Identification of the Privileged Position in the Imidazo[1,2-a]pyridine Ring of Phosphonocarboxylates for Development of Rab Geranylgeranyl Transferase (RGGT) Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.7b00811

Members of the Rab GTPase family are master regulators of vesicle trafficking. When disregulated, they are associated with a number of pathological states. The inhibition of RGGT, an enzyme responsible for post-translational geranylgeranylation of Rab GTPases represents one way to control the activity of these proteins. Because the number of molecules modulating RGGT is limited, we combined molecular modeling with biological assays to ascertain how modifications of phosphonocarboxylates, the first reported RGGT inhibitors, rationally improve understanding of their structure-activity relationship. We have identified the privileged position in the core scaffold of the imidazo[1,2-a]pyridine ring, which can be modified without compromising compounds' potency. Thus modified compounds are micromolar inhibitors of Rab11A prenylation, simultaneously being inactive against Rap1A/Rap1B modification, with the ability to inhibit proliferation of the HeLa cancer cell line. These findings were rationalized by molecular docking, which recognized interaction of phosphonic and carboxylic groups as decisive in phosphonocarboxylate localization in the RGGT binding site.

Full text of DOI:10.1021/acs.jmedchem.7b00811

Subscriber access provided by PEPPERDINE UNIV
Article
Identification of the privileged position in the imidazo[1,2-
a]pyridine ring of phosphonocarboxylates for development
of Rab geranylgeranyl transferase (RGGT) inhibitors
Aleksandra Ka#mierczak, Damian Kusy, Sanna P. Niinivehmas, Joanna Gmach, #ukasz
Joachimiak, Olli T Pentikäinen, Edyta Gendaszewska-Darmach, and Katarzyna M. Blazewska
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00811 • Publication Date (Web): 27 Sep 2017
Downloaded from http://pubs.acs.org on September 28, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Identification of the privileged position in the imidazo[1,2-a]pyridine ring of
phosphonocarboxylates for development of Rab geranylgeranyl transferase (RGGT)
inhibitors
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b#
a#
c
a
Aleksandra Kaźmierczak, Damian Kusy, Sanna P. Niinivehmas, Joanna Gmach, Łukasz
Joachimiak,a Olli T. Pentikäinen,c,d Edyta Gendaszewska-Darmach,b Katarzyna M.
Błażewskaa*
a
Institute of Organic Chemistry, Faculty of Chemistry, Lodz University of Technology,
Żeromskiego Str. 116, 90-924 Łódź, Poland.
b
Institute of Technical Biochemistry, Faculty of Biotechnology and Food Sciences, Lodz
University of Technology, Stefanowskiego Str. 4/10, 90-924 Łódź, Poland.
c
University of Jyvaskyla, Department of Biological and Environmental Science &
Nanoscience Center, P.O.Box 35, FI-40014 University of Jyvaskyla, Finland.
d
University of Turku, Institute of Biomedicine, FI-20520 Turku, Finland
ABSTRACT
Members of the Rab GTPase family are master regulators of vesicle trafficking. When
disregulated, they are associated with a number of pathological states. The inhibition
of RGGT, an enzyme responsible for post-translational geranylgeranylation of Rab
GTPases represents one way to control the activity of these proteins. Since the
number of molecules modulating RGGT is limited, we combined molecular modeling
with biological assays to ascertain how modifications of phosphonocarboxylates, the
first reported RGGT inhibitors, rationally improve understanding of their structure-
activity relationship. We have identified the privileged position in the core scaffold of
the imidazo[1,2-a]pyridine ring, which can be modified without compromising
compounds’ potency. Thus modified compounds are micromolar inhibitors of
Rab11A prenylation, simultaneously being inactive against Rap1A/Rap1B
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
modification, with the ability to inhibit proliferation of the HeLa cancer cell line.
These findings were rationalized by molecular docking, which recognized interaction
of phosphonic and carboxylic groups as decisive in phosphonocarboxylate
localization in the RGGT binding site.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION
Rab GTPases constitute the most abundant class among the Ras superfamily of small
GTP-binding proteins. They are considered master regulators of vesicle trafficking
and their secretion to the extracellular matrix. Rab geranylgeranyl transferase (RGGT,
Rab GGTase, GGT-II) is responsible for post-translational geranylgeranylation of
Rab GTPases (Fig. 1). Attachment of lipophilic prenyl group(s) is crucial for their
correct subcellular membrane localization and functioning. Most Rab GTPases
undergo double prenylation catalyzed by RGGT, with the formation of a thioether
bond between two C-terminal cysteines and 20-carbon isoprenoid chains, derived
from geranylgeranyl pyrophosphate (GGPP).
The abnormal activities of RGGT and some Rab proteins have been identified in a
number of diseases, including neurodegenerative and infectious disorders1-3 as well as
in several cancer types, in some cases being connected with increased tumor
aggression.4-6 Specifically, members of the Rab11 family (Rab11A, Rab11B,
Rab25/Rab11C) along with their effectors exert an important role in development of
cancers of multiple lineages, including pancreas, skin, breast, colon, endometrial,
lung, ovarian, renal, prostate and bladder.1, 7
ACS Paragon Plus Environment

Page 3 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. The isoprenoid biosynthesis pathway. NBP: nitrogen-containing
bisphosphonates; FPPS: farnesyl pyrophosphate synthase; GGPPS: geranylgeranyl
pyrophosphate synthase; FT: farnesyl transferase; GGT-1: geranylgeranyl
transferase 1; FPP: farnesyl pyrophosphate; GPP: geranyl pyrophosphate; GGPP:
geranylgeranyl pyrophosphate; RGGT: Rab geranylgeranyl transferase.
It has been also reported that disruption of prenylation by inhibition of more upstream
enzymes of isoprenoid pathway, e.g. farnesyl pyrophosphate synthase, FPPS (by
nitrogen-containing bisphosphonates (NBPs)), can reduce proliferation of malignant
cells. Even though the molecular target of NBPs is farnesyl pyrophosphate synthase,
their molecular mechanism of action might be related to downstream depletion of
geranylgeranyl pyrophosphate, a substrate for geranylgeranyl transferases (Fig. 1).8-10
Additionally, the unexpectedly low cytotoxicity of farnesyl transferase (FT)
inhibitors, which should disrupt prenylation of the oncogenic K-Ras GTPase, is
associated with alternative prenylation of this protein by geranylgeranyl
transferases.10-12 The combined therapy using both, farnesyl transferase and
geranylgeranyl transferase inhibitors, may overcome this surrogate prenylation.10, 11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Therefore, identification of specific inhibitors of RGGT and their application as tools
to study RGGT and Rab-connected processes may open new avenues for therapeutic
intervention and advances in understanding of Rab proteins prenylation.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The α-phosphonocarboxylate (PC), 3-PEHPC13 derived from risedronate,
bisphosphonate anti-osteoporotic drug, was the first identified inhibitor of RGGT,
which showed moderate but selective activity. Up to now, several other classes of
RGGT inhibitors have been reported,14-18 with the most active analogs being
1
6
developed from a tetrahydrobenzodiazepine core. None of them were studied with
respect to inhibition of first or second geranylgeranylation process, except for
phosphonocarboxylates, which were reported to prohibit the introduction of the
second geranylgeranyl group to Rabs, therefore selectively affecting Rabs, which
require double prenylation.19
Further studies led to the identification of the more potent RGGT
phosphonocarboxylate inhibitors. We found that other analogs of heterocyclic
bisphosphonates show greater activity against RGGT than originally reported 3-
PEHPC,20-22 while those derived from aminoalkyl BP are inactive.22 Oxidation of
nitrogen in the pyridine ring of 3-PEHPC to N-oxide, did not improve activity against
RGGT.23 We demonstrated that the most active derivatives of currently known
phosphonocarboxylates contain imidazo[1,2-a]pyridine (3-IPEHPC and its analogs)20,
21
22
or the imidazole ring in their structure.
Thus, here we describe the synthesis of an extended set of phosphonocarboxylate
inhibitors (Fig. 2), varying in lipophilicity, size and geometry of substituent in order
to probe their affinity towards RGGT. We investigated analogs of 3-IPEHPC with the
diversely substituted imidazo[1,2-a]pyridine ring (1a-1c, 1e-1r, 2a, 2b, 2d, 2h) as
well as analogs with a modified phosphonic acid moiety, in the form of
ACS Paragon Plus Environment

Page 5 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
hydroxy(phenyl)phosphoryl or ethyl(hydroxy)phosphoryl residues (1s-t, 2s-t). We
have also synthesized compounds potentially resembling GGPP, one of the RGGT
2
4
substrate (1u, 1w, 1y, 2z), where a prenyl chain was mimicked with a polyphenyl
moiety. Since it is suggested that PCs interact with the enzyme cavity storing the
geranylgeranyl residue after first prenylation,15 we hypothesized that such prenyl-
mimicking moiety could inhibit RGGT by interacting with this lipophilic pocket.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Structures of studied compounds.
We examined the potency of thus synthesized compounds by investigating their effect
on RGGT-driven geranylgeranylation of Rab11A, one of the most characterized
molecule among all Rab proteins, along with nonselective prenylation of
Rap1A/Rap1B, a strict substrate of GGTase-1.
RESULTS AND DISCUSSION
Synthesis of novel phosphonocarboxylate analogs
We applied synthetically robust approach, in which the functionalized
phosphonocarboxylate scaffold was built through condensation of an aldehyde 3 and
2
5
triethyl phosphonoacetate 5 or diethyl phosphinoacetate 6 (Scheme 1, route 1). That
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
way we obviated the formation of the doubly substituted analog and necessity of
preparing the alkylating entity, as was needed for the previously used low-yielding
2
1
strategy. While this new approach worked well for most studied analogs, it turned
out troublesome for derivatives of 2-substituted imidazo[1,2-a]pyridines (Fig. 2,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
compounds 1a-c), because preparation of corresponding starting aldehydes required
low-yielding and tedious formylation, e.g. through Vilsmeier-Haack reaction.20
Therefore, instead of using aldehydes 3 as the starting materials, we synthesized 2-
substituted imidazo[1,2-a]pyridine analogs 4 and subjected them to Friedel-Crafts
alkylation with ethyl 2-(diethoxyphosphoryl)acrylate 7 (Scheme 1, route 2), according
2
6
to procedure applied for 2-phenylindolizine. Unfortunately, none of these strategies
worked for imidazo[1,2-a]pyridine analogs substituted with halogen atom in the
position 5, therefore only representatives with alkyl and aryl substituent in this
location were obtained.
Scheme 1. Two strategies used for the synthesis of phosphonocarboxylates, presented
on the example of 3-IPEHPC analogs. Reagents and conditions: (a) Triethyl
o
phosphonoacetate, TiCl
4
, TEA, DCM, -20 C to rt; (b) NaBH
4
, NiCl
2
x6H O, MeOH, -
2
ACS Paragon Plus Environment

Page 7 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
o
o
4
0 C; (c) 7, 1,4-dioxane, AlCl (cat), reflux; (d) NFSI, BuLi, THF, -70 C/ 20
3
o
minutes, -20 C/ 1 h; (e) 12 M HCl, reflux.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Route 1 required synthesis of imidazo[1,2-a]pyridine-3-carbaldehydes 3 (Fig. 3a). For
their synthesis, appropriate 2-aminopyridine analogs were used in the reaction with
bromomalonaldehyde.27 Apart from commercially available halogen and methyl
substituted 2-aminopyridine, phenyl substituted analogs had to be first synthesized
either from halogen-containing 2-aminopyridines or from the corresponding halogen-
substituted aldehydes 3, in the Suzuki-Miyaura coupling with phenylboronic acid.28
Polyphenyl aldehydes 3u, 3w, 3y, 3z (Fig. 3), were synthesized according to literature
procedures (see SI for experimental details).29, 30
Figure 3. Starting materials for the synthesis of target compounds, representing a)
heterocycle and aromatic part; b) phosphonoacetate and phosphinoacetate
derivatives.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Thus obtained aldehydes 3 were subjected to TiCl -mediated condensation with
4
phosphonoacetate 5 or phosphinoacetate 6, in the presence of TEA (Scheme 1),
following the procedure applied previously by Tanabe for condensation of triethyl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
phosphonoacetate and benzaldehyde. The products 8 in the form of a E/Z mixture
were obtained with yields in the range 40-84%.32
Compounds 8 were subjected to hydrogenation catalyzed by Pd-C. However, that
reaction was accompanied by partial reduction of the pyridine ring, as was observed
3
3
previously for 2-(imidazo[1,2-a]pyridin-2-yl)acetates. Therefore, the reduction was
carried out with NaBH and NiCl hydrate instead, to obtain products 9 with yields
4
2
3
4
5
9-98%. While non-halogen containing analogs 8 could be reduced using conditions
o
of 15 min at -20 C, reduction of halogen-containing derivatives required shorter time
o
(
8 min) and lower temperature (-40 C), due to their propensity towards halogen-
carbon bond cleavage (up to 30% of dehalogenated product was detected if such
precaution was not undertaken). As we proved for selected examples, condensation
and reduction steps can be run subsequently, without purification of an intermediate
vinyl analog 8, giving comparable yields to procedure involving its purification.
Thus obtained analogs
9
were subjected to fluorination using N-
fluorobenzenesulfonimide (NFSI) and butyl lithium, according to procedure
previously applied for fluorination of bisphosphonates.35 Fluorinated products 10
were obtained with 41-98% yields (Scheme 1). The last step involved hydrolysis of
desoxy 9 and fluoro 10 esters under acidic conditions. Final products were obtained
with 55-99% yields36 upon precipitation. The same procedure was applied in the
synthesis of phosphinoacetate analogs 1s-t and 2s-t.
For the synthesis of analogs with the imidazo[1,2-a]pyridine ring substituted in the
position 2, compounds 4 (obtained by condensation of suitable 2-aminopyridine and
ACS Paragon Plus Environment

Page 9 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
chloroacetone or bromoacetophenone) were subject to Friedel-Crafts alkylation,
leading to esters 9 with 55-63% yields (Scheme 1, route 2). This approach seem
advantageous compared with route 1, the advantages being: (a) starting materials 4
are obtained in yields >70%, while synthesis of appropriate aldehydes 3 requires low
yielding formylation; (b) substrates can be directly transformed to product 9, thereby
eliminating one reaction step; (c) avoidance of dehalogenation upon reduction, see
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(b). However, we found it applicable only to analogs of imidazo[1,2-a]pyridines, in
which the position 2 is occupied by substituent. In such cases, selective alkylation
takes place only on carbon 3, the position of imidazo[1,2-a]pyridine most activated.37,
3
8
Thus obtained products of alkylation 9 were subjected to fluorination (72-95%
yield) and hydrolysis (53-89% yield), according to the same protocol as the one
described above.
In order to study the role of chirality in interaction between RGGT and -fluorinated
phosphonocarboxylate analogs, we carried out the resolution of enantiomers on the
2
0
example of compound 1i. We used previously developed HPLC method, utilizing
weak anion exchange chiral column introduced by Lindner et al.39
All synthesized fluorine-containing compounds 1 were screened for their biological
activity, using cell-based assays. Based on previous studies,21 we predicted that
desoxy analogs 2 should be less potent, therefore representatives of such were
screened only in selected cases. In the case of 2d, a fluoro analog was not obtained,
leaving the 5-methyl substituted compound 1e as the sole fluorinated representative of
this position.
The cytotoxic efficacy of phosphonocarboxylates against HeLa cancer cell line
Inhibition of prenyltransferases or enzymes in the mevalonate pathway may be
associated with decreased cell viability.13, 40, 41 Therefore, the newly synthesized
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
phosphonocarboxylate analogs were initially evaluated for their antiproliferative
activity against the human epithelial adenocarcinoma cell line (HeLa), using
previously described 3-IPEHPC20, 22 as a reference. Taking into account that activities
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
2
of tested compounds could be attenuated or modulated by serum proteins, influence
of FBS-containing and FBS-free media was also investigated (Table 1).
All analogs of 3-IPEHPC, in which a hydroxyl group on the branching carbon atom
between the phosphonic and carboxylic groups (C-α) was exchanged for a hydrogen
atom (2a, 2b, 2d, 2h), had no effect on the number of viable HeLa cells up to
maximum concentration tested (2 mM) in both serum-containing or serum-free
medium. The same effect was observed also for 2-(alkyl/aryl(hydroxy)phosphoryl)-
propanoic acid analogs (1s-t, 2s-t).
Introduction of a phenyl (1a, 1n), a methyl (1b, 1o) or a bromine (1p) substituent into
the position 2 or 8 of the imidazo[1,2-a]pyridine ring of fluoro analogs did not show
any effect on HeLa cell proliferation either. Substitution at the 7 (1j-m) or 5 (1e)
position in turn had negligible or no effect when cells where cultured in complete
medium, but in most cases significantly increased cytotoxic effect when applied under
serum-free condition.
Remaining fluoro analogs modified at the 6 position of the imidazo[1,2-a]pyridine
ring (1f-i) demonstrated cytotoxic effect under both medium conditions, being the
most potent molecules, with IC50 values ranging from 68 µM to 323 µM in fasting
medium and from 267 µM to 521 µM in FBS-containing medium. They showed
similar or even stronger effect when compared to the reference compound, 3-
IPEHPC. It is noteworthy that introduction of an additional methyl substituent in the 2
position of compound 1i (1c) led to significant limitation of antiproliferative activity
against HeLa cells (IC50 = 1952 µM in complete medium).
ACS Paragon Plus Environment

Page 11 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Exchanging a heterocycle with biphenyl or benzylphenyl moieties resulted in no or
weak cytotoxic effect in DMEM supplemented with FBS (1u, 1w, 1y), while the
desoxy analog containing a triphenyl residue (2z) turned out to be very potent. The
IC50 values in fasting medium were in the range of 318 - 930 µM for compounds 1u,
1w, 1y, and the most active appeared to be 2z with IC50 below 25 µM.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Functionalization of the imidazo[1,2-a]pyridine ring in the position 6 increases
potency of 3-IPEHPC analogs against Rab prenylation.
All synthesized fluorine-containing compounds 1 and selected representatives of
desoxy analogs 2 were screened for their ability to inhibit RGGT and GGT-1 in intact
cells, which also allowed assessing their cellular availability. As previously,22 the
assays were based on determination of the unprenylated forms of Rab11A and
Rap1A/Rap1B, which are modified with geranylgeranyl groups by RGGT and GGT-
1
, respectively. We reasoned that independent effect observed either for Rab11A or
Rap1A/Rap1B would indicate selective inhibition of RGGT or GGT-1, while
comparable effect for both proteins would indicate that mevalonate pathway enzymes,
e.g. GGPPS or FPPS, might be affected, since they are providing prenyl
pyrophosphate substrate for both transferases (Fig. 1).
Our choice of the method for activity assessment was dictated by the fact that most
assays on isolated RGGT utilize diversely labeled, non-commercially available GGPP
analogs, often requiring multi-step synthesis, and being labile due to the presence of
pyrophosphate moiety. The differences in IC50 values obtained in different in vitro
16
assays suggest that they may measure different modes of inhibition, none of them
being adjusted to inhibitors of second geranylgeranylation, such as
phosphonocarboxylates. For such inhibitors using labeled GGPP may lead to
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ambiguous readout, since both, the properly doubly geranylgeranylated as well as
mono geranylgeranylated intermediate will be functionalized with a visualizing
handle. Therefore, we studied the activity of compounds through fractionation of
cellular lysates and immunochemical detection of mis-prenylated Rab11A present in
cytosol.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The inhibitory activity of the set of investigated compounds 1-2 was initially tested at
1
00 µM concentration under serum-free conditions and the Rab11A enrichment in the
cytosol fraction was examined with Western Blot analysis (Table 1). Treatment with
the series of GGPP mimetics (1u, 1w, 1y, 2z) and with desoxy analogs of 3-IPEHPC
(2a, 2b, 2d, 2h) did not deplete Rab11A from membrane fractions of HeLa cells.
Among five analogs modified at the 2 or 8 position of the imidazo[1,2-a]pyridine
ring, only two, compounds 1a and 1o, have shown an effect on Rab11A prenylation.
7
-Substituted fluoro analogs (1j-m) did not inhibit prenylation of Rab11A, but they
did show ability to reduce HeLa cells viability in serum-free medium. The 5-
substituted fluoro analog, compound 1e, has shown inhibition of Rab11A prenylation
at 100 µM and significant reduction of HeLa cell viability, but only in serum-free
medium.
The set of phosphinic acid derivatives (1s-t, 2s-t) also weakly inhibited Rab11A
prenylation at 100 µM, despite their non-cytotoxic character.
Interestingly, all the fluoro analogs containing substituent at the position 6 of the
imidazo[1,2-a]pyridine ring (1f-i), enriched Rab11A protein level in cytosolic
fraction, which correlates well with their ability to reduce HeLa cell viability. As an
example of a desoxy analog, compound 2h was also tested and it had no effect on
both, HeLa cells viability and Rab11A potency. It supports the previously observed
trend21, 43 that fluorinated analogs are more active than their desoxy counterparts.
ACS Paragon Plus Environment

Page 13 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. Effects of synthesized PC analogs on HeLa cell viability in serum-
a
supplemented and serum-free medium , and Rab11A prenylation tested at a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
concentration of 100 µMb.
Reduction of HeLa
cell viability
Inhibition of
Rab11A
(IC50/µM)d
prenylation
at
1
2
3
Entr Compoun
R / R / R
in medium
in
(see structure)c
concentratio
n of 100 µM
y
d
supplemente serum-
d with 10%
FBS
free
mediu
m
1
2
3
.
3-
H
585
167
+
+
-
IPEHPCc
.
1ae
1b
2-Ph
2-Me
no effect
no effect
no
effect
.
no
effect
4
5
.
.
1ce
1ee
2-Me-6-Br
5-Me
1952
1487
879
224
-
+
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
7
8
.
.
.
1fe
1ge
6-Ph
6-Me
6-Cl
6-Br
6-Br
6-Br
7-Ph
7-Me
7-Cl
7-Br
8-Ph
510
358
68
222
323
156
NE
+
+
+
+
-
1he
1i e
521
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
9.
504
1
1
1
0.
1i-E1f
1i-E2f
1j e
NE
1.
2.
267
130
186
1147
351
634
+
-
1579
13.
14.
1k
no effect
no effect
no effect
no effect
-
1l
-
1
1
5.
6.
1m
-
1n
no
-
effect
1
7.
1oe
1p
8-Me
no effect
no effect
747
no
+
-
effect
1
8.
8-Br
no
effect
19.
1re
122
-
20.
2a
2-Ph
no effect
no
-
effect
ACS Paragon Plus Environment

Page 15 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
2
1.
2.
2b
2d
2h
2-Me
5-Ph
6-Cl
no effect
no effect
no effect
no
-
-
-
effect
no
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
effect
23.
no
effect
1sg
Ph
Et
no effect
no effect
no
effect
+
+
+
2
2
2
4.
5.
6.
1tg
no
effect
2s
Ph
no effect
no
effect
2
7.
2t
Et
no effect
no
+
effect
28.
1u
1w
1y
2z
m-Ph
p-Ph
no effect
no effect
no effect
778
930
318
427
<25
-
-
-
-
2
3
3
9.
0.
1.
m-Bn
m-C
6
4
H (p-Ph)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
aHeLa cells were treated with the compounds for 72 h and then viable cell number was assayed. Represented data
were calculated from the means of eight tested concentrations from at least 3 independent experiments.
bHeLa cells were treated for 48 h with 100 µM concentration of the compounds, then lysed and separated into
cytosolic and membrane-rich fractions and Western blotted for Rab11A and β-actin in cytosolic fractions. Data
shown in the table, from at least 3 independent experiments, indicates for which compounds higher bands intensity
compared to control (untreated cells) was observed.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
c
For analogs modified in the imidazo[1,2-a]pyridine ring, the position of modification is given according to the
numbering of heterocycle.
d
“no effect” stands for lack of response in cells treated with up to 2 mM concentration of inhibitor.
e
Compounds in bold were evaluated to inhibit Rab11A and Rap1A/Rap1B prenylation in wide concentration
range.
f
The enantiomers of 1i are distinguished as E1 and E2, based on their retention times during chiral HPLC
separation. The enantiomer with the shorter retention time is labeled as E1, and the one with the longer retention
time, as E2.
g
Compounds demonstrated insignificant activity to inhibit Rab11A prenylation (based on densitometry analysis),
which was the reason why they have not been tested in wide concentration range.
Thirteen compounds (3-IPEHPC,1a,1c,1e-j,1o,1r, 2s-t) were selected for full panel 6
doses assay to determine the lowest effective dose inhibition of Rab11A prenylation
based on their antiproliferative activity (IC50 values) and the initial Western Blot
screening at a single concentration of 100ꢀµM (see last column in Table 1). Selected
compounds were assessed for their ability to inhibit Rab11A and Rap1A/Rap1B
prenylation by RGGT and GGTase 1, respectively.
A
Compound
1f
1g
1h
1i
Conc. (µM)
Rab11A
Lov
0
1
5
10 25 50 100
Lov
0
1
5
10 25 50 100
Lov
0
1
5
10 25 50 100
Lov
0
1
5 10 25 50 100
Rap1A/Rap1B
β-actin
ACS Paragon Plus Environment

Page 17 of 94
Journal of Medicinal Chemistry
1
2
3
B
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Effect of 6-substituted analogs of 3-IPEHPC (1f-i) on Rab11A and Rap1A/Rap1B prenylation in HeLa cells. Cells
were treated for 48 h with the indicated concentrations of PCs (µM) and 10 μM lovastatin (Lov) acting as a positive control in
serum-free medium. Subsequently cells were lysed and fractionated into cytosolic and membrane-rich fractions. Cytosolic
fractions containing unprenylated proteins were western blotted for Rab11A, Rap1A/Rap1B and β-actin (A) and the immunoblot
bands were quantified by densitometry analysis, normalized to β-actin, and displayed as a percentage of controls (B). Data
represented mean ± SEM from at least three independent experiments, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 and ****p ≤ 0.0001.
Since compounds 1j and 1r did not show any effect on Rab11A prenylation
inhibition, their significant potency to reduce HeLa cell viability (Table 1) should be
associated with different mechanism of action.
Compounds that showed no influence on HeLa cell viability (1a, 1o, 2s-t) were
proven to be weak inhibitors of Rab11A prenylation. They demonstrated similar
effect in the case of Rap1A/Rap1B prenylation, except for 1a, which turned out to be
inactive even at 100 µM concentration. 1e was another weak inhibitor, with equal
LEDs obtained for both Rab11A and Rap1A/Rap1B prenylation. Such results suggest
that those PCs may inhibit an enzyme of the mevalonate pathway such as GGPPS or
FPPS (Fig. 1), responsible for synthesis of geranylgeranyl pyrophosphate, a substrate
necessary for prenylation mediated by RGGT or GGTase 1. It cannot be excluded
though that these compounds inhibit both, RGGT and GGT-1, at similar
concentrations.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The α-fluorine analogs of 3-IPEHPC with the imidazo[1,2-a]pyridine ring substituted
in the position 6 were the most potent inhibitors of Rab11A prenylation as measured
by LED (10 μM for 1g, 1h; 25 µM for 1f, 1i; Table 2). At the same time
Rap1A/Rap1B prenylation was not affected (LED > 100 µM), except for 1h (LED =
100 µM). Our data suggest that they are even stronger selective RGGT inhibitors than
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
22
a reference (±)-3-IPEHPC (LED for inhibition of Rab11A = 25 µM or 50 µM in our
research model). Densitometry analysis confirmed their efficacy, showing at least 2.5-
fold enrichment of unprenylated Rab11A in the cytosol fraction of treated cells
compared to untreated control (Table 2). Compound 1c combining structural features
characteristics for 1i (bromine in the position 6) and 1b (methyl in the position 2) did
not inhibit Rab11A and Rap1A/Rap1B prenylation up to 100 µM. It confirmed that
C6 substitution of the imidazo[1,2-a]pyridine ring is the key and privileged position
among all analogs tested within this study, and simultaneous modification at other
positions reduces compound’ activity.
Similarly to the trend observed in viability assay (Table 1, 1i-E2: IC50 is 267 µM in
full medium and 130 µM in serum-free medium; no effect for 1i-E1), out of two
enantiomers of 1i, only one had the desired biological activity as measured by LED,
confirming previous findings that chirality plays a pivotal role in interaction between
phosponocarboxylates and RGGT.20
a
Table 2. Effects of selected PC analogs on Rab11A and Rap1A/Rap1B prenylation in
HeLa cells.
ACS Paragon Plus Environment

Page 19 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Densitometry
analyses of the
quantity of Rab11A
in cytosolic fractions
after treatment with
Inhibition of
Rab11A
Inhibition of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Rap1A/Rap1B
Compound
R1
R2
prenylation
prenylation
(LED/µM)
1
00 µM
(
LED/µM)
concentrations of
compounds (% of
control ± SD)b
1
.
3-
H
OH
5
0
258 ± 109
181 ± 81
129 ± 38
100
IPEHPCc
2
3
.
.
1a
2-Ph
OH
OH
100
no effect
no effect
2-Me-6-
Br
1
c
no effect
4
.
1e
1f
5-Me
6-Ph
6-Me
6-Cl
OH
OH
OH
OH
OH
OH
OH
100
25
10
10
25
10
155 ± 56
262 ± 102
335 ± 69
100
5.
no effect
no effect
100
6
7
.
.
1g
1h
312 ± 87
8.
1i
6-Br
6-Br
7-Me
357 ± 121
1539 ± 1143
no effect
no effect
9
1
.
1i-E2
0.
not
1
k
no effect
100
118 ± 20
180 ± 70
determined
1
1.
1o
8-Me
OH
100
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
2.
3.
2s
2t
H
H
Ph
Et
50
133 ± 36
143 ± 42
100
50
100
a
LED = lowest effective dose. No effect = LED is not included in concentration up to 100 µM.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
b
Densitometry analysis was performed with ImageLab™ Software (Bio-Rad) and relative unprenylated protein
band intensity was normalized to β-actin and quantified with respect to controls (untreated cells).
c
Inhibition of Rab11A prenylation (LED) by (S)-3-IPEHPC was determined to be 25 M.
Fractionation of prenylated and non-prenylated proteins prior to Western Blot
Triton X-114 is the nonionic mild detergent, commonly used to extract membrane
proteins from biological samples. It rarely denatures membrane proteins and allows
phase separation above its clouding point (about 22°C). It was previously used for cell
lysis and separation of hydrophobic prenylated proteins from their more polar
counterparts.22, 44 However, separation into hydrophilic and hydrophobic phases by
Triton X-114 extraction is not highly effective and may be substantially improved.
Our experience shows that the procedure based on the use of Triton X-114 is very
tedious, partly due to inefficient lysis of cells and temperature dependent partitioning,
both factors lowering protein yields and leading to contamination of cytosolic fraction
with membrane proteins. While this method should leave no doubt when applied to
4
4
proteins which are monoprenylated, e.g. by GGT-1 or FT, in case of RGGT, which
is an enzyme capable of mono- as well as double geranylgeranylation, the partition
between detergent-rich and aqueous phase may raise the question about localization
of not completely prenylated Rab GTPases. Similar problem may arise when using
activity assays, based on attachment to Rabs an isoprenoid analogs functionalized
either with a bioorthogonal, affinity or fluorescent handle. While this method is
efficient for protein identification, it can give ambiguous read-out for inhibitors of
ACS Paragon Plus Environment

Page 21 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
second prenylation, since both, the proper product of double geranylgeranylation as
well as the mono geranylgeranylated intermediate will be tagged and therefore
provide positive signal.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Therefore, knowing that previously studied phosphonocarboxylates are inhibitors of
second geranylgeranylation,19, 22 and based on our previous studies, where we have
shown that disruption of prenylation by phosphonocarboxylates leads to
mislocalization of Rab proteins into cytosol, instead of Triton X-114 we applied
commercial kit, dedicated to the enrichment of integral membrane proteins and
membrane-associated proteins from cultured mammalian cells and eliminates the
hassle of phase separation based on hydrophobicity. Here it was used for the first time
for fractionation of prenylated proteins and it turned out highly reproducible and
efficient for this type of studies.
22
Molecular docking
Since the crystal structure of phosphonocarboxylate-RGGT complex is not available,
we elaborated the model of their interactions using molecular docking and new
biological data. We determined that interactions of the acidic residues, phosphonic
and carboxylic groups, guide the location of the novel inhibitors in the RGGT binding
area. The suggested binding mode differs slightly from our previous study,21 but
depends now more on interactions that ensure that compounds are well stabilized into
the binding site. As these novel compounds are smaller and chemically distinct from
inhibitors of other classes that have been co-crystallized with RGGT,15-18 the binding
mode likely differs. In our induce fit docking simulations the phosphonic group of the
novel RGGT inhibitors coordinates to zinc ion (Fig. 5). In addition, docking suggests
that the phosphonic group could form a hydrogen bond with D238B. Furthermore,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
several amino acids, K235B, D238B, Y241B and K105A, are in hydrogen bonding
distance of the carboxylic acid group of the novel inhibitors (Fig. 5).
All the active compounds (1f, 1g, 1h and 1i) are suggested to have similar binding
mode (Fig. 5). Their substituents in the position 6 (Ph, Me, Cl, Br, respectively) face
W244B indole side chain yielding to favorable packing. H190B and Y107A are in
hydrogen bonding distance of N in the position 1 of the imidazo[1,2-a]pyridine ring.
However, when Me is added to the position 2 (compound 1c) this substitution
prevents K105A from forming a hydrogen bond to the carboxylic acid, affecting
disruptively the binding efficiency (Fig. 5; Table 2).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As far as chirality is concerned, we observed that one of the enantiomers of the active
compound (1i) is mainly responsible for racemate activity towards RGGT. Based on
previous studies when higher potency towards RGGT was detected for (S)-3-IPEHPC
2
0
isomer, we tentatively extended this stereochemical preference to fluorine analogs.
Such preference for (S)-isomer might come from fluorine interaction with C240B
and/or W244B, while in (R)-isomer there is no interaction with fluorine (see Figure
S294). Having two acceptors around, donor adjusts itself to both, which might be
responsible for stronger interaction than either one alone.
The methyl group present in other positions of the imidazo[1,2-a]pyridine ring, 5, 7 or
8 (compounds 1e, 1k and 1o, respectively) cause a slight shift in the binding mode
compared to e.g. the pose of compound 1g when they pursue hydrophobic interactions
with W52B, R144B, G192B, Y195B, W244B and F289B. Due to that shift, Y107A
cannot form hydrogen bond with N in the position 1 of the imidazo[1,2-a]pyridine
ring and thus some binding potential is lost.
The presence of a phenyl group in the position 2 of compound 1a strongly influences
the binding mode, making it very different from other analogs. It changes the
ACS Paragon Plus Environment

Page 23 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
interactions formed with the carboxylic acid and the phosphonic group of the
inhibitor. Still, the same amino acids, H190B, K235B, D238B, and Y241B, remain in
hydrogen bonding distance. W244B and F289B form a flat hydrophobic layer for
hydrophobic packing of both the imidazo[1,2-a]pyridine ring and the phenyl group.
However, hydrogen bonding with N in the position 1 in the imidazo[1,2-a]pyridine
ring is lost and with that also binding potential.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Based on docking simulations, inhibitor 2s binds in principle similarly as e.g.
compound 1c or 1g (Fig. 5) but phenyl group in the phosphinate shifts the binding
pose so that aromatic carbocycle forms π-π stacking interaction with F289B. With this
interaction some binding potential is gained. However, simultaneously H190B or
Y107A have difficulties to reach to hydrogen bonding distance of N in the position 1
of the imidazo[1,2-a]pyridine ring and thus some potential is likely lost. Basically,
compound 2t binds similarly as compound 2s. But because 2t has smaller,
hydrophobic substitution in the phosphinate part, the shift in the binding mode
towards F289B is larger and then H190B and Y107A have even bigger difficulties to
reach N in the position 1 in the imidazo[1,2-a]pyridine ring to form the hydrogen
bond.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. An example of the binding interactions of the novel RGGT inhibitors
suggested by docking simulations. The phosphonic and the carboxylic acid groups
guide the novel RGGT inhibitors to the close proximity of the catalytic zinc ion and
several amino acids with hydrogen bonding ability (K235B, D238B, Y241B and
K105A). In the binding of compound 1c (purple) and compound 1g (turquoise)
K105A has a crucial role. When methyl group is added to the position 2 (compound
1
c) this substitution prevents K105A from forming a hydrogen bond to the carboxylic
acid weakening the binding (Table 2). The black dashed lines indicate interactions
between the inhibitor and the protein. Used atom colors: C in compounds 1c and 1g
are in purple and in turquoise, respectively. O=red, N=blue, P=orange, S=yellow,
F=bright green, Zn=dark green, Br=burgundy.
ACS Paragon Plus Environment

Page 25 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Conclusions
We have elaborated convenient protocols for the synthesis of novel -fluoro and -
desoxy phosphonocarboxylates functionalized with heterocycle or polyphenyl
moieties. We have identified four phosphonocarboxylates (1f, 1g, 1h, 1i) among
them, which inhibit prenylation of Rab11A and show reduction of human HeLa
cancer cell line viability. Three compounds have shown selectivity towards RGGT
and no inhibition of geranylgeranylation of Rap1A/Rap1B protein up to 100 M
concentration. All active compounds bear a substituent in the same location of the
core scaffold, carbon C-6 of the imidazo[1,2-a]pyridine ring. Therefore, it was
designated as the leading position for future modifications. Such choice is supported
by molecular docking studies, which identified the coordination of phosphonic group
to zinc cation, stabilized by several amino acid residues being in the hydrogen
bonding distance to phosphonic and carboxylic groups as well as nitrogen atom in the
position 1 of imidazo[1,2-a]pyridine ring, as the guiding interaction in the complex.
We evaluated compounds’ potency by immunochemical detection of not completely
prenylated Rab11A proteins, taking into consideration the possible ambiguity of their
separation from fully digeranylgeranylated proteins. For the first time here we applied
the commercially available kit for fractionation of membrane from cytosolic proteins,
which in our hands exceeded previously used for this application Triton X-114, in
terms of efficiency of lysis and reproducibility.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Phosphonocarboxylates as inhibitors of second prenylation mediated by RGGT,
mostly affect digeranylgeranylated and not monogeranylgeranylated Rab proteins.
Since it is still difficult to determine how specific prenylated proteins mediate
different pathological processes, e.g. cancer cell growth and survival, such selectivity
should be advantageous. Considering that inhibition of RGGT affects over 60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
proteins, such broadness of targets pose a threat of high toxicity. It remains to be
determined if inhibition only of the second prenylation can lead to some salvage
effect on cell well-being and not be so deleterious to healthy cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The works on development of new analogs as more potent inhibitors of RGGT and
tools for studying Rab GTPase-associated processes combined with the studies on
elaboration of the efficient asymmetric synthesis of phosphonocarboxylates and
confirmation of the preferred stereochemistry for interaction with RGGT are in
progress now.
Experimental Section
General
Synthesis. All reagents were purchased from commercial sources and used as
obtained, unless specified otherwise. Thin-layer chromatography was performed
using silica gel 60 with F254 indicator on alumina plates. Compounds were purified
using Flash chromatograhy and 40-63 m silica gel. Solvent ratios for the purification
of compounds by flash chromatography are reported as volume ratios (v/v).
Preparative HPLC for purification of compounds 1s, 1t was performed using Gilson
Prep equipped with UV-VIS-156 detector and semipreparative column Kromasil 100-
5-C18 (5 m, 10 x 250 mm). Enantiomers of 1i were separated using Chiralpak QN-
AX column (Chiral Technologies Europe; 0.46 cm x 15 cm), according to the method
previously described for 3-IPEHPC.20 Each enantiomer was re-purified using the
same chiral column. Column was eluted isocratically (1 mL/min) with 0.7 M TEAAc
containing 75% MeOH at pH 5.8. The optical purity of enantiomers was evaluated
based on HPLC analysis using Chiralpak QN-AX column. The enantiomer with the
shorter retention time, 4.9 minute, has been labeled as E1, while the one with the
longer retention time, 5.7 minute, as E2.
ACS Paragon Plus Environment

Page 27 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
NMR spectra were measured at 250.13 or 700 MHz for H NMR, 62.90 or 170
13
31
MHz for C NMR, 283 or 101.30 MHz for P NMR on Bruker Avance DPX 250
and Bruker Avance II Plus 700 spectrometrs, respectively. Chemical shifts (δ) are
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
reported in parts per million (ppm) relative to: internal residual CHCl
3
in CDCl
3
(δ
1
13
7.26 H NMR) or CDCl
3
signal in C NMR (δ 77.16); internal residual HDO in D O
2
1
31
31
13
(
δ 4.79 H NMR) or external 85% H
3
PO (δ 0 ppm P NMR). P NMR and C
4
NMR spectra were proton-decoupled. Coupling constants (J) are quoted in Hz. The
assignment of the signals in H NMR and 13C NMR was supported by two-
dimensional experiments (COSY, HMQC, HMBC, DEPT135). In case of compounds
obtained as diastereomeric mixtures, the ratio of diastereomers was determined based
on 31P NMR and (if possible) sufficiently separated signals in H NMR. For
1
1
1
13
compounds 8 the description of H and C NMR shows only major (at least 80%)
isomer, (Z)-8, unless stated otherwise. For compounds 9s-t and 10s-t, chemical shifts
1
13
in H and C NMR are enlisted for both diastereoisomers, if sufficiently separated,
without distinguishing isomers. The purity of biologically tested compounds was
evaluated by elemental analysis. All target compounds are >95% pure.
®
Biological studies. PrestoBlue Cell Viability Reagent, Mem-PER™ Plus Membrane
Protein Extraction Kit, and all reagents for cell culture were purchased from Life
Technologies (Carlsbad, CA, USA). Mem-PERTM Plus Membrane Protein Extraction
Kit was used for cell lysis and separation of cytosolic and membrane-rich fractions.
Bradford Protein Assay and Clarity™ Western ECL Substrate were obtained from
Bio-Rad (Hercules, CA, USA). Protease inhibitor cocktail and lovastatin were
purchased from Sigma (Saint Louis, MO, USA). Primary antibodies against Rab11A
and Rap1A/Rap1B were obtained from Abcam (Cambridge, UK) and primary
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 94
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
antibodies against β-actin along with secondary HRP-linked antibodies were
purchased from Cell Signaling Technology (Beverly, MA, USA).
HeLa cell culture. The cervical epithelial carcinoma HeLa cell line was purchased
from the American Type Cell Collection (ATCC). Cells were cultured in Dulbecco’s
modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(FBS) containing 100 IU/ml penicillin and 100 μg/ml streptomycin. Cells were
incubated at 37°C in a humidified atmosphere of 95% air and 5% CO
2
.
For biological studies, PCs were dissolved in PBS at stock concentrations or 10 mM,
and the pH adjusted to about 7. Compounds were stored at 4 °C or -20 °C (for longer
storage) prior to use.
Molecular docking. Compounds were drawn with Maestro 2016.2 (Schrödinger
Release 2016-2: Maestro, Schrödinger, LLC, New York, NY, 2016). Both (/all)
stereoisomers were drawn for 11 compounds selected for molecular docking
(compounds in Table 2: 1a, 1c, 1e, 1f, 1g, 1h, 1i, 1k, 1o, 2s, 2t). First traditional
molecular docking with Glide was tested. Protein structure for docking, (PDB ID
1
6
45
4GTS) , was downloaded from the Protein Data Bank and prepared with Protein
Preparation Wizard. In docking, protein was held rigid except hydroxyl and thiol
groups were allowed to rotate. In ligand binding larger conformational changes in the
protein structure, according to the induced fit concept, are possible. In addition, the
binding area of RGGT is spacious and the sensitivity of traditional molecular docking
is not high enough. To better model the binding interactions of novel RGGT
inhibitors induced fit docking (IFD) in Maestro 2016.2 was performed. In IFD side
chains of amino acids in the defined distance from the ligand are allowed to move
freely and thus subtleties of the interactions between the ligand and the protein can be
considered more accurately than with traditional molecular docking. In IFD procedure
ACS Paragon Plus Environment

Page 29 of 94
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
center of the docking box was set to catalytic zinc ion coordinate since zinc ion is
considered important in binding. After the Glide molecular docking part of the IFD
procedure 20 poses were accepted for Prime refinement. Protein amino acid residues
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
6
were refined in 5 Å radius of the docking poses. OPLS3 force field was used in the
IFD procedure. In the end ten complexes were saved for the visual examination.
General procedure for hydrolysis. Synthesis of compounds 1 and 2. Compound 9/10
(
0.50 mmol) was dissolved in 12 M HCl (5.5 mL). After 4-5 h of reflux the mixture
was evaporated to dryness. The products were obtained as solids with 56-99% yields,
upon precipitation (using water or petroleum/diethyl ether) or by sole removal of
solvents.
2-Fluoro-3-(2-phenylimidazo[1,2-a]pyridin-3-yl)-2-phosphonopropanoic acid (1a).
Yield: 53%, scale: 0.093 g. Elemental analysis: C16
H
14FN
2
O
5
P (H O)1.35,
2
(C
2
H
5
OH)0.55, Calculated: C: 49.62, H: 4.87; N: 6.77, Found: C: 49.66; H: 4.88; N:
1
O pH 7) δ: 3.88 (ddd, 2JHH = 16.0, 3JPH = 8.7, 3JFH = 4.4,
6
.82, H NMR (700 MHz, D
2
CH
2
CPF, 1H), 4.17 (ddd, 3JFH = 37.3, 2JHH = 16.0, 3JPH = 1.3, CH
2
CPF, 1H), 7.02 (td,
3JHH = 6.9, 4JHH = 1.1, CHAr(6), 1H), 7.41 (ddd, 3JHH = 9.0, 3JHH = 6.9, 4JHH = 1.2,
CHAr(7), 1H), 7.50 (tt, 3JHH = 7.9, 4JHH = 1.2, CHAr, 1H), 7.58 (m, CHAr(8), CHPh, 3H),
7
.85 (dd, 3JHH = 8.2, 4JHH = 1.2, CHPh, 2H), 8.63 (bd, 3JHH = 6.9, CHAr(5), 1H), 31P
13
NMR (283 MHz, D
2
O pH 7) δ: 10.39 (d, 2JPF = 71.9), C NMR (176 MHz, D
2
O):
2
1
8.7 (bd, 2JFC = 22.0, CH
2
CPF, 1C), 99.7 (dd, 1JFC = 192.1, 1JPC = 140.0, CH
2
CPF,
C), 112.4 (s, CHAr(6), 1C), 115.1 (s, CHAr(8), 1C), 117.9 (dd, 3JPC = 12.7, 3JFC = 3.2,
C
Ar(3), 1C), 126.3 (s, CHAr(7), 1C), 126.5 (d, 5JPC = 5.8, CHAr(5), 1C), 128.0 (s, CHPh
,
,
1
C), 128.6 (s, CHPh, 2C), 128.9 (s, CHPh, 2C), 134.2 (s, CHPh, 1C), 142.8 (s, CAr(2)
1C), 145.1 (s, CAr(9), 1C), 176.3 (dd, 2JPC = 21.5, 2JFC = 3.3, CO
, 1C).
2
ACS Paragon Plus Environment