and UPPP are
thus of interest as targets for novel anti-infective agents. In
mycobacteria, cell wall biosynthesis is different; it begins with
the formation of cis-FPP, not trans-FPP, followed by formation
of cis-decaprenyl (not undecaprenyl) diphosphate, and is not
discussed further here.
Our research group previously reported the lipophilic bi-
sphosphonate 6 to have modest activity against E. coli (IC50 ~
30 mm)[3] and to exhibit potent synergy (fractional inhibitory
concentration index, FICI=0.18) with fosmidomycin (7), which
inhibits the MEP pathway (Figure 2).[4] We also reported several
benzoic acids, as well as diamidines such as 8, to inhibit UPPS,
and compound 8 was found to act synergistically with methi-
cillin against methicillin-resistant Staphylococcus aureus
(MRSA).[5] Compound 8 was also active in vivo in a mouse
model of infection. We also found that the fertility drug clomi-
phene (9) is a UPPS inhibitor,[6] as well as an uncoupler of oxi-
dative phosphorylation;[7] the two effects presumably contrib-
ute to cell growth inhibition. In this study, we synthesized and
tested a series of bisphosphonates, phosphonates, and carbox-
ylic acids and tested them for prenyl synthase activity and for
activity against both Gram-positive and Gram-negative bacte-
ria.
Results and Discussion
[a] J. Desai,+ Prof. E. Oldfield
Farnesyl diphosphate synthase as an antibacterial target
Center for Biophysics and Quantitative Biology,
University of Illinois at Urbana–Champaign,
1110 West Green Street, Urbana, IL 61801 (USA)
E-mail: eo@chad.scs.uiuc.edu
FPPS is the target for bisphosphonate bone resorption drugs
such as zoledronate (10), though 10 has essentially no antibac-
terial activity. However, as noted above, in earlier work[3] we
found that the lipophilic bisphosphonate 6 (an FPPS inhibitor
first made by Widler et al.[8] as a potential bone resorption
drug) had weak but measurable (~30 mm) activity against the
Gram-negative bacterium E. coli. We also found activity against
FPPS from the trypanosomatid parasite Trypanosoma brucei
(TbFPPS) and we solved the X-ray crystallographic structure of
[b] Dr. Y. Wang,+ Dr. K. Wang, Dr. S. R. Malwal, Prof. E. Oldfield
Department of Chemistry,
University of Illinois at Urbana–Champaign,
600 South Mathews Avenue, Urbana, IL 61801 (USA)
[+] These authors contributed equally to this work.
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cmdc.201600343.
ChemMedChem 2016, 11, 1 – 12
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 1. Illustration of selected molecules involved in cell wall biosynthesis in most bacteria. Also shown are sites of action of some antibiotics and potential
targets, discussed in the text.
Figure 2. Structures of substrates and inhibitors discussed in the text.
a
TbFPPS·6 complex.[9] Interestingly, the crystal structure
cussed more below), it appeared that the aromatic group in 6
(or its analogues) might interact with the electron-rich Tyr in
the bacterial YS motif. We thus hypothesized that it might be
possible to obtain improved activity by incorporating electron-
withdrawing substituents on the phenyl group in 6 or its ana-
logues, leading to a tyrosine–inhibitor charge-transfer interac-
tion. In contrast, the addition of an electron-donating substitu-
ent might decrease activity, basically as we suggested in earlier
work on bisphosphonate inhibitors of ATP-mediated HIV-1 re-
verse transcriptase catalyzed excision of chain-terminating 3’-
azido-3’-deoxythymidine.[11] We therefore made three ana-
logues of 6, compounds 11–13, and tested them against vari-
ous FPPSs as well as in bacterial cell growth inhibition assays.
As can be seen in Figure 2, in addition to adding substitu-
ents (difluoro, bromo or methoxy) to the phenyl group, we re-
placed the amine with an imidazolium group, as in other
work[12] we found such species to be potent FPPS inhibitors
showed the phenyl group in 6 to be close (~4 ꢁ) to a tyrosine
residue (Y99), which in human FPPS (HsFPPS) is phenylalanine
(F99). This Phe residue in HsFPPS is thought to be involved in
limiting chain elongation to C15 and is present in FPPS isoforms
from other eukaryotes such as that from Schistosoma manso-
ni.[10] A partial alignment of TbFPPS, HsFPPS, SmFPPS, as well
as Gram-positive and Gram-negative bacterial FPPSs (S. aureus,
Bacillus subtilis, E. coli, Acinetobacter baumannii, Klebsiella pneu-
moniae, and Pseudomonas aeruginosa) is shown in Figure 3.
The bacterial FPPSs are different from those of eukaryotes in
that there are two extra amino acids in the first aspartate-rich
motif (FARM, which is involved in catalysis), plus, there is a con-
served YS sequence: both motifs are highlighted in yellow in
Figure 3. In our earlier work we found that longer- or shorter-
chain analogues of 6 had less activity than 6 against FPPS and
E. coli cell growth.[3] Also, based on structure alignments (dis-
&
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 3. ClustalW alignments of T. brucei, human, S. mansoni, and six bacterial FPPSs. The YS pair is present in all of the bacterial FPPSs, and the Tyr residue is
proposed here to interact with electron-deficient aryl groups in bisphosphonate inhibitors. FARM=first aspartate-rich motif; SARM=second aspartate-rich
motif; wavy lines represent residues omitted from the alignment, for clarity.
with good in vivo activity. Moreover, we replaced the 1-OH
group with 1-H because the hydroxy group is involved (in
other bisphosphonates) in potent bone binding.[13] These ana-
logues also appeared to be more readily synthesized than the
corresponding amines. We then tested all three compounds
against FPPSs from E. coli, P. aeruginosa, T. brucei, H. sapiens,
and S. mansoni, as well as against human geranylgeranyl di-
phosphate synthase (HsGGPPS). The results are shown in
Table 1 and Figure S1. As can be seen in Table 1, two of the
three bisphosphonates had activity in the ~70–600 nm range
against EcFPPS, PaFPPS, TbFPPS, HsFPPS, SmFPPS, and
HsGGPPS, generally similar to that observed (under the same
assay conditions) with zoledronate, although zoledronate is
only a weak HsGGPPS inhibitor,[12] whereas 11–13 are all active,
consistent with previous work on other lipophilic bisphospho-
nates.
analogue 13. This behavior was originally predicted, so the dif-
ferences between the EcFPPS/PaFPPS and other FPPS inhibi-
tion results with compounds 11–13 may be due to interactions
with a Tyr residue (Y79), in addition to small differences be-
tween the type II (EcFPPS; eubacterial) and type I (eukaryotic)
FPPS structures. At present, we do not have the structures of
11, 12, or 13 bound to a bacterial FPPS. However, Figure 4
However, with the two bacterial enzymes EcFPPS and
PaFPPS, there is essentially no FPPS inhibition by the methoxy
Table 1. IC50 values of 11, 12, and 13 against FPPS (in mm), GGPPS (in
mm), bacteria (mgmLꢀ1), and human cells in (mgmLꢀ1).[a]
Enzymes/
Organisms
11
12
13
Figure 4. Stereo representation of superimposition of TbFPPS·6 (PDB ID:
2P1C) with EcFPPS (PDB ID: 1RQJ). The side chains of HY (red) and YS
(yellow) amino acids are shown in stick format, compound 6 is shown in
purple, the first aspartate-rich motif (FARM) is shown in magenta, and Mg2+
ions are shown as orange spheres.
EcFPPS
PaFPPS
TbFPPS
HsFPPS
SmFPPS
HsGGPPS
0.12ꢁ0.025
0.32ꢁ0.068
0.12ꢁ0.026
0.065ꢁ0.012
0.57ꢁ0.11
1.6ꢁ0.28
>100
>100
0.58ꢁ0.15
0.16ꢁ0.029
0.12ꢁ0.026
0.33ꢁ0.054
0.33ꢁ0.089
0.51ꢁ0.20
0.084ꢁ0.0083
0.50ꢁ0.065
0.53ꢁ0.11
0.39ꢁ0.073
shows a stereo structural alignment of TbFPPS·6 (PDB ID:
2P1C) with EcFPPS (PDB ID: 1RQJ) from which it can be seen
that the conserved YS residues in EcFPPS could align with the
(electron-rich) Tyr in close proximity (~3 ꢁ) to the phenyl
group in the TbFPPS·6 structure, consistent with a role for
a charge-transfer interaction[11] with inhibitors with electron-
withdrawing phenyl substituents. This is a hypothesis, of
course, but clearly there are major differences in FPPS inhibi-
tion between 11, 12, and 13 in the bacterial and eukaryotic
S. aureus
B. subtilis
>100
>100
>100
>100
39ꢁ13
26ꢁ5.9
M. smegmatis
E. coli
A. baumannii
K. pneumonia
P. aeruginosa
HEK293
6.8ꢁ3.0
2.1ꢁ0.57
2.8ꢁ0.70
2.3ꢁ0.49
3.5ꢁ0.78
27ꢁ4.0
7.5ꢁ2.4
1.7ꢁ0.38
2.0ꢁ0.50
2.4ꢁ0.43
3.3ꢁ0.68
36ꢁ8.6
20ꢁ5.3
>100
>100
>100
>100
82ꢁ22
[a] Data are the meanꢁSD for duplicate experiments.
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
systems, and charge transfer seems a likely reason for the dif-
ferences observed.
MICðABÞ MICðBAÞ
FICI ¼ FICA þ FICB ¼
þ
MICðAÞ
MICðBÞ
We next tested 11–13 in cell growth inhibition assays: first
against the Gram-positives S. aureus and B. subtilis, as well as
Mycobacterium smegmatis, second against the Gram-negatives
E. coli, P. aeruginosa, K. pneumoniae and A. baumannii, and third
against a human embryonic kidney cell line (HEK293). Results
are shown in Table 1 and Figures S2 and S3. There was little or
no activity of 11–13 against S. aureus or B. subtilis, but there
was some activity against M. smegmatis (corresponding to an
IC50 value of ~7 mgmLꢀ1 for 11 and 12). However, the difluoro
species 11 as well as the bromo species 12 had promising ac-
tivity against all four Gram-negative bacteria in the ~1–
4 mgmLꢀ1 range, but the methoxy analogue 13 was inactive
(Table 1). These results are of interest as they indicate that in
the Gram-negatives, 11–13 exhibit the same overall pattern of
activity in cells as that observed for EcFPPS and PaFPPS
enzyme inhibition (Table 1), for which 13 was inactive. Why
there is less activity against S. aureus and B. subtilis is not
known, and naturally could involve both differences in uptake/
efflux as well as FPPS inhibition, although the latter possibility
seems unlikely given the strong sequence similarities in the
active site regions (Figure 3). Therefore, how can we further
test whether FPPS is actually a target for 11 and 12 in the
Gram-negative organisms?
for which FICA and FICB are the fractional inhibitory concentra-
tions of drugs A and B, MIC(A) and MIC(B) are the MIC values
of drugs A and B acting alone, and MIC(AB) and MIC(BA) are
the MIC values of the most effective combination of drug A or
B in the presence of drug B or A. Using this method, FICI
values of <0.5 represent synergy, >0.5 and <1.0 represent ad-
ditivity, >1 and <2 represent an indifferent effect, and ꢂ2
represents antagonism.[15] In addition, we evaluated isobolo-
grams using the method of Berenbaum.[16] FICI values are
shown in Table 2 and isobolograms in Figure 5. We found FICI
values of 0.39ꢁ0.15 (E. coli), 0.64ꢁ0.23 (A. baumannii), 0.22ꢁ
0.026 (K. pneumoniae), and 0.44ꢁ0.12 (P. aeruginosa); Table 2.
FICI values of <0.5 are generally taken to indicate a synergistic
interaction, so these results are consistent with 11 inhibiting
the same pathway as does fosmidomycin, that is, isoprenoid
biosynthesis.
Another possible test to determine whether FPPS is a target
is to see if cell growth inhibition can be “rescued” by the addi-
tion of farnesol. The use of farnesol, farnesyl diphosphate, ger-
aniol, and geranylgeranyl diphosphate has been used previ-
ously—in mammalian cell lines as well as in trypanosomes
(parasitic protozoa)—to assess the effects of bisphosphonates
on cell growth, and in mammalian cell lines it is known that
there kinases convert geranylgeraniol to geranylgeranyl di-
phosphate (GGOH!GGPP),[17] with GGOH rescuing cells from
bisphosphonate growth inhibition. There are also kinases that
can convert farnesol to farnesyl diphosphate (FOH!FP), for
example, but whether such prenol kinases are present in the
To help answer this question, we first investigated whether
there were synergistic effects between bisphosphonate 11 and
fosmidomycin (7) in E. coli, A. baumannii, K. pneumoniae, and
P. aeruginosa, basically as we described previously for 6 in
E. coli. More specifically, we determined the fractional inhibito-
ry concentration index (FICI) values for each combination
using the FICI formula:[14]
Table 2. Combination of 11 with fosmidomycin (fos) against Gram-negative bacteria.[a]
Organism
MIC fos
[mgmLꢀ1
FIC fos
MIC 11
[mgmLꢀ1
]
FIC 11
FICI[b]
]
E. coli
5
5
5
3
0.14ꢁ0.13
0.27ꢁ0.057
0.12ꢁ0.014
0.18ꢁ0.066
6
6
6
0.24ꢁ0.023
0.37ꢁ0.17
0.39ꢁ0.15
0.64ꢁ0.23
0.22ꢁ0.026
0.44ꢁ0.12
A. baumannii
K. pneumoniae
P. aeruginosa
0.097ꢁ0.012
0.27ꢁ0.057
10
[a] Data are the meanꢁSD for duplicate experiments. [b] Fractional inhibitory concentration index.
Figure 5. Isobolograms for bacterial cell growth inhibition by 11 and fosmidomycin (7) against a) E. coli, b) A. baumannii, c) K. pneumoniae, and d) P. aerugino-
sa. The fractional inhibitory concentration indices (FICI) are listed in Table 2. The mean FICI value is 0.42ꢁ0.17, indicating synergistic activity (in most cases).
&
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
organisms studied here is not known, so caution needs to be
exercised in interpreting the results. We grew all four Gram-
negative bacteria in the absence or presence of 200 mm FOH
resulting in, on average, an approximate eightfold increase in
IC50 for growth inhibition by the bisphosphonates 11 and 12,
consistent with an FPPS target (Figure 6).
Table 3. IC50 values of compounds 11, 12, 32, and 41 in E. coli, A. bau-
mannii, K. pneumoniae, P. aeruginosa, E. coli overexpressing P. aeruginosa
FPPS (PaFPPS+ +), and E. coli overexpressing E. coli UPPS (EcUPPS+ +).
[a]
Organism
Inhibitor
Rescuing
IC50 [mgmLꢀ1
]
Agent (RA)
without RA
with RA
E. coli
E. coli
11
12
11
12
11
12
11
12
11
12
32
41
Farnesol
Farnesol
Farnesol
Farnesol
Farnesol
Farnesol
Farnesol
Farnesol
PaFPPS+ +
PaFPPS+ +
EcUPPS+ +
EcUPPS+ +
2.1ꢁ0.57
1.7ꢁ0.38
2.8ꢁ0.70
2.0ꢁ0.50
2.3ꢁ0.49
2.4ꢁ0.43
3.5ꢁ0.78
3.3ꢁ0.68
1.6ꢁ0.35
2.5ꢁ1.1
19ꢁ4.1
28ꢁ9.0
18ꢁ4.9
22ꢁ2.4
13ꢁ1.8
21ꢁ4.2
23ꢁ7.8
22ꢁ3.5
19ꢁ2.4
15ꢁ3.6
86ꢁ17
107ꢁ15
Next, we carried out growth inhibition assays of 11 and 12
against an E. coli strain that overexpresses PaFPPS,[18] again re-
sulting in a considerable (~8-fold) increase in the IC50 values
for both 11 and 12 (Table 3). Based on the patterns of FPPS
and cell growth inhibition by 11–13, synergistic activity with
fosmidomycin, partial rescue of cell growth inhibition by FOH,
and FPPS overexpression results, we conclude that the activity
of 11 and 12 in Gram-negative bacteria is due—at least in
part—to FPPS inhibition. The two active compounds 11 and
12 contain electron-withdrawing substituents and are likely to
undergo charge-transfer interactions with the conserved Tyr79
(Figures 2 and 3), an interaction that would be absent with the
methoxy side chain containing species 13. It is also possible
that there are additional targets, but the key point is that 11
and 12 are the first bisphosphonates to have promising (1–
4 mgmLꢀ1) IC50 values against these four Gram-negative bacte-
ria—values that are about an order of magnitude higher than
found with 6. FPPS is at least one of the targets involved, and
there was little effect on HEK293 cell growth.
A. baumannii
A. baumannii
K. pneumoniae
K. pneumoniae
P. aeruginosa
P. aeruginosa
E. coli
E. coli
E. coli
E. coli
6.4ꢁ1.3
9.0ꢁ1.5
[a] Data are the meanꢁSD for duplicate experiments.
M. smegmatis, E. coli, A. baumannii, K. pneumoniae, and P. aeru-
ginosa. Compounds 16, 17, 23, 24, 26, 28, 29, and 34–38 were
found to be inactive. Results for all active compounds are
shown in Table S1 (SMILES structures in Table S2) and Fig-
ure S5, and results for the most active compounds are listed in
Table 4. Several compounds containing long akyl side chains,
such as 32 and 41, had activity in the 2–6 mgmLꢀ1 range
against Gram-positive, but not Gram-negative bacteria
(Table 4). These results then raise the question: how do these
compounds inhibit bacterial cell growth?
Alkyl phosphonate and carboxylate inhibitors
We next synthesized a series of hydroxy-substituted alkyl phos-
phonic acids and carboxylic acids. Our initial interest in these
types of compound (structures shown in Figure S4) was that
they might inhibit the enzyme dihydroxyacid dehydratase
(DHAD), which in many bacteria is essential for branched-chain
amino acid biosynthesis; for example, Flint and Nudelman re-
ported the 2,3-dihydroxyoctanoic acid 14 (Figure 2) and the 1-
hydroxy-2-methylpropylphosphonic acid 15 to be inhibitors of
DHAD from plants and E. coli.[19] The compounds were shown
to bind enzymes that contain iron–sulfur clusters, but it ap-
peared to us that they might also have prenyl transferase in-
hibitory activity. We therefore synthesized and tested a series
of dihydroxy acids and hydroxyphosphonates (16–41, Fig-
ure S4) and tested them against the same panel of bacteria as
described for the bisphosphonates: S. aureus, B. subtilis,
We reasoned that isoprenoid biosynthesis might again be
one target area since there is some similarity between the
more potent bacterial cell growth inhibitors—the lipophilic hy-
droxyphosphonates 32 and 41—and GPP or FPP, substrates in-
volved in many prenyl transferase reactions. To determine
whether isoprenoid biosynthesis inhibition was a likely target,
we carried out 12 FICI determinations using the 1-hydroxy-
phosphonate 32 together with either a known antibacterial
isoprenoid/cell wall biosynthesis inhibitor (fosmidomycin, car-
benicillin, vancomycin, ampicillin, bacitracin, cefotaxime, fosfo-
mycin) or an antibacterial compound that does not target
these pathways (kanamycin, tetracycline, chloramphenicol,
spectinomycin, sulfamethoxazole, trimethoprim). We obtained
Figure 6. Partial rescue of cell growth inhibition by farnesol or PaFPPS overexpression. Addition of 200 mm farnesol (FOH) to the growth medium increases
the IC50 values of 11 and 12 for cell growth inhibition by a factor of ~8. a) E. coli K-12 with FOH; b) A. baumannii with FOH; c) K. pneumoniae with FOH;
d) P. aeruginosa with FOH; and e) E. coli BL21(DE3) with PaFPPS overexpression.
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Table 4. IC50 values of dihydroxy acids and hydroxyphosphonates against bacterial cells and enzymes.
IC50 [mgmLꢀ1
Compound
]
IC50 [mm]
S. aureus B. subtilis M. smegmatis E. coli A. baumannii K. pneumoniae P. aeruginosa
SaUPPS EcUPPP
5.2
1.7
3.1
39
14
8.1
6.4
8.1
>100
>100
>100
39
>200
2.0
3.7
5.9
6.4
37
0.73
0.92
5.0
8.7
6.2
24
>100
25
>100
>100
>100
2.4
3.7
>100
>100
21
>100
>100
25
11
6.4
10
7.0
9.0
14
2.8
3.4
data on two organisms: B. subtilis, chosen because it uses the
non-mevalonate (MEP) pathway and is inhibited by fosmido-
mycin (which targets 1-deoxy-d-xylulose 5-phosphate reductoi-
somerase, DXR), and S. aureus, which uses the mevalonate
pathway.
gets, based on structure, would be FPPS, UPPS, and UPPP, as
these are all in the isoprenoid/cell wall biosynthesis pathway
(where we see synergy) while for example, prenyl synthases,
such as those involved in quinone biosynthesis, are not.
We then tested all compounds against EcFPPS, SaUPPS, and
EcUPPP[20] using phosphate-release assays.[21] There was no ac-
tivity (IC50 >300 mm) against EcFPPS or PaFPPS (data not
shown), but several compounds (Figures S7 and S8) were
found to be active against UPPS and UPPP, and dose–response
curves for the most active species are shown in Figure 8a,b.
Both the 1-OH and 2-OH phosphonates (32 and 41) were
active against SaUPPS, with IC50 values of ~1–3 mm. For the
UPPP activity inhibition assay, we found that 32 and 41 inhibit-
ed UPPP with IC50 values of ~1–4 mm (corresponding to Ki ~
300–980 nm, assuming competitive inhibition).[22] Under the
same assay conditions, the IC50 value for UPPP inhibition by ba-
citracin was 32 mm (Figure 8b). So, compounds 32 and 41 in-
Typical experimental (isobologram) results are shown in
Figure 7 and Figure S6, and the FICI results are listed in Table 5.
The results indicate that for each case in which the second in-
hibitor targets isoprenoid/cell wall biosynthesis, there is a syn-
ergistic interaction (FICIavg =0.31ꢁ0.11, n=13), whereas for
cases in which the second inhibitor is not involved in isopre-
noid/cell wall biosynthesis, there is an indifferent (albeit not
antagonistic) effect (FICIavg =1.53ꢁ0.19, n=11). We again use
the definition that FICI values <0.5 mean synergy, 0.5–1 addi-
tivity, 1–2 indifference, and >2 antagonism. These results
strongly support the idea that this hydroxyphosphonate, 32,
targets isoprenoid/cell wall biosynthesis. Some possible tar-
Figure 7. Representative isobolograms for 32 with antibiotics having known mechanisms of action. a) 32+fosmidomycin in B. subtilis showing synergy
(FICI=0.27) of 32 with a cell wall biosynthesis inhibitor (that targets 1-deoxy-d-xylulose-5-phosphate reductoisomerase (DXR) in the non-mevalonate path-
way); b) 32+sulfamethoxazole in B. subtilis showing an indifferent effect (FICI=1.78) of 32 with a nucleic acid biosynthesis inhibitor (that targets dihydropter-
oate synthase); c) 32+bacitracin in S. aureus showing synergy (FICI=0.24) of 32 and a cell wall biosynthesis inhibitor (that targets UPPP); d) 32+tetracycline
in S. aureus showing an indifferent effect (FICI=1.61) of 32 with a protein synthesis inhibitor (that targets ribosome function).
&
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Table 5. Effects of the addition of various known bacterial cell growth inhibitors on the inhibition of B. subtilis and S. aureus cell growth by compound 32.
The cell targets of the known inhibitors are listed, together with the FICI values (and errors) for the combinations.[a]
Antibiotic
MIC antibiotic [mgmLꢀ1
]
FIC antibiotic
B. subtilis
MIC 32 [mgmLꢀ1
]
FIC 32
FICI
fosmidomycin
carbenicillin
cefotaxime
vancomycin
fosfomycin
ampicillin
5
5
0.25ꢁ0.043
0.25ꢁ0.043
0.30ꢁ0.044
0.33ꢁ0.058
0.21ꢁ0.025
0.18ꢁ0.041
0.12ꢁ0.028
10
10
10
10
10
10
10
0.013ꢁ0.0037
0.097ꢁ0.049
0.075ꢁ0.012
0.080ꢁ0.059
0.22ꢁ0.052
0.048ꢁ0.023
0.23ꢁ0.010
0.27ꢁ0.046
0.34ꢁ0.048
0.38ꢁ0.056
0.41ꢁ0.12
0.43ꢁ0.077
0.23ꢁ0.063
0.35ꢁ0.13
Cell Wall
Biosynthesis
Inhibitors
2.5
0.5
200
0.5
200
bacitracin
Protein
Biosynthesis
Inhibitors
kanamycin
tetracycline
chloramphenicol
1.5
5
0.5
0.61ꢁ0.097
0.82ꢁ0.21
0.55ꢁ0.13
10
10
10
0.95ꢁ0.097
0.80ꢁ0.15
0.93ꢁ0.32
1.55ꢁ0.19
1.61ꢁ0.35
1.47ꢁ0.45
Nucleic Acid
Inhibitors
sulfamethoxazole
trimethoprim
200
0.5
0.87ꢁ0.23
0.77ꢁ0.18
10
10
0.91ꢁ0.25
1.78ꢁ0.48
1.37ꢁ0.26
0.60ꢁ0.084
S. aureus
carbenicillin
cefotaxime
vancomycin
fosfomycin
ampicillin
15
2
1.5
200
0.5
200
0.27ꢁ0.015
0.21ꢁ0.048
0.19ꢁ0.019
0.13ꢁ0.022
0.30ꢁ0.095
0.21ꢁ0.039
10
10
10
10
10
10
0.052ꢁ0.0077
0.18ꢁ0.057
0.32ꢁ0.022
0.40ꢁ0.11
0.34ꢁ0.052
0.18ꢁ0.036
0.35ꢁ0.11
0.24ꢁ0.048
Cell Wall
Biosynthesis
Inhibitors
0.16ꢁ0.033
0.052ꢁ0.014
0.056ꢁ0.012
0.027ꢁ0.0090
bacitracin
kanamycin
tetracycline
chloramphenicol
spectinomycin
1.5
0.5
5
0.81ꢁ0.17
0.83ꢁ0.38
0.71ꢁ0.11
0.45ꢁ0.091
10
10
10
10
0.83ꢁ0.17
0.93ꢁ0.13
0.73ꢁ0.21
0.84ꢁ0.22
1.64ꢁ0.34
1.76ꢁ0.51
1.44ꢁ0.33
1.29ꢁ0.31
Protein
Biosynthesis
Inhibitors
40
Nucleic Acid
Inhibitors
sulfamethoxazole
trimethoprim
200
15
0.75ꢁ0.20
0.69ꢁ0.32
10
10
0.99ꢁ0.17
0.53ꢁ0.13
1.74ꢁ0.37
1.22ꢁ0.45
inhibitors targeting cell wall biosynthesis
inhibitors targeting nucleic acids and protein biosynthesis
inhibitors targeting cell wall biosynthesis
0.31ꢁ0.14
1.56ꢁ0.15
0.31ꢁ0.080
1.52ꢁ0.23
B. subtilis
S. aureus
Mean FICIs:
inhibitors targeting nucleic acids and protein biosynthesis
[a] The mean value for isoprenoid/cell wall biosynthesis inhibitors is FICIavg =0.31ꢁ0.11, n=13 (indicating synergistic activity); that for other inhibitors is FI-
CIavg =1.53ꢁ0.19, n=11 (indicating an indifferent although not antagonistic effect). Data are the meanꢁSD for duplicate experiments.
Figure 8. Enzyme and cell growth inhibition by the dihydroxyacid 20 and hydroxyphosphonates 32 and 41. a) SaUPPS inhibition; b) EcUPPP inhibition togeth-
er with result for bacitracin; c) partial rescue of E. coli cell growth inhibition by EcUPPS overexpression.
hibit both UPPS as well as UPPP, consistent with the observa-
tion that 32 shows synergistic activity with all isoprenoid/cell
wall biosynthesis inhibitors tested. The dihydroxy acid 20
showed weak inhibition against SaUPPS and no inhibition (>
200 mm) against EcUPPP, resulting in less activity in cells. We
also found that overexpressing EcUPPS caused an approximate
sixfold increase in the IC50 value for E. coli cell growth inhibi-
tion by both 32 and 41 (Figure 7c and Figure S9). We therefore
conclude that both UPPS as well as UPPP are likely targets for
the most potent hydroxyphosphonates. Because 32 and 11 in-
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
hibit different enzymes involved in isoprenoid biosynthesis, we
also tested for synergistic activity against E. coli cell growth. As
expected, 32 and 11 have synergistic activity with an FICI
value of 0.31ꢁ0.099 (Figure S10). Finally, we tested 20, 32, and
41 for their effects on growth of the human embryonic kidney
cell line HEK293. The MIC values for cell growth inhibition
were in the range of 90–410 mgmLꢀ1 (Figure S3), so these com-
pounds are not highly toxic toward at least this human cell
line.
Experimental Section
General methods: All chemicals were reagent grade. 1H and
13C NMR spectra were obtained on Varian Unity spectrometers
(Palo Alto, CA, USA) at 400 and 500 MHz for 1H and at 100 and
125 MHz for 13C. Elemental analyses were carried out in the Univer-
sity of Illinois Microanalysis Laboratory. HPLC–MS analyses were
performed with an Agilent LC–MSD Trap XCT Plus system (Agilent
Technologies, Santa Clara, CA, USA) with an 1100 series HPLC
system including a degasser, an autosampler, a binary pump, and
a multiple wavelength detector. All final compounds were ꢂ95%
pure as determined by quantitative spin count NMR (qNMR) or
HPLC–MS, and structures were characterized by 1H NMR and
HRMS.
Conclusions
Chemical syntheses: Hydrogen (2-(3-(3-(3,4-difluorophenyl)prop-
yl)-1H-imidazol-3-ium-1-yl)-1-phosphonoethyl)
phosphonate
The results we report herein are of interest for several reasons.
We synthesized a broad range of compounds—bisphospho-
nates, hydroxyphosphonates, and dihydroxyacids—to identify
novel inhibitors of isoprenoid biosynthesis that also inhibit
bacterial cell growth. Based on earlier work on E. coli cell
growth inhibition and the structure of a T. brucei FPPS–bi-
sphosphonate complex, we made novel bisphosphonates con-
taining electron-withdrawing aryl substituents that inhibited
E. coli, K. pneumoniae, A. baumannii, and P. aeruginosa cell
growth in the 1–4 mgmLꢀ1 range. Growth inhibition was par-
tially rescued (an increase in IC50 by a factor of ~8) by the addi-
tion of farnesol, and in cells overexpressing FPPS, FICI values
for one bisphosphonate with fosmidomycin were 0.42ꢁ0.17,
consistent with an isoprenoid biosynthesis target. We also pro-
duced a series of dihydroxy acids and hydroxyphosphonates
which we found to inhibit UPPS as well as UPPP, along with
Gram-positive (but not Gram-negative) bacterial cell growth.
These compounds, as well as the bisphosphonates, acted syn-
ergistically with other known inhibitors of isoprenoid biosyn-
thesis/bacterial cell wall growth. Overall, the results are of in-
terest because we have found several new prenyl synthase in-
hibitors that target FPPS, UPPS, or UPPP, that also have low mi-
cromolar activity against numerous pathogenic bacteria.
(11). To a solution of 3,4-difluorobenzaldehyde a (1.4 g, 10 mmol)
in 20 mL methanol was added triethyl phosphonoacetate (3 mL,
15 mmol; Scheme 1). The reaction mixture was cooled to 08C and
into it was slowly added 2 mL sodium methoxide (30 wt% solution
in methanol). The mixture was stirred overnight at room tempera-
ture. The solvent was removed under reduced pressure, and the
crude material partitioned between ethyl acetate and water. The
organic layer was dried over Na2SO4, and solvent removed under
reduced pressure. The product was purified by column chromatog-
raphy (silica gel, eluting with hexane/ethyl acetate 30:1) to give
the unsaturated ester b as an oil (1.7 g, 80%). To a solution of
b (1.7 g, 8 mmol) in methanol (100 mL) was added palladium on
charcoal (0.6 g, 10%) under hydrogen. The reaction mixture was
stirred at room temperature for 3 h. Solvents were removed under
reduced pressure. The product was partitioned between water and
ethyl acetate, the organic layer separated, dried over Na2SO4, and
solvent removed under reduced pressure. The ester product c was
obtained as a colorless oil (1.6 g, 95%). To a solution of lithium alu-
minum hydride (160 mg, 4 mmol) in anhydrous tetrahydrofuran
(THF) was slowly added a solution of c (1.6 g, 8 mmol) in THF
(30 mL). The reaction mixture was stirred at room temperature
overnight. Then, water (12 mL), sodium hydroxide (4n, 12 mL) and
finally water (36 mL) were added at 08C. The mixture was stirred,
and the resulting salts filtered through a pad of Celite, washing
with ethyl acetate (100 mL). The product was treated with water
Scheme 1. Synthesis route to FPPS inhibitors 11–13.
&
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
and CH2Cl2 and the organic layer separated, dried over Na2SO4, and
solvent removed under reduced pressure to give the alcohol prod-
uct d as an oil (1.0 g, 90%). To a solution of d (1.0 g, 7 mmol) in
100 mL CH2Cl2 was added triphenylphosphine (2.3 g, 8 mmol). The
mixture was stirred at room temperature, and N-bromosuccinimide
(NBS, 1.4 g, 8 mmol) added in small portions. The mixture was
stirred overnight then washed with water and extracted with
hexane. The organic layer was separated, dried over Na2SO4, and
solvent removed under reduced pressure to give the bromide
product e as a light-yellow solid (1.5 g, 95%). Imidazole (1.0 g,
15 mmol) was dissolved in THF (30 mL), and NaH (240 mg,
10 mmol) was added. The mixture was stirred at room temperature
for 30 min, then e (1.5 g, 7 mmol) was added. The reaction mixture
was heated at 808C and stirred overnight, then quenched with
water, and the mixture was extracted with diethyl ether. The or-
ganic layer was separated, dried over Na2SO4, and solvent removed
under reduced pressure. The crude product was purified by
column chromatography (silica gel, hexane/ethyl acetate 2:1). The
imidazole product f was obtained as a light-yellow solid (1.05 g,
70%). Paraformaldehyde (1.35 g, 45.0 mmol) and diethylamine
(0.68 g, 9.33 mmol) were dissolved in dry methanol (30 mL), with
warming. A solution of tetraethyl methylenebisphosphonate g
(2.7 g, 9.33 mmol) in dry methanol (30 mL) was added at 208C, and
the mixture stirred for five days. The reaction mixture was then
concentrated under vacuum, toluene (20 mL) was added, and the
solution was concentrated again. This last step was then repeated
to remove all traces of methanol, yielding tetraethyl 1-methoxy-
methyl methylene 1,1-bisphosphonate h as a colorless oil (2.9 g,
95%). Intermediate h was added to pTSA (cat) and toluene
(100 mL), and the mixture was heated for 16 h at reflux with
a Soxhlet apparatus charged with 4 ꢁ molecular sieves. The mix-
ture was allowed to cool to 208C, then washed with water (2ꢂ
20 mL). Drying (MgSO4) and concentrating under vacuum gave i as
a colorless viscous oil (2.7 g, 99%). Intermediate i (2.7 g, 9 mmol)
was dissolved in dry CH2Cl2, cooled to 08C, and Me3SiBr (12 mL,
90 mmol) was added dropwise over 30 min. The mixture was
stirred for two days at room temperature. The solvent was then
evaporated, and the residue was dried under vacuum for ~1 h.
Then, 80 mL dry methanol was added, and the mixture was stirred
for 20 min at room temperature. The solvent was removed, and
the brown oil was dried overnight to give compound j (1.6 g,
95%). Intermediate f (220 mg, 1 mmol) was dissolved in AcOH,
then j (206 mg, 1.1 mmol) was slowly added. The mixture was
stirred at reflux for three days. The AcOH was then evaporated
under vacuum. The residue was suspended in methanol and soni-
cated for 1 min. The mixture was then centrifuged, and the liquid
layer discarded. The latter procedure was repeated twice. Then,
the remaining residue was dried under vacuum overnight to give
500 MHz): d=8.62 (s, 1H), 7.39 (s, 1H), 7.24 (s, 1H), 7.00 (m, 2H),
6.80 (m, 1H), 4.43 (dt, J=7.0, 12.5 Hz, 2H), 4.02 (t, J=7.5 Hz, 2H),
2.50 (t, J=7.5 Hz, 2H), 2.41 (tt, J=7.0, 12.5 Hz, 1H), 2.04 ppm (m,
2H); 31P NMR (D2O, 202 MHz): d=15.29 ppm; ESI HRMS: m/z [M+
H]+ calcd for C14H19F2N2O6P2+: 411.0686, found: 411.0690. The
purity of the product was determined by qNMR (potassium hydro-
gen phthalate as standard, 99.95%): 98.3%.
General procedure for 2,3-dihydroxy acids: 2,3-Dihydroxy acids
were synthesized according to a published method (Scheme 2a).[23]
To a stirred mixture of the corresponding aldehyde (10 mmol) and
tris-(trimethylsilyloxy) ethylene (2.9 g, 10 mmol), 1–2 drops of
tin(IV) chloride were added (exothermic). After stirring for 3 h, 2:2:1
acetic acid/THF/water (20 mL) was added. The solutions were
stirred for 10 min, and solvents were removed under reduced pres-
sure at 50–608C. Toluene was added and removed, twice. The resi-
dues were recrystallized from chloroform/methanol to give the 2,3-
dihydroxy acids.
General procedure for (1-hydroxy)phosphonic acids: 1-Hydro-
phosphonic acids were synthesized according to a published
method (Scheme 2b).[24] To a stirred solution of diethyl phosphite
(1.4 g, 10 mmol) and triethylamine (1.0 g, 10 mmol), the corre-
sponding aldehyde (10 mmol) was added dropwise, at room tem-
perature (exothermic). After addition was complete, the reaction
mixture was kept at room temperature overnight. The mixtures
were concentrated, and the residues were purified by column
chromatography on silica (hexane/ethyl acetate) to give the 1-hy-
droxyalkyl phosphonates, which were then treated with 6m hydro-
chloric acid (20 equiv). The resulting solutions were held at reflux
under N2 for 1–2 days. The reaction mixtures were then concen-
trated and subjected to high vacuum, yielding the 1-hydroxyphos-
phonic acids.
General procedure for (2-hydroxy)phosphonic acids: To a solution
of diethyl methylphosphonate (152 mg, 1 mmol) in THF (5 mL) was
added 1.2 equiv 1.6m nBuLi at ꢀ788C, Scheme 2c. The solution
was stirred at the same temperature for 1 h. A solution of corre-
sponding aldehyde (1 mmol) in THF (5 mL) was slowly added. The
resulting solution was stirred for 30 min at ꢀ788C and then
quenched by the addition of saturated aqueous NH4Cl. The aque-
ous phase was extracted with Et2O. The combined organic layers
were concentrated and the product purified by column chroma-
tography on silica (hexane/ethyl acetate) to give the 2-hydroxyalkyl
phosphonate, which was then treated with 6m hydrochloric acid
(20 equiv). The resulting solution was held at reflux under N2 for
1–2 days, concentrated, then subjected to high vacuum, yielding
the 2-hydroxyphosphonic acids.
E. coli growth inhibition assay: IC50 values for E. coli cell growth
inhibition were determined using a microdilution method. Station-
1
the final product k (11) as a white solid (3.1 g, 85%). H NMR (D2O,
Scheme 2. General methods for synthesis of: a) 2,3-dihydroxy acids; b) 1-hydroxyphosphonic acids; c) 2-hydroxyphosphonic acids.
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
ary overnight starter cultures of E. coli (K-12 or BL21(DE3) strains),
were diluted 1000-fold and grown to an OD600 value of ~0.3. These
log-phase cultures were then diluted 500-fold into fresh LB broth
to generate a working solution; 200 mL of working solution was
transferred into each well of a 96-well culture plate (Corning 3370).
Inhibitors were then added at 1 mm and sequentially diluted three-
fold to 46 nm, keeping volume and culture broth composition con-
stant. Plates were incubated for 12 h at 378C, shaking at 200 rpm.
Absorbance at 600 nm was then measured to assess bacterial cell
growth. IC50 values were determined using nonlinear regression,
whereas minimum inhibitory concentration (MIC) values in the syn-
ergy assays were calculated by using a Gompertz function in
Prism 5 (GraphPad Software Inc., La Jolla, CA, USA).
(200 mgmLꢀ1 to 90 ngmLꢀ1 for bacitracin, fosfomycin, and sulfame-
thoxazole) in the presence half-MIC concentrations of 11 and 32, in
addition to a threefold gradient of 11 and 32 ranging from
40 mgmLꢀ1 to 18 ngmLꢀ1 in the presence of half-MIC concentra-
tions of each antibiotic. New MIC values were calculated by using
a Gompertz function in Prism 5 (GraphPad Software Inc., La Jolla,
CA, USA).
Enzyme inhibition assays: FPPS, SaUPPS, and EcUPPP were ex-
pressed and purified as described previously.[3,5,9,10,12,19,22] FPPS and
UPPS assays were carried out using a phosphate release assay.[5,21]
Depending on the solubility, bisphosphonates, hydroxyphospho-
nates, and dihydroxy acid inhibitors were prepared as 10 mm stock
solutions in DMSO or basic water (pH~10) and then serially diluted
from 1 mm to 1 nm. Inhibitors were incubated with 25 ng SaUPPS
at room temperature for 10 min in a pH 7.5 buffer (50 mm HEPES,
150 mm NaCl, 10 mm MgCl2, and 0.02% n-dodecyl-b-d-maltopyra-
noside) before adding a reaction mixture containing 5 mm FPP,
50 mm IPP, 3 UmLꢀ1 purine nucleoside phosphorylase, 1 UmLꢀ1 in-
organic phosphatase, and ~600 mm 7-methyl-6-thioguanosine
(MESG), again in the same buffer. For FPPS inhibition assay, inhibi-
tors were incubated with 25 ng of various FPPS enzymes at room
temperature for 10 min in a pH 7.0 buffer (10 mm HEPES, 150 mm
NaCl, 5 mm MgCl2) before adding a reaction mixture containing
50 mm GPP, 50 mm IPP, 3 UmLꢀ1 purine nucleoside phosphorylase,
1 UmLꢀ1 inorganic phosphatase, and ~600 mm 7-methyl-6-thiogua-
nosine (MESG), again in the same buffer. FPPS and UPPS reactions
were monitored for 15 min with the rate of increase in absorbance
at 360 nm taken as the rate of FPP or UPP synthesis, respectively.
IC50 values were calculated by using Prism 5 (GraphPad Software
Inc., La Jolla, CA, USA). The UPPP inhibition assay was carried out
using a malachite-green reagent as described previously.[25] The
same 10 mm inhibitor stock solutions and assay buffer as for the
SaUPPS assays were used to test for UPPP inhibition. Inhibitors
were incubated with 20 nm EcUPPP at room temperature for
15 min before adding FPP to 35 mm. Reaction mixtures were incu-
bated at 378C for 20 min, then quenched by adding 30 mL mala-
chite-green reagent. In this assay, the phosphate released from FPP
reacts with ammonium molybdate to form phosphomolybdate
(yellow), which then forms a complex (lmax ~620 nm) with mala-
chite-green, used to assess phosphatase activity. Phosphate release
was measured at 620 nm and quantified based on a phosphate
standard curve, and the OD620 values used to construct dose–re-
sponse curves.
Gram-negative bacterial cell growth inhibition assay: As with
the E. coli inhibition assays, overnight cultures (in cation-adjusted
Mueller–Hinton broth, CAMHB) of A. baumannii (Bouvet and Gri-
mont, ATCC 19606); K. pneumoniae (subsp. pneumoniae Schroeter
Trevisan ATCC 27736); and P. aeruginosa (PA01) were diluted 1000-
fold (in fresh CAMHB) to create a working solution. Working solu-
tions were then transferred into flat-bottom 96-well plates, and in-
hibitors added at 1 mm and sequentially diluted threefold to
46 nm. Plates were incubated at 378C, shaking at 200 rpm, over-
night. The OD600 value was then measured to determine bacterial
growth inhibition.
B. subtilis growth inhibition assay: An overnight starter culture (in
LB broth) of B. subtilis (subsp. subtilis (Ehrenberg) Cohn ATCC 6051)
was diluted 1000-fold (in fresh LB media) to create a working solu-
tion. Working solutions were then transferred into flat-bottom 96-
well plates, and inhibitors were added at 1 mm and sequentially di-
luted threefold to 46 nm. Plates were incubated at 378C, shaking
at 200 rpm overnight. The OD600 value was then measured to de-
termine bacterial growth inhibition.
S. aureus growth inhibition assay: An overnight starter culture of
S. aureus (Newman strain) in tryptic soy broth was diluted 1000-
fold in fresh tryptic soy media to create a working solution. Work-
ing solutions were transferred into flat-bottom 96-well plates and
inhibitors added at 1 mm and sequentially diluted threefold to
46 nm. Plates were incubated at 378C, shaking at 200 rpm over-
night. The OD600 value was then measured to determine bacterial
growth inhibition.
HEK293 toxicity assay: A frozen stock of human embryonic kidney
cells (HEK293 ATCC CRL-1573) was used to grow a first generation
of cells in DMEM (4.5 gLꢀ1 glucose with l-glutamine) containing
10% fetal bovine serum (FBS) and 1% penicillin–streptomycin
(10000 UmLꢀ1). This generation was harvested in 0.25% trypsin/
2.1 mm EDTA, and cells were counted under a light microscope. A
working solution was generated containing 105 cells per mL, which
was then transferred into a flat-bottom 96-well plate for 36 h. At
this time, 20 mL of inhibitor solutions ranging from 1 mm to 46 nm
were added, and cells were allowed to grow in the presence of the
inhibitors for an additional 24 h. MTT solution (10 mL, 5 mgmLꢀ1 in
PBS) was then added to each well and incubated for 4 h. HCl in
isopropanol (100 mL of 100 mm) was added to each well, and ab-
sorbance at 570 nm was measured. MIC values were calculated by
using a Gompertz function in Prism 5 (GraphPad Software Inc., La
Jolla, CA, USA).
Supporting Information: Enzyme and cell growth inhibition tables
and graphs, isobolograms, compound synthesis and characteriza-
1
tion, HPLC purity results, H NMR spectra of inhibitor compounds.
Acknowledgements
This work was supported by the United States Public Health Ser-
vice (NIH grants CA158191 and GM065307), a Harriet A. Harlin
Professorship, and the University of Illinois Foundation/Oldfield
Research Fund. The authors thank Dr. Robert Schnell and Profes-
sor Gunter Schneider for providing the P. aeruginosa FPPS expres-
sion system and Professor Douglas Mitchell for providing the bac-
teria.
Synergy/antagonism assays: To investigate possible synergistic in-
teractions between compound 11 and fosmidomycin as well as
compound 32 and a range of antibiotics, we carried out two-drug
combination assays. Bacteria were incubated with a threefold gra-
dient of antibiotic typically ranging from 40 mgmLꢀ1 to 18 ngmLꢀ1
Keywords: cell wall biosynthesis · drug discovery · Gram-
negative pathogens · membrane proteins · Staphylococcus
aureus
&
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[1] Antibiotic Resistance Threats in the United States, 2013, US Department
of Health and Human Services, Centers for Disease Control and Preven-
tion (CDC): http://www.cdc.gov/drugresistance/pdf/ar-threats-2013-
508.pdf.
[2] a) E. Oldfield, X. Feng, Trends Pharmacol. Sci. 2014, 35, 664–674; b) E. D.
Brown, G. D. Wright, Nature 2016, 529, 336–343.
[13] S. Mukherjee, C. Huang, F. Guerra, K. Wang, E. Oldfield, J. Am. Chem.
Soc. 2009, 131, 8374–8375.
[14] a) G. M. Eliopoulos, R. C. Moellering in Antibiotics in Laboratory Medicine
(Ed.: V. Lorian), Williams & Wilkins Publishing Co., 1998, pp. 330–396;
b) P. K. Singh, B. F. Tack, P. B. McCray, M. J. Welsh, Am. J. Physiol. Lung
Cell Mol. Physiol. 2000, 279, L799–805.
[15] European Committee for Antimicrobial Susceptibility Testing (EUCAST)
of the European Society of Clinical Microbiology and Infectious Dieases
(ESCMID), Clin. Microbiol. Infect. 2000, 6, 503–508.
[16] M. C. Berenbaum, Pharmacol. Rev. 1989, 41, 93–141.
[17] a) J. E. Fisher, M. J. Rogers, J. M. Halasy, S. P. Luckman, D. E. Hughes, P. J.
Masrachia, G. Wesolowski, R. G. Russell, G. A. Rodan, A. A. Reszka, Proc.
Natl. Acad. Sci. USA 1999, 96, 133–138; b) E. van Beek, C. Lçwik, G.
van der Pluijm, S. Papapoulos, J. Bone Miner. Res. 1999, 14, 722–729.
[18] J. W. Schmidberger, R. Schnell, G. Schneider, Acta. Crystallogr. Sect. D
2015, 71, 721–731.
[19] D. H. Flint, A. Nudelman, Bioorg. Chem. 1993, 21, 367–385.
[20] M. F. Hsu, T. F. Yu, C. C. Chou, H. Y. Fu, C. S. Yang, A. H. Wang, PLoS One
2013, 8, e56363.
[21] a) M. D. Hartley, A. Larkin, B. Imperiali, Bioorg. Med. Chem. 2008, 16,
5149–5156; b) M. R. Webb, Proc. Natl. Acad. Sci. USA 1992, 89, 4884–
4887.
[22] H. Y. Chang, C. C. Chou, M. F. Hsu, A. H. Wang, J. Biol. Chem. 2014, 289,
18719–18735.
[23] A. Wissner, Synthesis 1979, 1, 27–28.
[24] L. Maier, H. Sporri, Phosphorus Sulfur Silicon Relat. Elem. 1992, 70, 39–
48.
[25] A. A. Baykov, O. A. Evtushenko, S. M. Avaeva, Anal. Biochem. 1988, 171,
266–270.
[3] A. Leon, L. Liu, Y. Yang, M. P. Hudock, P. Hall, F. Yin, D. Studer, K. J. Puan,
C. T. Morita, E. Oldfield, J. Med. Chem. 2006, 49, 7331–7341.
[4] H. Jomaa, J. Wiesner, S. Sanderbrand, B. Altincicek, C. Weidemeyer, M.
Hintz, I. Turbachova, M. Eberl, J. Zeidler, H. K. Lichtenthaler, D. Soldati, E.
Beck, Science 1999, 285, 1573–1576.
[5] W. Zhu, Y. Zhang, W. Sinko, M. E. Hensler, J. Olson, K. J. Molohon, S. Lin-
dert, R. Cao, K. Li, K. Wang, Proc. Natl. Acad. Sci. USA 2013, 110, 123–
128.
[6] M. A. Farha, T. L. Czarny, C. L. Myers, L. J. Worrall, S. French, D. G. Conra-
dy, Y. Wang, E. Oldfield, N. C. Strynadka, E. D. Brown, Proc. Natl. Acad.
Sci. USA 2015, 112, 11048–11053.
[7] X. Feng, W. Zhu, L. A. Shurig-Briccio, S. Linder, C. Shoen, R. Hitchings, J.
Li, Y. Wang, N. Baig, T. Zhou, B. K. Kim, D. C. Cric, Proc. Natl. Acad. Sci.
USA 2015, 112, E7073–E7082.
[8] L. Widler, K. A. Jaeggi, M. Glatt, K. Muller, R. Bachmann, M. Bisping, A.-R.
Born, R. Cortesi, G. Guiglia, H. Jeker, et al., J. Med. Chem. 2002, 45,
3721–3738.
[9] R. Cao, C. K. Chen, R. T. Guo, A. H. Wang, E. Oldfield, Proteins Struct.
Funct. Bioinf. 2008, 73, 431–439.
[10] P. D. Ziniel, J. Desai, C. L. Cass, C. Gatto, E. Oldfield, D. L. Williams, Anti-
microb. Agents Chemother. 2013, 57, 5969–5976.
[11] Y. Song, J. M. Chan, Z. Tovian, A. Secrest, E. Nagy, K. Krysiak, K. Bergan,
M. A. Parniak, E. Oldfield, Bioorg. Med. Chem. 2008, 16, 8959–8967.
[12] Y. Xia, Y. L. Liu, Y. Zie, W. Zhu, F. Guerra, S. Shen, N. Yeddule, W. Fischer,
W. Low, X. Zhou, Y. Zhang, E. Oldfield, I. M. Verma, Sci. Transl. Med.
2014, 6, 263ra161.
Received: July 7, 2016
Published online on && &&, 0000
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
11
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
J. Desai, Y. Wang, K. Wang, S. R. Malwal,
E. Oldfield*
Against the wall: Bisphosphonates in-
hibit the growth of Gram-negative bac-
teria (Acinetobacter baumannii, Klebsiella
pneumoniae, Escherichia coli, and Pseu-
domonas aeruginosa) at ~1–4 mgmLꢀ1
levels, targeting farnesyl diphosphate
synthase, whereas monophosphonates
inhibit Gram-positive bacteria (Staphylo-
coccus aureus and Bacillus subtilis) by
targeting undecaprenyl diphosphate
synthase and undecaprenyl diphosphate
phosphatase.
&& – &&
Isoprenoid Biosynthesis Inhibitors
Targeting Bacterial Cell Growth
&
ChemMedChem 2016, 11, 1 – 12
www.chemmedchem.org
12
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!