Synthesis and characterization of novel phosphonocarboxylate inhibitors of RGGT

Article abstract of DOI:10.1016/j.ejmech.2014.06.062

Phosphonocarboxylate (PC) analogs of the anti-osteoporotic drugs, bisphosphonates, represent the first class of selective inhibitors of Rab geranylgeranyl transferase (RabGGTase, RGGT), an enzyme implicated in several diseases including ovarian, breast and skin cancer. Here we present the synthesis and biological characterization of an extended set of this class of compounds, including lipophilic derivatives of the known RGGT inhibitors. From this new panel of PCs, we have identified an inhibitor of RGGT that is of similar potency as the most active published phosphonocarboxylate, but of higher selectivity towards prenyl pyrophosphate synthases. New insights into structural requirements are also presented, showing that only PC analogs of the most potent 3rd generation bisphosphonates inhibit RGGT. In addition, the first phosphonocarboxylate-derived GGPPS weak inhibitor is reported.

Full text of DOI:10.1016/j.ejmech.2014.06.062

European Journal of Medicinal Chemistry 84 (2014) 77e89
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis and characterization of novel phosphonocarboxylate
inhibitors of RGGT
,
,
,
Fraser Coxon b 1, Łukasz Joachimiak a 1, Arafath Kaja Najumudeen c 1, George Breen b,
Joanna Gmach a, Christina Oetken-Lindholm c, Rebecca Way b, James Dunford d,
Daniel Abankwa , Katarzyna M. Błazewska
c
a, *
_
a
_
ꢀ ꢀ
Institute of Organic Chemistry, Lodz University of Technology, Zeromskiego 116, 90-924 Łodz, Poland
b Musculoskeletal Programme, Institute of Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen AB252ZD, UK
c
€
Turku Centre for Biotechnology, Åbo Akademi University, Tykistokatu 6B, 20520 Turku, Finland
d University of Oxford, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, The Botnar Research Center, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Phosphonocarboxylate (PC) analogs of the anti-osteoporotic drugs, bisphosphonates, represent the first
class of selective inhibitors of Rab geranylgeranyl transferase (RabGGTase, RGGT), an enzyme implicated
in several diseases including ovarian, breast and skin cancer. Here we present the synthesis and bio-
logical characterization of an extended set of this class of compounds, including lipophilic derivatives of
the known RGGT inhibitors. From this new panel of PCs, we have identified an inhibitor of RGGT that is of
similar potency as the most active published phosphonocarboxylate, but of higher selectivity towards
prenyl pyrophosphate synthases. New insights into structural requirements are also presented, showing
that only PC analogs of the most potent 3rd generation bisphosphonates inhibit RGGT. In addition, the
first phosphonocarboxylate-derived GGPPS weak inhibitor is reported.
Received 21 February 2014
Received in revised form
14 June 2014
Accepted 27 June 2014
Available online 28 June 2014
Keywords:
Phosphonocarboxylates
Bisphosphonates
Geranylgeranylation
RGGT
© 2014 Elsevier Masson SAS. All rights reserved.
GGPPS
1. Introduction
osteoclast-mediated bone resorption by targeting the mevalonate
pathway enzyme farnesyl pyrophosphate synthase (FPPS) (Fig. 1)
Compounds bearing a phosphonocarboxylate (PC) scaffold, in
which phosphonate and carboxylate groups are linked via one
carbon atom have received considerable attention as potential
antiviral [1e3] and anticancer agents [4,5]. They are therefore
structurally similar to pyrophosphate, an important endogenous
regulator of mineralization, but unlike pyrophosphate they are
resistant to hydrolysis.
The interest in nitrogen-containing PCs was initiated during
structureeactivity relationship (SAR) studies of the well estab-
lished nitrogen-containing bisphosphonates (BP), which inhibit
[6].
The first PC (1c, 3-PEHPC, NE-10790), derived from bisphosph-
onate 2, in which one phosphonic group was exchanged for a car-
boxylic group, has no effect on FPPS, but was found to be the first
selective inhibitor of RGGT [7], inhibiting osteoclast activity in vitro
[7], in vivo [8] and exhibiting anti-cancer properties both in vitro
and in an animal model [9,10]. Comparable potency was observed
for analogs 1aeb [4]. Much higher potency against RGGT was
achieved with the subsequent generation of PCs, compounds 3aec,
derived from BP 4 (Fig. 2) [11,12].
RGGT is responsible for geranylgeranylation of Rab proteins
(Fig. 1), the largest family of small GTPases. Both RGGT and Rab
proteins have recently been implicated in numerous diseases
including cancer, neurological disorders, bacterial and viral in-
fections [13e15]. To date a few classes of RGGT inhibitors have been
identified, including PC derivatives of BPs [4,7,11,12], tripeptide
analogues [16,17], compounds derived from GGTase 1 inhibitors,
with pentasubstituted pyrrolidine analogs [18], compounds
Abbreviations: PC, phosphonocarboxylate; BP, bisphosphonate; RGGT, Rab ger-
anylgeranyl transferase, Rab GGTase, GGTase 2; TAG tunnel, a tunnel adjacent to
GGPP binding site in RGGT; FTase, farnesyl transferase; FPPS, farnesyl pyrophos-
phate synthase, farnesyl diphosphate synthase; GGTase 1, geranylgeranyl trans-
ferase 1; NFSI, N-fluoro-N-(phenylsulfonyl) benzenesulfonamide.
* Corresponding author.
_
E-mail address: katarzyna.blazewska@p.lodz.pl (K.M. Błazewska).
These authors contributed equally.
1
http://dx.doi.org/10.1016/j.ejmech.2014.06.062
0223-5234/© 2014 Elsevier Masson SAS. All rights reserved.

78
F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
Fig. 1. Mevalonate pathway and transferases responsible for prenylation of selected proteins, including the sites of action of the most potent BPs and PCs (enzymes are shown in
blue, inhibitors are shown in red). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2. Structures of the first PC inhibitors of RGGT (1 and 3) and the BPs (2 and 4) from which they were derived.
derived from FTase inhibitor, based on a tetrahydrobenzodiazepine
scaffold [19,20] and triazole-based BPs [21].
potent, clinically used BPs. Until now, only analogs bearing a het-
erocyclic amine side chain have been tested (Fig. 2: 1aec, 3aec)
[4,7,11,12].
The lack of a crystal structure of the RGGTePC complex limits
the possibilities for the rational design of more potent PC inhibitors
and therefore further elaboration of the pharmacophore model by
synthesizing and evaluating the biological activity of new analogs is
required.2 It has been proposed that PCs interact with the TAG
tunnel of RGGT, a tunnel adjacent to the GGPP binding site that is
not present in the other prenyl transferases, FTase and GGT-1,
making it a promising site for selective targeting of RGGT [17].
The TAG tunnel is thought to accommodate the first GG-cysteine
prior to addition of the second geranylgeranyl group, explaining
why the PCs inhibit only the second geranylgeranylation step [22].
We based our studies on the trend observed for the PCeBP pairs
that have already been investigated (1e2 and 3e4), which implied
a possible correlation between structureeactivity relationships for
inhibition of FPPS by BPs and RGGT by PCs with respect to the
nitrogen-containing group.
The second group of compounds that we generated were ana-
logs of PCs with reported potency against RGGT (1c, 3c), in which
the hydroxyl group was exchanged for an alkyl substituent (Fig. 4).
The influence of a hydrophobic substituent that mimics the ger-
anylgeranyl pyrophosphate (GGPP) substrate of RGGT on PC activity
has not been studied before. We synthesized five analogs of 3-
PEHPC (1deh), modified with alkyl chains of different lengths
and one analog of 3-IPEHPC (3g). We expected that an increase in
the hydrophobicity of the PCs might result in increased activity due
to the inhibitor fitting into the GGPP binding pocket and/or
improved cell permeability [23].
2. Results
2.1. Chemistry
Therefore, we focused on the synthesis of 5, the PC analog of the
potent 3rd generation heterocyclic BP, zoledronic acid 9, and PC
analogs (6e8) of the 2nd generation aminoalkyl BPs, pamidronic
10, alendronic 11 and ibandronic 12 acids (Fig. 3), which are all
A variety of approaches to the synthesis of PCs have been
described. They include formation of CeP bond, via introduction of
phosphonic group either in the ArbuzoveMichaelis reaction of
trialkyl phosphite with
enolate and chlorodialkyl phosphite [25], or reaction of diethyl
phosphite with -ketoester [11]. In other methods the carboxylic
moiety is introduced into the carbon adjacent to the phosphonate
a-bromoesters [24], or in a reaction of
2
a
Even though the crystal structures of RGGT with other inhibitors were solved,
attempts to dock newly designed PCs into enzyme did not give coherent results.

F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
79
Fig. 3. Structures of novel PC analogs (5e8) derived from the second and third generation BPs (9e12) (as discussed in the Results, structures of PC derivatives containing hydroxy
substituent (X ¼ OH, 5e8c) were not synthesized).
Fig. 4. Structures of analogs of known PC inhibitors of RGGT, modified in position C-
a
with lipophilic chains (in circles, marked in green) of different lengths. (For interpre-
tation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)
group using lithium alkylphosphonate and diethyl carbonate [26]
or CO2 [27]. Yet another approach is based on alkylation of tri-
alkyl phosphonoacetate [27,28] or functionalization of trialkyl 2-
phosphonoacrylate via Michael-type addition [5,29,30]. Up to
now few examples of nitrogen-containing PCs have been generated
[4,11,12].
Scheme 1. Synthesis of PC analogs of zoledronic acid. Conditions and reagents: (i)
imidazole (1.02 eq), CHCl3, 10 min, rt; (ii) 12 M HCl, reflux; (iii) NaH (1.7 eq), Selectfluor
(1.5 eq), THF, 0 ꢀC to rt, 1.25 h; (iv) 12 M HCl, reflux.
Our first attempts were directed at the synthesis of PCs with the
hydroxyl group retained in the molecule (Fig. 3, compounds 5e8c).
We considered two routes for construction of such derivatives,
The target compounds' structural diversity required separate
approaches to synthesise precursory esters. Subsequent fluorina-
tion of esters was carried out using either Selectfluor, according to a
procedure used previously for fluorination of other PCs [4,12,39], or
using N-fluoro-N-(phenylsulfonyl) benzenesulfonamide (NFSI),
according to a procedure applied for fluorination of selected BPs
[40]. Fluorination yields were affected by the susceptibility of the
products to decomposition, as a result of PeC bond cleavage under
basic conditions [32,39]. Free acids were obtained upon hydrolysis
using 12 M HCl under reflux.
either reaction of dialkyl phosphite with appropriate a-ketoester
(an approach successfully applied in the synthesis of 1c and 3c)
[11,31] or by opening an epoxide ring with an appropriate nucle-
ophile in easily available ethyl 2-(diethoxyphosphoryl)oxirane-2-
carboxylate [32]. Both approaches failed; the first route due to
problems with obtaining precursory a-ketoesters, the second route
due to phosphonateephosphate rearrangement [33], that could not
be suppressed.
Triester 14a was obtained via Michael-type addition of imid-
azole to easily available 13 [29,32,41] (Scheme 1). The reaction was
immediate (a few minutes) and quantitative. Due to the com-
pound's instability (see Supporting information for details), we
immediately subjected it to hydrolysis (98%) or fluorination (47%).
Precursors of 6a and 7a were obtained by applying the Arbuzov
BPs and PCs have different SARs regarding the presence of the
hydroxyl group. Exchange of the hydroxyl for a hydrogen or
halogen atom in BPs leads to a significant decrease of activity
against FPPS [4]. By contrast, the analogous switch in PCs has
smaller effect on the ability to inhibit Rab prenylation [4,12]. This
characteristic was crucial for our efforts, as it enabled us to change
the target PCs from hydroxy- (5e8c) into desoxy- (5e8b) and flu-
oro- (5e8c) analogs. The fluorine atom is considered bioisosteric to
CeH, C]O and CeOH groups [34] and its role in molecules of
pharmacological use is gaining in importance [35e37].3
reaction (Scheme 2). Appropriate
a-bromoesters 17 and 18 were
synthesized from - or -amino acids 15 and 16. N-Phthalyl prod-
g
d
ucts were brominated in the alpha position [42], and subjected to
reaction with triethyl phosphite, producing triester precursors 19a
and 20a with 82e100% yields. Fluorination of esters gave products
19b and 20b with 49% yields. Free acids 6aeb and 7aeb were ob-
tained upon acidic hydrolysis with 66e85% yields.
3
Compound 7a could be also obtained by alkylation of 21 with 22
(Scheme 3). However, when t-BuOK was used as a base, conversion
Also, analogs bearing hydroxyl group may reversibly form dimer complex with
boron on exposure with borosilicate glassware [38].

80
F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
Scheme 4. Synthesis of PC analogs of ibandronic acid. Conditions and reagents: (i) LDA
(2.2 eq), THF, hexane, ꢁ78 ꢀC, then (EtO)2CO (1 eq), to rt, 20 h; (ii) 80% AcOH, 60 ꢀC,
3 h; (iii) N-methyl-N-penthylamine (0.9 eq), NaBH(OAc)3 (0.9 eq), DCM, 4 h, rt; (iv)
1.6 M BuLi, NFSI; (v) 12 M HCl, reflux.
Scheme 2. Synthesis of PC analogs of pamidronic and alendronic acids. Conditions and
reagents: (i) phthalic anhydride (1.02 eq), 180 ꢀC, 35 min; (ii) SOCl2 (5 eq), CCl4, 1.25 h
reflux, then NBS (1.2 eq) with HBr(aq) (cat), 1.1 h, reflux, then EtOH; (iii) P(OEt)3
(1.92 eq), 165 ꢀC, 4 h; (iv) NaH (1.7 eq), Selectfluor (1.5 eq), THF, 0 ꢀC to rt, 1.25 h; (v)
12 M HCl, reflux.
benzenesulfonamide (NFSI) [40] led to the fluorination product 29b
with 50% yield. The free acid 8b was obtained with 90% purity.
A second group of compounds, modified with a lipophilic chain
(Scheme 5), was synthesized by the alkylation reaction of 30 with
picolyl chloride or with chloro-2-methyl-imidazo[1,2-a]pyridine.5
Synthesis of the
a-alkyl phosphonoacetate 30 was achieved
using the Arbuzov reaction [24], giving compounds in gram
quantities and 61e71% yields. Alternatively, alkylation of 21 with
alkyl bromide can be applied, but this approach invariably gives a
mixture of 30, substrate and the product of dialkylation, impeding
product isolation.
The analogs 30 thus obtained were subjected to alkylation with
picolyl chloride (Scheme 5), using conditions reported previously
for alkylation of triethyl phosphonoacetate with benzyl-like chlo-
rides [4,12] and a-alkyl phosphonoacetate with allyl bromide [26],
giving pure products in 19e56% after column chromatography.
Optimization efforts did not improve this process. Free acids 1 were
obtained upon hydrolysis of 31 with 32e72% yields.
Synthesis of
a-alkylated analogs of 3-IPEHPC proved more
challenging. In the alkylation reaction, we either observed 30geh
recovery or the expected triester 32geh in low yields 11e14%. The
free acid 3g was synthesized with 93% purity.
These novel PCs were evaluated for their ability to inhibit RGGT,
other prenyl transferases and/or mevalonate pathway enzymes in
intact cells, which has the advantage over isolated enzyme assays of
assessing the cellular availability of the compounds at the same
time. Two complementary and independent techniques were used.
Scheme 3. Synthesis of PC analogs of pamidronic acid: second method. Conditions and
reagents: (i) t-BuOK (2.2 eq), THF, 5 d, rt; (ii) 3 M HCl (g)/EtOH, 1 h 40 min, rt; (iii) 12 M
HCl, reflux; (iv) NaH (1.7 eq), Selectfluor (1.5 eq), THF, 1.25 h, 0 ꢀC to rt.
did not exceed 70%. Higher conversion was achieved under
different conditions (using different bases, solvents and PTC cata-
lysts), but then it was compromised by lower selectivity, yielding
also the product of dialkylation 24.4 Triester 23a was fluorinated
with 49% yield.
2.2. Biological evaluations
The synthesis of 8a was based on introducing a carboxyester
group to the phosphonate 26 (Scheme 4) [26]. Subsequently, the
acetal moiety was hydrolysed to produce aldehyde 28 which was
subjected to reductive amination with N-methyl-N-penthylamine
[44]. Hydrolysis of 29a led to the expected product 8a. Fluorination
of 29a with Selectfluor gave a mixture of unidentified products.
Efforts to overcome this limitation by fluorination of precursory 27
or 28 failed. Application of N-fluoro-N-(phenylsulfonyl)
2.2.1. Biological activity testing using Rab-NANOPS
In order to test the biological activity of the compounds, we used
a recently described cell-based FRET-assay that allows for the
detection of functional membrane anchorage of lipid-modified
proteins [45,46]. Three FRET-biosensors that report on steady-
state protein farnesylation (Ras-NANOPS), as well as RGGT-
dependent geranylgeranylation of dual- (Rab5-NANOPS) and
mono-prenylated Rab proteins (Rab8-NANOPS) were first used to
determine the potency of compounds at 500 mM.
We found that three compounds, 5a, 5b and 1h significantly
reduced the FRET-signal of Rab5-NANOPS (Fig. 5A), but not of Ras-
4
Synthesis of the compound 23a was reported in patent [43], using NaH in THF.
No product of dialkylation was reported. In contrast, according to our studies
conditions reported therein were giving mixture of starting compound with 23a
and dialkylation product 24.
5
The introduction of alkyl chain into scaffold containing heterocyclic ring, was
expected to be inferior, due to the possibility of forming N-alkylation side product.

F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
81
Scheme 5. Synthesis of PC analogs of selected BPs modified with alkyl chain in C-a position. Conditions and reagents: (i) picolyl chloride (1 eq), NaH (3.15 eq), DMF/THF; (ii) 12 M
HCl, reflux.
Fig. 5. Cell-based activity testing of PC analogs using FRET-biosensors. (A) BHK21 were transfected with Rab5-NANOPS and screened with the indicated PC analogs at a final
concentration of 500 M for 24 h. The Emax values describe the cellular FRET-response of the indicated inhibitors as described in Experimental section. A decrease in the FRET
m
reports on loss of membrane anchorage, cellular redistribution or perturbed nanoclustering of the biosensors. Dotted line indicates the average Emax of the non-treated control. Data
are presented as mean Emax s.e.m (n ¼ 3): (*) p < 0.05 (***) p < 0.001 indicate statistical significance vs. non-treated control using one-way ANOVA followed by Dunnett's test as
described in Experimental Section. (B and C) FRET-response data like in (A) of Ras- and Rab8-NANOPS, respectively. Treatment with compactin (5 mM), a HMG-CoA inhibitor that
blocks synthesis of isoprenoids and therefore protein prenylation, serves as a positive control and reduces the Emax value significantly in AeC. (D) Doseeresponse curves of the effect
of 5ae5b, 1c and 1h on the Emax values of Rab5-NANOPS expressed in BHK cells (n ¼ 6).
Table 1
Effect of active PC compounds on prenylation of Rab11 and Rap1a, cell viability in Hela cells, response of FRET-NANOPS. LED ¼ lowest effective dose for inhibition of
prenylation.
Compound
Reduction of Hela
Inhibition of Rab11
Inhibition of Rap1A
FRET-response of NANOPSc
cell viability (IC50
/m
M)
prenylation (LED/
m
M)
prenylation (LED/mM)
Rab5a (IC50
/m
M)
Rab21 (IC50
No effectb
1060 500
980 180
253 45
/m
M)
Rab8a (IC50
/m
M)
Ras (IC50/mM)
8a
1c
1h
5a
5b
3c
4
No effect
>2000
No effect
650
850
500
1000
250
500
25
10
25
1000
No effect
500
No effect
No effect
400a
No effectb
630 300
360 60
360 40
No effectb
No effectb
No effectb
No effectb
No effectb
n.d.
No effectb
No effectb
No effectb
No effectb
No effectb
n.d.
40
3
72
1
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
1
2.5
0.25
0.25
0.25
0.25
n.d.
n.d.
n.d.
n.d.
9
a
LED for inhibition of Rap1A prenylation in J774 cells [11].
b
c
No effect at concentrations 500 and 1000
mM.
Detection specificity: Rab5a and Rab21: dual prenylation and inhibition of RGGT; Rab8a: mono prenylation and inhibition of RGGT; Ras: inhibition of farnesylation and/or
palmitoylation.

82
F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
expressed constitutively active Rab5aQ79L that was N-terminally
tagged with the fluorescent protein mCitrine in BHK21 cells and
monitored its localization in the cell using confocal microscopy.
This Rab5-construct localized to well recognizable hyperfused
endosomes (Fig. 6A). However, treatment with compactin led to a
strong cytoplasmic redistribution (Fig. 6B). Both 5a and 5b
completely relocalized the Rab5-construct to the cytoplasm,
consistent with a potent inhibition of its prenylation (Fig. 6C, D). On
the other hand, 1h leads to a redistribution of the construct to
perinuclear structures and more punctate labelling of endosomes
(Fig. 6E).
2.2.2. Biological testing in Hela cells
The novel PCs were also assessed for their ability to inhibit RGGT
and/or mevalonate pathway enzymes in intact cells by assaying for
the unprenylated forms of Rab11 and Rap1A, which are modified
with geranylgeranyl groups by RGGT and GGTase 1, respectively;
inhibition of Rab11 prenylation alone is indicative of a specific
RGGT inhibitor, while inhibition of prenylation of both GTPases at
similar concentrations is indicative of a GGPPS or FPPS inhibitor
(see Fig. 1). In addition, we assessed the ability of these PCs to
reduce the viable cell number in cultures of Hela cells as a separate
indicator of inhibition of prenylation, since inhibition of prenyl
transferases or mevalonate pathway enzymes is known to reduce
cell viability [7,47e49].
As expected, the first reported PC inhibitor of RGGT, 1c, weakly
inhibited prenylation of Rab11, while bisphosphonate 9 inhibited
prenylation of both Rab11 and Rap1A, consistent with its known
ability to inhibit FPPS (see Fig. 1) [7,47]. Consistent with the FRET
assay, among the first group of synthesized compounds (Fig. 3)
compounds 5aeb showed inhibition of RGGT (Fig. 7).
PC analogues of pamidronic and alendronic acids, 6aeb or 7aeb,
did not inhibit prenylation of Rab11 or Rap1A in Hela cells at con-
centrations up to 1 mM, indicating that they do not inhibit RGGT or
enzymes of the mevalonate pathway that are involved in pre-
nylation (Fig. 8). Accordingly, only 5aeb analogs reduce Hela cell
viability, correlating well with the effects on Rab11 prenylation
(Fig. 9C, Table 1), while compounds 6aeb and 7aeb had no effect
on the viable cell number of Hela cells after 72 h of treatment with
up to 2 mM (Fig. 9A).
Fig. 6. PC analogs mislocalize Rab5aQ79L in BHK cells. BHK cells expressing mCitrine
tagged Rab5aQ79L (i.e. the acceptor of the Rab5-NANOPS FRET-pair) were treated with
(A) PBS (control), (B) compactin (5 mM), (C) 5b, (D) 5a, (E) 1h (all three at 500 mM).
Following treatment, cells were fixed and imaged using a confocal microscope. Shown
are representative images from 2 independent experiments. Scale bars are represen-
tative for all images and corresponds to 20 mm.
The similarity in potency between 5aeb and 3c for reducing
Hela cell viability suggests that the possible weak inhibition of
GGPPS by 3-IPEHPC (3c) [11] does not significantly contribute to its
cellular toxicity, despite the evidence that inhibition of prenylation
of other geranylgeranylated proteins (i.e. those modified by GGTase
I) has a more profound effect on cell viability than RGGT inhibition
[7,50]. In agreement with this, weak inhibition of GGPPS by 1h
(discussed later) also had no effect on the viability of Hela cells
(Fig. 9B). The correlation between the inhibition of Rab prenylation
NANOPS or Rab8-NANOPS (Fig. 5B, C). Analysis with an additional
Rab-biosensor for dual-geranylgeranylated Rab proteins, Rab21-
NANOPS, confirmed exactly the activity of the identified three
compounds (Supporting information: Fig. S108). Dose response
analysis with Rab5-NANOPS revealed IC50 e values of 360 40 (for
5a), 40 3 (for 5b) and 360 60 mM (for 1h) (Fig. 5D, Table 1).
We next tested what effect these compounds have on the sub-
cellular distribution of the dual-prenylated Rab5. To this end we
Fig. 7. Effect of zoledronic acid analogues 5aeb on protein prenylation in Hela cells. Hela cells were treated for 48 h with the indicated concentration of PCs (mM); (on the left) or
with 0.5 mM 5a, 5b or 1 mM 1h (on the right). Cells were then lysed and the prenylated and unprenylated proteins separated using triton X-114 fractionation. Aqueous phases
containing unprenylated proteins were separated by electrophoresis and western blotted for Rab11, Rap1A or actin.

F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
83
Fig. 8. Effect of desoxy- and fluoro-analogues of alkyl PCs on protein prenylation in Hela cells. Hela cells were treated for 48 h with the indicated PCs (1 mM unless otherwise
stated). Cells were then lysed and the prenylated and unprenylated proteins separated using triton X-114 fractionation. Aqueous phases containing unprenylated proteins were
separated by electrophoresis and western blotted for Rab11, Rap1A or actin.
and reduction of Hela cell viability also indicates that the effects of
these novel compounds on viability are due to inhibition of Rab
prenylation, and not due to other unrelated cellular effects. Similar
results for these compounds were found in RAW264 macrophages,
demonstrating that the effects are not cell type specific (data not
shown).
Compound 8a, the PC analog of another 2nd generation
bisphosphonate, ibandronic acid (12), inhibited prenylation of both
Rab11 and Rap1A at 1 mM (Table 1). This suggests that rather than
inhibiting RGGT, this compound may inhibit an enzyme of the
mevalonate pathway such as GGPPS or FPPS (see Fig. 1). However,
the low potency of this compound is reflected in Hela cell viability
assays, in which it had no effect at up to 2 mM (Fig. 9A).
Several analogues of 3-PEHPC in which the hydroxyl group was
exchanged for an alkyl substituent were tested for biological ac-
tivity. As in the FRET assay the shorter chain analogues, (C1: 1d, C3:
1e and C6: 1g) were all inactive at up to 1 mM (both Rab11 and
Rap1A prenylation). However, the octyl (C8) analogue 1h, inhibited
both Rab11 and Rap1A prenylation at 1 mM, but not at 0.25 mM or
lower concentrations (Fig. 10, Table 1). This data suggests that,
similar to 8a, this novel PC does not inhibit RGGT but inhibits either
GGPPS or FPPS. Alternatively, it is possible that both RGGT and
GGTase I could be inhibited at similar concentrations by both of
these PCs.
To exclude the latter possibility, we determined the ability of
replenishing cells with GGPP to overcome the inhibition of pre-
nylation; since GGPP is downstream of GGPPS/FPPS it should rescue
the effect of inhibitors of these enzymes, but not the effect of RGGT
or GGTase I inhibitors (see Fig. 1) [11,51]. Indeed, we found that
neither inhibition of Rab11 prenylation by 3-PEHPC (a RGGT in-
hibitor) nor inhibition of Rap1A prenylation by GGTI-298 (a GGTase
I inhibitor) could be rescued by GGPP (Fig. 11A, B). By contrast,
inhibition of prenylation of both these proteins by zoledronic acid 9
(an FPPS inhibitor), 1h or 8a could be completely prevented by
GGPP, indicating that these PCs are possibly GGPPS or FPPS in-
hibitors. In isolated enzyme assays, we found that 1h dose-
dependently inhibited GGPPS activity (IC50 127 mM), however 8a
had no effect at up to 1 mM (Fig. 11C), implying that it likely targets
FPPS instead. At high concentrations, 3-IPEHPC (3c) also appears to
inhibit GGPPS, although this compound is also a very potent in-
hibitor of RGGT [11].
3. Discussion
The cellular assays identified two PC analogs as potent and
specific inhibitors of RGGT (5b > 5a) (Table 1). 5b was as potent as
3-IPEHPC (3c), which is the most potent PC inhibitor of RGGT yet
identified. However, 5b is more selective, since at high
Fig. 9. Effect of PCs on Hela cell viability. Hela cells were treated with PCs for 72 h, then
the viable cell number was determined by the Alamar blue assay. Data are the mean
from 4 independent experiments.

84
F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
Fig. 10. Effects of alkyl chain length in 3-PEHPC (1d, e, g, h) on protein prenylation in Hela cells. Hela cells were treated for 48 h with the indicated PCs (concs in mM; 1 mM unless
otherwise stated). Cells were then lysed and the prenylated and unprenylated proteins separated using triton X-114 fractionation. Aqueous phases containing unprenylated proteins
were separated by electrophoresis and western blotted for Rab11, Rap1A or actin.
Fig. 11. (A, B) 1h and 8a likely inhibit GGPPS in Hela cells. Hela cells were treated with PCs or 10 mM GGTI-298 for 48 h in the presence or absence of 15 mM GGPP. Cells were then
lysed and the prenylated and unprenylated proteins separated using triton X-114 fractionation. Aqueous phases containing unprenylated proteins were separated by electrophoresis
and western blotted for Rab11, Rap1A or actin. (C) The effect of 5a, 8a, 1h on activity of the GGPPS in an isolated enzyme assay. 40 nM GGPPS was preincubated for 10 min with
compounds before initiation of the reaction by addition of substrate. Data represents the mean SEM (n ¼ 3). Compound 1h is a weak inhibitor of GGPPS.
concentrations 3c also inhibits Rap1a prenylation, most likely due
to inhibition of GGPPS [11].
other fluoro-PC analogues that we have studied are of similar po-
tency to their parent PCs [4,12].
The cell-based FRET-assays clearly showed that 5a and 5b both
significantly decreased the membrane localization of dual-
geranylgeranylated Rabs, but not of monoprenylated Rab8 or far-
nesylated H-Ras derived polypeptide. This observation is consistent
with previous studies that reported PCs as inhibitors of the second
geranylgeranylation reaction catalysed by RGGT [22].
Increasing hydrophobicity by introducing an alkyl moiety on the
branching carbon atom between the phosphonic and carboxylic
groups, C-a, completely abolished the ability of 3-PEHPC to inhibit
RGGT in Hela cells. This result extends the previously observed
trend [4,12], where the desoxy analogs 1a and 3a, modified with the
smallest possible hydrophobic substituent, a hydrogen atom, were
slightly less potent than other studied compounds (1bec and
3bec). Therefore, the presence of a polar group in this position
seems to be important for inhibition of RGGT by PCs.
Interestingly, 1h, which has a C8 side chain shows weak activity
against an enzyme of the mevalonate pathway, GGPPS, which ac-
commodates a C15-chain substrate and C20-chain product. This
structural trend corresponds well with literature data for alkylated
bisphosphonate-derived GGPPS inhibitors.6 The ibandronic acid
Interestingly, of all the PCs tested here, those that are analogues
of a heterocyclic BP (3rd generation bisphosphonate: 9) appear to
be active RGGT inhibitors (as all previously reported PC inhibitors of
RGGT, 1aec and 3aec), whereas analogues of the alkyl BPs (2nd
generation BPs: 10, 11, 12) appear to be inactive in this respect. Such
difference in activity between these two groups of compounds may
reflect the difference in the nitrogen character. All active derivatives
(1aec, 3aec, 5aeb) have a nitrogen atom in the aromatic ring, at
the same distance from the phosphonic group. Also, the most
potent representatives of PCs (5aeb and 3aec) contain two nitro-
gen atoms within a differentially substituted imidazole ring. The
activity order between 5a (desoxy-analogue) and 5b (fluoro-
analogue) is broadly in agreement with our previous studies of
desoxy and halogen-substituted PC analogues, in which fluoro-
derivatives 1b and 3b were from 2 to 10 times more potent than
their desoxy counterparts [4,12]. It is unclear whether 5b would be
more potent than 5c (a compound that has yet to be synthesized);
6
According to Oldfield [52,53] the activity of monoalkylated hydroxy
bisphosphonates against GGPPS increases with the length of the side chain, with
the C9 chain analogues being the most potent. Of the doubly alkylated analogs,
those with two octyl groups showed significant activity against GGPPS. In addition,
Wiemer [54] reported that a risedronic acid derivative, with the hydroxyl group
exchanged for a geranyl group possessed activity against GGPPS.

F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
85
analogue, 8a, is most likely a weak inhibitor of FPPS in Hela cells,
since no effect on GGPPS activity could be detected.
methanol:ammonia (Sigma) were air-dried to remove most of the
solvent, then adjusted to a concentration of 2 mM using DMEM.
GGTI-298 (Sigma) was dissolved in 10 mM dithiothreitol in DMSO
to give a stock concentration of 10 mM, and stored at ꢁ20 ꢀC.
Based on the above we expected 1h to affect the monoger-
anylgeranylated Rab8-NANOPS biosensor, since deficiency in GGPP
should impair prenylation of both mono- and di-
geranylgeranylated Rabs. However, only a non-significant effect
was observed, which is consistent with the weak potency of the
compound. However, an alternative possibility could be ascribed to
the fact that FPP can also serve as a substrate for RGGT [55,56,59],
and in this case inhibition of GGPPS would cause an increase in FPP,
which might then be utilised for prenylation of Rab8 (and other Rab
proteins).
The inhibition of prenylation was confirmed by the disrupted
subcellular localization of the dual-prenylated Rab5aQ79L. Both 5a,
5b significantly mislocalize Rab5aQ79L to the cytosol, while 1h
leads to mistargeting of Rab5a to perinuclear structures. Complete
cytosol mislocalization induced by 5aeb may be supporting a hy-
pothesis that RGGT and monoprenylated (but dual-cysteine) Rabs
form a dead end complex in the presence of PC, thus depleting cells
of RGGT for prenylation of newly synthesized Rabs [22]. On the
other hand, the Rab5aQ79L-decorated perinuclear structures
detected in cells treated with 1h resemble those seen with C-ter-
minal mono-cysteine mutants of Rab5a [57], which occurs because
monoprenylation of normally diprenylated proteins leads to their
mislocalization to ER [58]. We therefore speculate (as suggested in
the previous paragraph) that when GGPPS is inhibited, Rab8 is
monoprenylated with a farnesyl group by RGGT [59].
5.1. Chemistry: general procedure for acidic hydrolysis
5.1.1. 3-(1H-imidazol-1-yl)-2-phosphonopropanoic acid (5a)
Compound 14a (1.29 g) was dissolved in 12 M HCl (30 mL). After
5 h reflux it was evaporated to dryness and precipitated using ab-
solute ethanol (7 mL). Precipitate was filtered off and rinsed with
minute amounts of ethanol, giving 0.935 g of product. 31P NMR
2
3
(D2O, pH 8): 9,72; 1H NMR: 3.12 (ddd, JHP ¼ 21.40, JHH ¼ 12.10,
3JHH ¼ 3.45, CHP, 1H), 4.40 (ddd, JHH ¼ 13.90, JHP ¼ 5.35,
2
3
2
3
3JHH ¼ 3.35, CHHCHP, 1H), 4.60 (ddd, JHH ¼ 13.90, JHP ¼ 4.15,
3JHH ¼ 12.10, CHHCHP, 1H), 7.13e7.14 (m, CHar, 1H), 7.30e7.31 (m,
CHar, 1H), 7.97 (bs, CHar, 1H); 13C NMR: 47.90 (s, CH2CHP, 1C), 49.34
(d, 1JCP ¼ 112.25, CHP, 1C), 120.33 (s, CHar, 1C), 122.56 (s, CHar, 1C),
2
135.92 (s, CHar, 1C), 172.92 (d, JCP ¼ 5.65, C(O), 1C). Elemental
analysis: C6H9N2O5P Calculated: C 32.74, H 4.12, N 12.73; Found: C
32.89, H 4.37, N 12.57; Yield: 98% (0.92 g).
5.1.2. 2-Fluoro-3-(1H-imidazol-1-yl)-2-phosphonopropanoic acid
(5b)
31P NMR (D2O, pH 5): 8.10 (d, 2JPF ¼ 66.85 Hz); 19F NMR: 169.27
2
(d, JPF ¼ 66.85); 1H NMR: 4.68e4.90 (signal overlapping with
3
3
2
water, CHHN, 1H), 5.01 (ddd, JHF ¼ 32.50, JHP ¼ 3.80, JHH ¼ 15.0,
CHHN, 1H), 7.44e7.49 (m, 2Har), 8.73e8.75 (m, 1Har); 13C NMR:
53.20 (dd, 2JCP ¼ 8.10, 2JCF ¼ 20.20, CH2N, 1C), 96.79 (dd, 1J ¼ 143.95,
196.20 Hz, CFP, 1C), 120.14 (s, Car), 123.61 (s, Car), 136.28 (s, Car),
172.02 (d, 2J ¼ 20.70 Hz, C]O); Elemental analysis: C6H8FN2O5P
Calculated: C 30.27, H 3.39, N 11.76; Found: C 30.15, H 3.20, N 11.38.
Scale: 0.406 g (14b); Yield: 87% (0.26 g).
4. Conclusions
In summary, we have developed simple routes for the synthesis
of novel PC analogs of anti-resorptive bisphosphonates, which were
predicted to extend the library of PC-type inhibitors of RGGT.
Indeed, biological evaluation of the compounds identified two an-
alogs of zoledronic acid, 5a and 5b, to be highly effective RGGT
inhibitors in intact cells, with 5b more potent than 5a and of
comparable potency and higher selectivity to the most potent and
selective PC inhibitor of RGGT reported thus far. Interestingly, a PC
analogue containing an 8-carbon side chain was found to have no
effect on RGGT, but to be a weak inhibitor of GGPPS. Further studies
on the synthesis and biological evaluation of new phosphono-
carboxylate analogs derived from the most promissing RGGT in-
hibitors found here, 5a and 5b, are in progress.
5.1.3. 4-Amino-2-phosphonobutanoic acid (6a)
31P NMR (D2O, pH 8): 18.58; 1H NMR: 1.95e2.25 (m, CH2CH,
2
3
2H), 2.61 (ddd, JHP ¼ 21.40, JHH ¼ 9.70, 4.65, CHP, 1H), 3.02 (bt,
3JHH ¼ 7.90, CH2NH2, 2H); 13C NMR: 24.67 (d, JCP ¼ 3.25, CH2CH,
2
1C), 37.24 (d, 3JCP ¼ 15.90, CH2NH2, 1C), 46.80 (d, 1JCP ¼ 117.40, CHP,
2
1C), 177.21 (d, JCP
¼
2.25, C(O)OEt); Elemental analysis:
C4H10NO5P, Calculated: C 26.24, H 5.50, N 7.65; Found: C 26.14; H
5.58; N 7.62; Scale: 0.38 g (19a); Yield: 79%.
5. Experimental section
NMR spectra were measured at 250.13 or 700 MHz for 1H NMR,
62.90 or 170 MHz for 13C NMR, 101.30 MHz for 31P NMR,
5.1.4. 4-Amino-2-fluoro-2-phosphonobutanoic acid (6b)
2
31P NMR (D2O, pH 7): 10.84 (d, JPF ¼ 75.75 Hz); 19F NMR:
235.31 MHz for 19F NMR. Chemical shifts (
d
) are reported in parts
per million (ppm) relative to: in 1H NMR: internal residual CHCl3 in
CDCl3 ( 7.26), or internal residual HDO in D2O (pH ~ 12, 4.76); in
2
ꢁ167.96 (d, JPF ¼ 75.75); 1H NMR: 2.30e2.63 (m, CH2CF, 2H);
3.02e3.23 (m, CH2NH2, 2H); 13C NMR: 32.56 (d, 2J ¼ 21.20, CH2CF,
1C), 36.40 (dd, 3J
¼
9.95, 4.85, CH2NH2, 1C), 98.20 (dd,
d
d
31P NMR: external 85% H3PO4 (0 ppm); in 13C NMR: CDCl3
(77.00 ppm); in case of solutions in D2O signals from residual sol-
vents were used as reference (EtOH (15.17 and 58.13) or acetone
(30.89 and 215.94 ppm); in 19F NMR: 70% CFCl3 in CDCl3 (0 ppm).
All mass spectra were recorded on the Finnigan MAT 95 double
focussing (BE geometry) mass spectrometer (Finnigan MAT, Bre-
men, Germany). Elemental analyses were performed on a CHNS
Elemental Analyser e Elementar Vario MICRO. For final com-
pounds' (1 and 5e8) precipitation the following solvents were
used: water, ethanol, acetone. Only in case of 3g, diethyl ether was
used.
1JCP,CF ¼ 189.20, 147.95, CPF, 1C), 175.94 (d, 2J ¼ 20.20, C(O)OH, 1C).
Elemental analysis: C4H9FNO5P(C2H5OH)0.43 Calculated: C 26.42, H
5.28; Found: C 26.58, H 5.47; Scale: 0.396 g (19b); Yield: 85%.
5.1.5. 5-Amino-2-phosphonopentanoic acid (7a)
31P NMR (D2O, pH 1): 17.44; 1H NMR: 1.70e2.06 (m, CHCH2CH2,
4H), 2.88 (ddd, 2JHP ¼ 22.35, 3JHH ¼ 10.20, 3JHH ¼ 3.70, CHP, 1H), 3.03
(bt, 3JHH ¼ 7.25, CH2NH2, 2H); 13C NMR: 22.22 (d, J ¼ 3.80, CH2, 1C),
23.96 (d, J ¼ 14.45, CH2, 1C), 37.16 (s, CH2NH2, 1C), 44.96 (d,
1JCP ¼ 120.10, CHP, 1C), 172.92 (d, 2JCP ¼ 4.95, COOH, 1C); Elemental
analysis: C5H12NO5P(HCl)0.8 Calculated: C 26.54, H 5.70, N 6.19;
Found: C 26.68; H 5.78; N 6.30; Yield: 66% (from 20a, scale: 0.30 g)
or 78% (from 23a, scale: 0.276 g).
For prenylation studies, PCs were dissolved in calcium-free PBS
at stock concentrations of 20e50 mM, and the pH adjusted to 7.6.
Compounds were stored at 4 ꢀC prior to use. FPP and GGPP in

86
F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
5.1.6. 5-Amino-2-fluoro-2-phosphonopentanoic acid (7b) (from
23b)
148.12 (s, CHAr (2), 1C), 179.86 (s, C]O, 1C). Elemental analysis:
13H20NO5P Calculated: C 51.83, H 6.69, N 4.65; Found: C 51.58, H
6.66, N 4.54: Scale: 31f (0.2 mmol), Yield: 72%.
C
2
31P NMR (D2O, pH 2): 10.39 (d, JPF ¼ 74.35 Hz); 19F NMR:
2
ꢁ176.96 (d, JPF ¼ 75.10); 1H NMR: 1.61e1.98 (m, CH2CH2, 2H),
2.09e2.45 (m, CH2CH2, 2H), 3.06 (bt, 3JHH ¼ 7.40, CH2NH2, 2H); 13C
NMR: 21.86 (dd, JCP,CF ¼ 10.25, 3.65, CH2CH2, 1C), 30.76 (d, J ¼ 20.40,
5.1.12. 2-Phosphono-2-(pyridin-3-ylmethyl)octanoic acid (1g)
3
31P NMR (D2O, pH 13): 22.31; 1H NMR: 0.72 (t, JHH ¼ 7.2,
1
CH2CH2, 1C), 39.70 (s, CH2NH2, 1C), 97.48 (dd, JCP,CF ¼ 147.95,
CH3(CH2)5, 3H), 0.93e1.22 (m, CH3(CH2)4CH2, 8H), 1.44e1.1.58 (m,
2
3
2
189.65, CFP, 1C), 173.64 (bdd, JCP,CF ¼ 24.05, 2.95, C(O), 1C); Scale:
CH3(CH2)4CH2, 2H), 2.90 (dd, JHP ¼ 9.1, JHH ¼ 14.0, PyCHHC, 1H),
3 2 3
0.17 g (23b); Yield: 73%.
3.52 (dd, JHP ¼ 7.3, JHH ¼ 14.0, PyCHHC, 1H), 7.27 (dd, JHH ¼ 7.9,
4.8, CHAr (5), 1H), 7.75 (dt, 3JHH ¼ 7.9, 4JHH ¼ 1.9, CHAr (4), 1H), 8.25 (d,
3JHH ¼ 4.8, CHAr (6), 1H), 8.36 (s, CHAr (2), 1H); 13C NMR: 11.57 (s,
CH3(CH2)5, 1C), 20.13, 27.94, 29.19 (3s, CH3(CH2)3(CH2)2, 3C), 23.61
5.1.7. 4-(Methyl(pentyl)amino)-2-phosphonobutanoic acid (8a)
31P NMR (D2O, pH 1): 17.99; 1H NMR: 0.86e0.91 (m, CH3(CH2)4,
3H), 1.28e1.36 (m, (CH2)2, 4H), 1.62e2.38 (m, CH2, 2H), 2.79e3.00
(m, CHP, 1H), 2.87 (s, CH3N, 3H), 3.03e3.39 (m, CH2NCH2, 4H); 13C
NMR: 11.20 (s, CH3CH2, 1C), 19.64, 21.34, 26.01 (3s, 3CH2, 3C), 21.65
3
(d, JCP ¼ 5.2, CH3(CH2)3CH2CH2, 1C), 31.25 (s, CH3(CH2)4CH2, 1C),
35.30 (s, PyCH2C, 1C), 53.45 (d, 1JCP ¼ 115.3, C, 1C), 121.67 (s, CHAr
(5),1C), 135.66 (d, 3JCP ¼ 14.9, CAr (3), 1C), 137.15 (s, CHAr (4),1C), 143.85
(s, CHAr (6), 1C), 148.17 (s, CHAr (2), 1C), 180.58 (s, C]O, 1C).
Elemental analysis: C14H22NO5P(C2H5OH)0.19 Calculated: C 53.30,
H 7.20, N 4.32; Found: C 53.36, H 7.12, N 4.32. Scale: 31g (0.3 mmol),
0.12 g; Yield: 36%.
2
(d, JCP ¼ 8.75, PCHCH2, 1C), 37.52 (s, CH3N, 1C), 46.66 (d,
1JCP ¼ 118.59, CHP, 1C), 54.14, 54.38 (s, d, CH2NCH2, 2C), 177.22 (d,
2JCP ¼ 4.60, COOH, 1C); Scale: 0.087 g (29a); Yield: 100%.
5.1.8. 2-Fluoro-4-(methyl(pentyl)amino)-2-phosphonobutanoic
acid (8b)
5.1.13. 2-Phosphono-2-(pyridin-3-ylmethyl)decanoic acid (1h)
2
3
31P NMR (D2O, pH 2): 7.89 (d, JPF ¼ 72.5 Hz); 1H NMR 0.88 (t,
31P NMR (D2O, pH 13): 22.05; 1H NMR: 0.82 (t, JHH ¼ 7.1,
3JHH ¼ 6.8 Hz, CH3CH2CH2, 3H); 1.23e1.38 (m, CH3(CH2)3CH2, 6H);
2.48e2.72 (m, CH2CHP, 2H), 2.87 (s, CH3N, 3H); 2.99e3.53 (m,
CH2NCH2, 4H). Scale: 0.10 g (29b). Yield: 100% (purity > 80%).
CH3(CH2)7, 3H), 0.97e1.22 (m, CH3(CH2)7CH2, 12H), 1.50e1.1.65 (m,
3
2
CH3(CH2)6CH2, 2H), 2.95 (dd, JHP ¼ 9.0, JHH ¼ 14.1, PyCHHC, 1H),
3
2
3
3.58 (dd, JHP ¼ 7.4, JHH ¼ 14.1, PyCHHC, 1H), 7.30 (dd, JHH ¼ 7.8,
3JHH ¼ 4.9, CHAr (5), 1H), 7.69 (d, JHH ¼ 7.8, CHAr (4), 1H), 8.16 (d,
3
5.1.9. 2-Methyl-2-phosphono-3-(pyridin-3-yl)propanoic acid (1d)
31P NMR (D2O, pH 6): 25.59; 1H NMR (700 MHz): 1.12 (d,
3JHP ¼ 14.6, CH3C, 3H), 2.84 (dd, 3JHP ¼ 8.7, 2JHH ¼ 13.4, PyCHHC, 1H),
3JHH ¼ 4.9, CHAr (6), 1H), 8.27 (s, CHAr (2), 1H); 13C NMR: 11.59 (s,
CH3(CH2)7C, 1C), 20.17, 26.59, 26.77, 28.21, 29.33 (5s,
3
CH3(CH2)5(CH2)2, 5C), 23.59 (d, JCP ¼ 5.3, CH3(CH2)5CH2CH2, 1C),
3
2
3
3.48 (dd, JHP ¼ 6.6, JHH ¼ 13.4, PyCHHC, 1H), 7.40 (dd, JHH ¼ 7.9,
4.9, CHAr (5),1H), 7.72 (dt, 3JHH ¼ 7.9, 4JHH ¼ 1.7, CHAr (4),1H), 8.36 (bd,
3JHH ¼ 1.7, CHAr (2), 1H), 8.39 (bdd, 3JHH ¼ 4.9, 4JHH ¼ 1.7, CHAr (6), 1H);
31.16 (s, CH3(CH2)6CH2, 1C), 35.13 (s, PyCH2C, 1C), 53.37 (d,
1JCP ¼ 115.1, C, 1C), 121.68 (s, CHAr (5), 1C), 135.85 (d, 3JCP ¼ 14.8, CAr
(3), 1C), 137.15 (s, CHAr (4), 1C), 143.83 (s, CHAr (6), 1C), 148.19 (s, CHAr
(2), 1C), 180.56 (s, C]O, 1C); Elemental analysis: C16H26NO5P
Calculated: C 55.97, H 7.63, N 4.08; Found: C 55.82, H 7.84, N 4.08;
Scale: 31h (0.48 mmol, 0.2 g); Yield: 56%.
13C NMR (700 MHz): 16.98 (s, CH3C, 1C), 38.45 (d, JCP ¼ 2.3,
2
1
PyCH2C, 1C), 50.32 (d, JCP ¼ 121, C, 1C), 123.74 (s, CHAr (5), 1C),
3
134.90 (d, JCP ¼ 15.6, CAr (3), 1C), 138.93 (s, CHAr (4), 1C), 146.48 (s,
CHAr (6), 1C), 149.72 (s, CHAr (2), 1C), 180.64 (s, C]O, 1C). Elemental
analysis: C9H12NO5P(EtOH)0.02(HCl)0.1 Calculated: C 43.48, H 4.93,
N 5.61; Found: C 43.21; H 4.73; N 5.58; Scale: 31d (0.55 mmol);
Yield: 47%.
5.1.14. 2-(Imidazo[1,2-a]pyridin-3-ylmethyl)-2-phosphonooctanoic
acid (3g)
31P NMR (D2O, pH 13): 20.87; 1H NMR: 0.54e0.66 (m,
CH3(CH2)5, 3H), 0.77e1.00 (m, CH3(CH2)4CH2, 8H), 1.69e1.91 (m,
CH3(CH2)4CH2, 2H), 3.22e3.31 (m, IPCHHC, 1H), 3.68e3.76 (m,
IPCHHC, 1H), 6.95e7.01 ( m, CHAr, 1H), 7.26e7.33 (m, CHAr, 1H), 7.37
(s, CHAr, 1H), 7.48e7.52 (m, CHAr, 1H), 8.42e8.46 (m, CHAr, 1H);
Scale: 32g (0.24 mmol, 0.1 g); Yield: 61%; (>93% purity).
5.1.10. 2-Phosphono-2-(pyridin-3-ylmethyl)pentanoic acid (1e)
31P NMR (D2O, pH 13): 18.11; 1H NMR: 0.89 (t, JHH ¼ 7.2,
3
CH3(CH2)2, 3H), 1.33e1.51 (m, CH3CH2CH2, 2H), 1.59e1.97 (m,
3
CH3CH2CH2, 2H), 3.29e3.50 (m, PyCH2C, 2H), 7.95 (dd, JHH ¼ 8.0,
4.8, CHAr (5), 1H), 8.55e8.62 (m, CHAr, 2H), 8.49 (m, CHAr (2,6), 2H),
8.76 (s, CHAr, 1H); 13C NMR: 11.93 (s, CH3(CH2)2, 1C), 16.15 (d,
3JCP ¼ 5.1, CH3CH2CH2, 1C), 28.37 (s, CH3CH2CH2, 1C), 33.13 (d,
2JCP ¼ 3.1, PyCH2C, 1C), 51.82 (d, 1JCP ¼ 122.7, C, 1C), 124.41 (s, CHAr
(5), 1C), 136.83 (s, CHAr (4), 1C), 137.24 (d, 3JCP ¼ 8.3, CAr (3), 1C), 140.63
(s, CHAr (6), 1C), 147.18 (s, CHAr (2), 1C), 174.70 (s, C]O, 1C);
Elemental analysis: C11H16NO5P Calculated: C 48.36, H 5.90, N
5.13; Found: C 48.12, H 6.03, N 5.09; Scale: 31e (0.3 mmol, 0.1 g);
Yield: 32%.
5.2. Biological evaluations
5.2.1. FRET-biosensor assay on a flow cytometer
Screens were performed as described earlier [45,46]. In brief,
BHK cells were transfected with either Rab5-, 8-, 21-NANOPS or
Ras-NANOPS. Rab21-NANOPS with constitutively active form of
Rab21 (Rab21Q76L) was constructed analogous to Rab5-NANOPS.
18 h after transfection cells were split to 96 well plates at
5 ꢂ 104 cells per well. After the cells have attached (usually 5e7 h),
they were treated with the phosphonocarboxylate (PC) analogs at a
5.1.11. 2-Phosphono-2-(pyridin-3-ylmethyl)heptanoic acid (1f)
3
31P NMR (D2O, pH 10.7): 22.80, 1H NMR: 0.74 (t, JHH ¼ 7.2,
final concentration of 500 mM for 24 h. All PC analogs were dis-
CH3(CH2)4, 3H), 1.02e1.32 (m, CH3(CH2)3(CH2)2, 6H), 1.46e1.59 (m,
solved in PBS with pH 7.0e7.4 and stored in ꢁ20 ꢀC until use.
Following treatment, the cells were detached with 10 mM EDTA in
PBS and fixed with equal volume of 4% PFA in PBS for 15 min at
room temperature. All assays were performed as three indepen-
dent experiments and each experiment had compactin as the in-
ternal control. Cytometry based FRET-measurements were made on
a FACS LSR II (BD bioscience) using the following filters for donor-
(405 nm excitation, 450/50 nm emission filter), acceptor- (488 nm,
excitation, 585/42 nm emission filter) and FRET-channel (405 nm
3
2
CH3(CH2)3CH2, 2H), 2.96 (dd, JHP ¼ 9.2, JHH ¼ 14.0, PyCHHC, 1H),
3.47 (dd, 3JHP ¼ 7.4, 2JHH ¼ 14, PyCHHC, 1H), 7.30 (dd, 3JHH ¼ 8.0, 4.9,
CHAr (5), 1H), 7.76 (dt, 3JHH ¼ 7.9, 4JHH ¼ 1.8, CHAr (4), 1H), 8.28 (bdd,
3JHH ¼ 4.9, 4JHH ¼ 1.8, CHAr (6), 1H), 8.37 (m, CHAr (2), 1H); 13C NMR:
11.63 (s, CH3(CH2)4, 1C), 20.10, 30.59, 30.78 (3s, CH3CH2CH2CH2CH2,
3
3C), 23.23 (d, JCP ¼ 4.5, CH3(CH2)2CH2CH2, 1C), 34.90 (s, PyCH2C,
1C), 53.03 (d, 1JCP ¼ 116.6, C, 1C), 121.67 (s, CHAr (5), 1C), 134.90 (d,
3JCP ¼ 13.7, CAr (3), 1C), 137.21 (s, CHAr (4), 1C), 144.02 (s, CHAr (6), 1C),

F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
87
excitation, 530/30 nm emission filter). The flow cytometer data
were analysed as described earlier [60,61]. In brief, doublet
discrimination was implemented to measure signals of single cells.
For normalized acceptor level calibration, cA, FITC beads (Bangs
Laboratories) with a defined size and fluorescein content were used
as described. A mCFPemCit fusion protein was used to calibrate for
the FRET efficiency and donoreacceptor ratio. Only cells with a
donor mole-fraction, xD ¼ 0.5 0.1 were analysed. The Emax value
was determined as described [45,46].
50 mM FPP. After 10 min, the assay was terminated by adding 200 ml
1:4 conc. HCl/Methanol. Reaction products were then extracted
with 0.4 mL scintillation fluid (Microscint E, Perkin Elmer) and
radioactivity in the upper phase determined by scintillation
counting. Data was analysed using Graphpad Prism.
Acknowledgement
KMB is grateful for financial support provided by Ministry of
Science and Higher Education in Poland (N N204 519839 and
IP2010 003070). FPC is grateful for financial support from the
Alliance for Better Bone Health. AKN is grateful for support by the
graduate school, National Doctoral Programme in Informational
and Structural Biology (ISB). This work was supported by the
Academy of Finland (252381) fellowship grant, the Sigrid Juselius
Foundation, the Cancer Society of Finland and the Marie-Curie
Reintegration Grant to DA.
5.2.2. Confocal imaging
BHK cells were grown on coverslips, transfected using JetPRIME
(Polyplus) for 24 h, treated with compounds for 24 h and fixed with
4% PFA (Sigma, #P6148) in PBS. The coverslips were mounted on
microscopic slides using Mowiol 4e88 (Sigma, # 81381). A Zeiss
LSM 510 confocal microscope with a 63ꢂ/1.4 oil DIC immersion
objective was used to record 12 bit 512 ꢂ 512 fluorescent images,
using 200 mm pinhole size and 0.09 mm pixel size in the frame mode
The authors thank Prof. Tadeusz Gajda and Prof. Charles E.
McKenna for helpful discussions, Ms. Marzena Kujawa for technical
assistance, Jonna Alanko for cloning of the Rab21-NANOPS.
with 8ꢂ averaging. The images were later processed in ImageJ.
5.2.3. Statistical analysis
Statistical differences between mean values of inhibitor treated
and untreated samples were analysed using one-way ANOVA fol-
lowed by Dunnett's test in GraphPad Prism (version 6.0b). Confi-
dence p-levels are indicated by asterisks, with * denoting p, 0.05, **
denoting p, 0.01, and *** denoting p, 0.001. The mean IC50 values for
inhibition were calculated from six measurements and data were
analysed in GraphPad Prism by nonlinear regression analysis on log
(inhibitor concentration) vs. normalized response with a Hill Slope
of ꢁ1.0 using the Marquardt method.
Appendix A. Supporting information
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2014.06.062.
References
[1] C.E. McKenna, L.A. Khawli, A. Bapat, V. Harutunian, Y. Chi Cheng, Inhibition of
herpesvirus and human DNA polymerases by a-halogenated phosphonoace-
tates, Biochem. Pharmacol. 36 (1987) 3103e3106.
5.2.4. Assessment of inhibition of prenylation by western blotting
The effects of the PC analogues on protein prenylation was
studied using triton X-114 fractionation, in which prenylated pro-
teins partition into the detergent-rich phase, whereas unprenylated
proteins remain in the aqueous phase [62]. Briefly, Hela cells were
seeded in DMEM into 12-well plates at 7 ꢂ 104 cells/well, then
treated the following day with PCs in fresh medium. 48 h later
(unless stated otherwise), cell monolayers were rinsed in PBS and
lysed in 20 mM Tris, 150 mM NaCl, pH 7.5, 1% triton X-114, a sample
taken for determination of protein concentration, then remaining
lysate incubated at 37 ꢀC for 10 min. Following centrifugation at
13,000 g for 2 min, the aqueous and detergent-rich phases were
separated, then triton X-114 added back to the aqueous phase to
1% v/v and the extraction process repeated. Aqueous phases,
equivalent to 20 mg of unfractionated lysate, were electrophoresed
on 12% gels and western blotted for Rab11 (an abundant, ubiquitous
[2] C.E. McKenna, J.N. Levy, L.A. Khawli, V. Harutunian, T. Gao Ye, M.C. Starnes,
A. Bapat, Y. Chi Cheng, Inhibitors of viral nucleic acid polymerases. pyro-
phosphate analogs, ACS Symp. Ser. 401 (1989) 1e16.
[3] S. Debarge, J. Balzarini, A.R. Maguire, Design and synthesis of
a-carboxy
phosphono-nucleosides, J. Org. Chem. 76 (2011) 105e126.
[4] M.S. Marma, Z.D. Xia, C. Stewart, F. Coxon, J.E. Dunford, R. Baron,
B.A. Kashemirov, F.H. Ebetino, J.T. Triffitt, R.G.G. Russell, C.E. McKenna, Syn-
thesis and biological evaluation of alpha-halogenated bisphosphonate and
phosphonocarboxylate analogues of risedronate, J. Med. Chem. 50 (2007)
5967e5975.
€
[5] O.I. Artyushin, S.N. Osipov, G.-V. Roschenthaler, I.L. Odinets, Propargyl-
substituted phosphonocarboxylates: efficient synthesis and application to
click chemistry, Synthesis (2009) 3579e3588.
[6] F.H. Ebetino, A.-M.L. Hogan, S. Sun, M.K. Tsoumpra, X. Duan, J.T. Triffitt,
A.A. Kwaasi, J.E. Dunford, B.L. Barnett, U. Oppermann, M.W. Lundy, A. Boyde,
B.A. Kashemirov, C.E. McKenna, R.G.G. Russell, The relationship between the
chemistry and biological activity of the bisphosphonates, Bone 49 (2011)
20e33.
[7] F.P. Coxon, M.H. Helfrich, B. Larijani, M. Muzylak, J.E. Dunford, D. Marshall,
A.D. McKinnon, S.A. Nesbitt, M.A. Horton, M.C. Seabra, F.H. Ebetino,
M.J. Rogers, Identification of a novel phosphonocarboxylate inhibitor of Rab
geranylgeranyl transferase that specifically prevents Rab prenylation in os-
teoclasts and macrophages, J. Biol. Chem. 276 (2001) 48213e48222.
[8] F.H. Ebetino, A.V. Bayless, J. Amburgey, K.J. Ibbotson, S. Dansereau,
A. Ebrahimpour, Elucidation of a pharmacophore for the bisphosphonate
mechanism of bone antiresorptive activity, Phosphorus Sulfur Silicon Relat.
Elem. 109e110 (1996) 217e220.
Rab), b-actin, or unprenylated Rap1A, and bands detected using a
LI-COR Odyssey Infrared Imager. Values quoted in the table are the
lowest effective dose (from 3 independent experiments), i.e. the
lowest dose at which unprenylated Rap1A or Rab11 was detected.
[9] A.J. Roelofs, P.A. Hulley, A. Meijer, F.H. Ebetino, R.G.G. Russell, C.M. Shipman,
Selective inhibition of Rab prenylation by a phosphonocarboxylate analogue
of risedronate induces apoptosis, but not S-phase arrest, in human myeloma
cells, Int. J. Cancer 119 (2006) 1254e1261.
5.2.5. Assessment of cell viability using Alamar blue
The number of viable Hela cells following treatment with PC
compounds for 72 h was assessed as previously described [7].
Briefly, Hela cells were seeded in 96 well plates at 5 ꢂ103 cells/well,
then treated the following day with PCs and BPs in replicates of 3.
The data expressed are the mean SEM of at least 4 independent
experiments.
[10] P.G.J. Fournier, F. Dauhine, M.W. Lundy, M.J. Rogers, F.H. Ebetino, P. Clezardin,
Lowering bone mineral affinity of bisphosphonates as a therapeutic strategy
to optimize skeletal tumor growth inhibition in vivo, Cancer Res. 68 (2008)
8945e8953.
_
[11] C.E. McKenna, B.A. Kashemirov, K.M. Błazewska, I. Mallard-Favier,
C.A. Stewart, J. Rojas, M.W. Lundy, F.H. Ebetino, R.A. Baron, J.E. Dunford,
M.L. Kirsten, M.C. Seabra, J.L. Bala, M.S. Marma, M.J. Rogers, F.P. Coxon, Syn-
thesis, chiral high performance liquid chromatographic resolution and
enantiospecific activity of a potent new geranylgeranyl transferase inhibitor,
2-hydroxy-3-imidazo[1,2-a]pyridin-3-yl-2-phosphonopropionic acid, J. Med.
Chem. 53 (2010) 3454e3464.
5.2.6. GGPP synthase assays
GGPP synthase was purified and assayed as described [63].
Briefly, compounds under investigation were pre-incubated with
4 pmol GGPPS for 10 min, then the reaction initiated by adding
_
[12] K.M. Błazewska, F. Ni, R. Haiges, B.A. Kashemirov, F.P. Coxon, C.A. Stewart,
20
ml substrate containing 50
mM [1e14C] IPP (4
mCi/mmol) and
R. Baron, M.J. Rogers, M. Seabra, F.H. Ebetino, C.E. McKenna, Synthesis,

88
F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
_
stereochemistry and SAR of a series of minodronate analogues as RGGT in-
hibitors, Eur. J. Med. Chem. 46 (2011) 4820e4826.
[13] E.E. Kelly, C.P. Horgan, B. Goud, M.W. McCaffrey, The Rab family of proteins:
25 years on, Biochem. Soc. Trans. 40 (2012) 1337e1347.
[14] C. Recchi, M.C. Seabra, Novel functions for Rab GTPases in multiple aspects of
tumour progression, Biochem. Soc. Trans. 40 (2012) 1398e1403.
[15] J.D. Ochocki, M.D. Distefano, Prenyltransferase inhibitors: treating human
[38] K.M. Błazewska, R. Haiges, B.A. Kashemirov, F.H. Ebetino, C.E. McKenna,
A serendipitous phosphonocarboxylate complex of boron: when vessel be-
comes reagent, Chem. Commun. 47 (2011) 6395e6397.
[39] M.S. Marma, L.A. Khawli, V. Harutunian, B.A. Kashemirov, C.E. McKenna,
Synthesis of alpha-fluorinated phosphonoacetate derivatives using electro-
philic fluorine reagents: perchloryl fluoride versus 1-chloromethyl-4-fluoro-
1,4-diazoniabicyclo[2.2.2]octane bis(tetrafluoroborate) (selectfluor (R)),
J. Fluor. Chem. 126 (2005) 1467e1475.
ailments from cancer to parasitic infections, MedChemComm
4 (2013)
476e492.
[40] Y.-S. Lin, J. Park, Joris W. De Schutter, X. Fang Huang, A.M. Berghuis, M. Sebag,
Y.S. Tsantrizos, Design and synthesis of active site inhibitors of the
human farnesyl pyrophosphate synthase: apoptosis and inhibition of ERK
phosphorylation in multiple myeloma cells, J. Med. Chem. 55 (2012)
3201e3215.
[16] Z. Guo, Y.W. Wu, K.T. Tan, R.S. Bon, E. Guiu-Rozas, C. Delon, U.T. Nguyen,
S. Wetzel, S. Arndt, R.S. Goody, W. Blankenfeldt, K. Alexandrov, H. Waldmann,
Development of selective RabGGTase inhibitors and crystal structure of a
RabGGTase-inhibitor complex, Angew. Chem. Int. Ed. 47 (2008) 3747e3750.
[17] K.-T. Tan, E. Guiu-Rozas, R.S. Bon, Z. Guo, C. Delon, S. Wetzel, S. Arndt,
K. Alexandrov, H. Waldmann, R.S. Goody, Y.-W. Wu, W. Blankenfeldt, Design,
synthesis, and characterization of peptide-based Rab geranylgeranyl trans-
ferase inhibitors, J. Med. Chem. 52 (2009) 8025e8037.
[18] M. Watanabe, H.D.G. Fiji, L. Guo, L. Chan, S.S. Kinderman, D.J. Slamon, O. Kwon,
F. Tamanoi, Inhibitors of protein geranylgeranyltransferase I and Rab ger-
anylgeranyltransferase identified from a library of allenoate-derived com-
pounds, J. Biol. Chem. 283 (2008) 9571e9579.
[19] R.S. Bon, Z. Guo, E. Anouk Stigter, S. Wetzel, S. Menninger, A. Wolf, A. Choidas,
K. Alexandrov, W. Blankenfeldt, R.S. Goody, H. Waldmann, Structure-guided
development of selective RabGGTase inhibitors, Angew. Chem. Int. Ed. 50
(2011) 4957e4961.
[20] C. Deraeve, Z. Guo, R.S. Bon, W. Blankenfeldt, R. DiLucrezia, A. Wolf,
S. Menninger, E. Anouk Stigter, S. Wetzel, A. Choidas, K. Alexandrov,
H. Waldmann, R.S. Goody, Y.-W. Wu, Psoromic acid is a selective and covalent
Rab-prenylation inhibitor targeting autoinhibited RabGGTase, J. Am. Chem.
Soc. 134 (2012) 7384e7391.
[41] M.F. Semmelhack, J.C. Tomesch, M. Czarny, S. Boettger, Preparation of 2-
(alkylthiomethyl)acrylates, J. Org. Chem. 43 (1978) 1259e1262.
[42] W.J. Kruper Jr., P.R. Rudolf, C.A. Langhoff, Unexpected selectivity in the
alkylation of polyazamacrocycles, J. Org. Chem. 58 (1993) 3869e3876.
[43] T. Nagata, M. Inoue, Y. Ashida, K. Noguchi, M. Ono, 3-(3-Cycloalkyl-1H-imi-
dazol-4-yl)propanoic acid derivatives as inhibitors of thrombin-activatable
fibrinolysis inhibitor (TAFIa), PCT Int. Appl. WO2011115064A1, 2011.
[44] Procedure applied from: N. Camper, Ch. Scott, M.E. Migaud, Synthesis of an
analogue of the bisphosphonate drug ibandronate for targeted drug-delivery
therapeutic strategies New J. Chem. 34 (2010) 949e955.
€
[45] M. Kohnke, S. Schmitt, N. Ariotti, A.M. Piggott, R.G. Parton, E. Lacey, R.J. Capon,
K. Alexandrov, D. Abankwa, Design and application of in vivo FRET biosensors
to identify protein prenylation and nanoclustering inhibitors, Chem. Biol. Oxf.
U. K. 19 (2012) 866e874.
ꢁ
€
[46] A.K. Najumudeen, M. Kohnke, M. Solman, K. Alexandrov, D. Abankwa, Cellular
FRET-biosensors to detect membrane targeting inhibitors of N-myristoylated
proteins, PLoS One 8 (2013) e66425.
[21] X. Zhou, S.V. Hartman, E.J. Born, J.P. Smits, S.A. Holstein, D.F. Wiemer, Triazole-
based inhibitors of geranylgeranyltransferase II, Bioorg. Med. Chem. Lett. 23
(2013) 764e766.
[22] R.A. Baron, R. Tavare, A.C. Figueiredo, K.M. Błazewska, B.A. Kashemirov,
C.E. McKenna, F.H. Ebetino, A. Taylor, M.J. Rogers, F.P. Coxon, M.C. Seabra,
Phosphonocarboxylates inhibit the second geranylgeranyl addition by Rab
geranylgeranyl transferase, J. Biol. Chem. 284 (2009) 6861e6868.
[23] J. Hwan No, F. De Macedo Dossin, Y. Zhang, Y.-L. Liu, W. Zhu, X. Feng, J.
Anny Yoo, E. Lee, K. Wang, R. Hui, L.H. Freitas-Junior, E. Oldfield, Lipophilic
analogs of zoledronate and risedronate inhibit Plasmodium geranylgeranyl
diphosphate synthase (GGPPS) and exhibit potent antimalarial activity, Proc.
Natl. Assoc. Sci. U. S. A. 109 (2012) 4058e4063.
[47] J.E. Dunford, K. Thompson, F.P. Coxon, S.P. Luckman, F.M. Hahn, C.
Dale Poulter, F.H. Ebetino, M.J. Rogers, Structure-activity relationships for
inhibition of farnesyl diphosphate synthase in vitro and inhibition of bone
resorption in vivo by nitrogen-containing bisphosphonates, J. Pharmacol. Exp.
Ther. 296 (2001) 235e242.
[48] A. Dudakovic, A.J. Wiemer, K.M. Lamb, L.A. Vonnahme, S.E. Dietz, R.J. Hohl,
Inhibition of geranylgeranyl diphosphate synthase induces apoptosis through
multiple mechanisms and displays synergy with inhibition of other iso-
prenoid biosynthetic enzymes, J. Pharmacol. Exp. Ther. 324 (2008)
1028e1036.
[49] A.J. Wiemer, R.J. Hohl, D.F. Wiemer, The intermediate enzymes of isoprenoid
metabolism as anticancer targets, Anti Cancer Agents Med. Chem. 9 (2009)
526e542.
[50] F.P. Coxon, M.H. Helfrich, H. Robert Van't, S. Sebti, S.H. Ralston, A. Hamilton,
M.J. Rogers, Protein geranylgeranylation is required for osteoclast formation,
function, and survival: Inhibition by bisphosphonates and GGTI-298, J. Bone
Min. Res. 15 (2000) 1467e1476.
ꢀ
_
[24] B. Fiszer, J. Michalski, Synthesis of organophosphorus compounds based on
phosphonoacetic ester and its analogs. Addition of phosphonoacetic ester,
alkylated phosphonoacetic esters, and phosphonoacetic nitrile to a, b-unsat-
urated esters and nitriles, Rocz. Chem. 28 (1954) 185e195.
[25] K. Lee, D.F. Wiemer, A convenient preparation of
a-phosphono esters and
lactones via carbon-phosphorus bond formation, Phosphorus Sulfur Silicon
Relat. Elem. 75 (1993) 87e90.
[26] H. Krawczyk, K. Wa˛sek, J. Ke˛dzia, J. Wojciechowski, W.M. Wolf, A general
stereoselective method for the synthesis of cyclopropanecarboxylates. A new
version of the homologous Horner-Wadsworth-Emmons reaction, Org. Bio-
mol. Chem. 6 (2008) 308e318.
[51] F.P. Coxon, A. Taylor, C.A. Stewart, R. Baron, M.C. Seabra, F.H. Ebetino,
M.J. Rogers, The gunmetal mouse reveals Rab geranylgeranyl transferase to be
the major molecular target of phosphonocarboxylate analogues of
bisphosphonates, Bone 49 (2011) 111e121.
[52] C.M. Szabo, Y. Matsumura, S. Fukura, M.B. Martin, J.M. Sanders, S. Sengupta,
J.A. Cieslak, T.C. Loftus, C.R. Lea, H.-J. Lee, A. Koohang, R.M. Coates, H. Sagami,
E. Oldfield, Inhibition of geranylgeranyl diphosphate synthase
[27] P. Coutrot, A. Ghribi,
A facile and general one-pot synthesis of 2-
oxoalkanephosphonates from diethylphosphonocarboxylic acid chlorides
by bisphosphonates and diphosphates: a potential route to new bone
and organometallic reagents, Synthesis (1986) 661e664.
[28] K. Hackeloer, G. Schnakenburg, S.R. Waldvogel, Oxidative coupling reactions
of 1,3-diarylpropene derivatives to dibenzo[a,c]cycloheptenes by PIFA, Eur. J.
Org. Chem. 2011 (2011) 6314e6319.
[29] E. Błaszczyk, H. Krawczyk, T. Janecki, 2-Diethoxyphosphoryl-4-
nitroalkanoates e versatile intermediates in the synthesis of
lactones and lactams, Synlett (2004) 2685e2688.
[30] L. Albrecht, B. Richter, H. Krawczyk, K. Anker Jøergensen, Enantioselective
antiresorption and antiparasitic agents, J. Med. Chem. 45 (2002) 2185e2196.
[53] K.-M.C. Cammy, M.P. Hudock, Y. Zhang, R.-T. Guo, R. Cao, J.H. No, P.-H. Liang,
T.-P. Ko, T.-H. Chang, S.-c. Chang, Y. Song, J. Axelson, A. Kumar, A.H.-J. Wang,
E. Oldfield, Inhibition of geranylgeranyl diphosphate synthase by
bisphosphonates: a crystallographic and computational investigation, J. Med.
Chem. 51 (2008) 5594e5607.
[54] R.J. Barney, B.M. Wasko, A. Dudakovic, R.J. Hohl, D.F. Wiemer, Synthesis and
biological evaluation of a series of aromatic bisphosphonates, Bioorg. Med.
Chem. 18 (2010) 7212e7220.
€
a-alkylidene-g-
organocatalytic approach to
a-methylene-d-lactones and d-lactams, J. Org.
Chem. 73 (2008) 8337e8343.
[55] U.T.T. Nguyen, J. Cramer, J. Gomis, R. Reents, M. Gutierrez-Rodriguez,
R.S. Goody, K. Alexandrov, H. Waldmann, Exploiting the substrate tolerance of
farnesyltransferase for site-selective protein derivatization, ChemBioChem 8
(2007) 408e423.
[56] Z. Guo, Y.W. Wu, D. Das, C. Delon, J. Cramer, S. Yu, S. Thuns, N. Lupilova,
H. Waldmann, L. Brunsveld, R.S. Goody, K. Alexandrov, W. Blankenfeldt,
Structures of RabGGTase-substrate/product complexes provide insights into
the evolution of protein prenylation, EMBO J. 27 (2008) 2444e2456.
[57] A.Q. Gomes, B.R. Ali, J.S. Ramalho, R.F. Godfrey, D.C. Barral, A.N. Hume,
M.C. Seabra, Membrane targeting of Rab GTPases is influenced by the pre-
nylation motif, Mol. Biol. Cell. 14 (2003) 1882e1899.
[31] F.H. Ebetino, A.V. Bayless, S.M. Dansereau, Phosphonocarboxylate compounds
pharmaceutical compositions, and methods for treating abnormal calcium
and phosphate metabolism, The Procter and Gamble Co., USA, US Patent
5,760,021, 1998.
[32] J.N. Levy, C.E. McKenna, Oxidations of triethyl a-phosphonoacrylate. Epoxi-
dation to triethyl a-phosphonoacrylate oxide by hypochlorite and formation
of triethyl dihydroxyphosphonoacetate with RuO4-periodate, Phosphorus
Sulfur Silicon Relat. Elem. 85 (1993) 1e8.
[33] B. Iorga, F. Eymery, P. Savignac, The synthesis and properties of 1,2-
epoxyalkylphosphonates, Synthesis (1999) 207e224.
[34] F.M.D. Ismail, Important fluorinated drugs in experimental and clinical use,
J. Fluor. Chem. 118 (2002) 27e33.
[35] S. Purser, P.R. Moore, S. Swallow, V. Gouverneur, Fluorine in medicinal
chemistry, Chem. Soc. Rev. 37 (2008) 320e330.
[36] I. Ojima, Exploration of fluorine chemistry at the multidisciplinary interface of
chemistry and biology, J. Org. Chem. 78 (2013) 6358e6383.
[58] U.T.T. Nguyen, A. Goodall, K. Alexandrov, D. Abankwa, Isoprenoid modifica-
tions, Protein Rev. Post Transl. Modifi. Health Dis. 13 (2011) 1e37.
[59] N.H. Thoma, A. Iakovenko, D. Owen, A.S. Scheidig, H. Waldmann, R.S. Goody,
€
K. Alexandrov, Phosphoisoprenoid binding specificity of geranylgeranyl-
transferase type II, Biochemistry 39 (2000) 12043e12052.
[60] D. Abankwa, H. Vogel, A FRET map of membrane anchors suggests distinct
microdomains of heterotrimeric G proteins, J. Cell. Sci. 120 (2007) 2953e2962.
[61] D. Abankwa, M. Hanzal-Bayer, N. Ariotti, S.J. Plowman, A.A. Gorfe, R.G. Parton,
J. Andrew McCammon, J.F. Hancock, A novel switch region regulates H-ras
membrane orientation and signal output, EMBO J. 27 (2008) 727e735.
ꢀ
ꢀ
~
[37] J. Wang, M. Sanchez-Rosello, J.L. Acena, P. Carlos del, A.E. Sorochinsky,
S. Fustero, V.A. Soloshonok, H. Liu, Fluorine in pharmaceutical industry:
fluorine-containing drugs introduced to the market in the last decade
(2001e2011), Chem. Rev. 114 (2014) 2432e2506.

F. Coxon et al. / European Journal of Medicinal Chemistry 84 (2014) 77e89
89
[62] F.P. Coxon, F.H. Ebetino, E.H. Mules, M.C. Seabra, C.E. McKenna, M.J. Rogers,
Phosphonocarboxylate inhibitors of Rab geranylgeranyl transferase disrupt
the prenylation and membrane localization of Rab proteins in osteoclasts
in vitro and in vivo, Bone 37 (2005) 349e358.
[63] K.L. Kavanagh, K. Guo, J.E. Dunford, X. Wu, S. Knapp, F.H. Ebetino, M.J. Rogers,
R.G.G. Russell, U. Oppermann, The molecular mechanism of nitrogen-
containing bisphosphonates as antiosteoporosis drugs, Proc. Natl. Assoc. Sci.
U. S. A. 103 (2006) 7829e7834.