α,β-Methylene-ADP (AOPCP) Derivatives and Analogues: Development of Potent and Selective ecto-5′-Nucleotidase (CD73) Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.5b00802

ecto-5′-Nucleotidase (eN, CD73) catalyzes the hydrolysis of extracellular AMP to adenosine. eN inhibitors have potential for use as cancer therapeutics. The eN inhibitor α,β-methylene-ADP (AOPCP, adenosine-5′-O-[(phosphonomethyl)phosphonic acid]) was used as a lead structure, and derivatives modified in various positions were prepared. Products were tested at rat recombinant eN. 6-(Ar)alkylamino substitution led to the largest improvement in potency. N⁶-Monosubstitution was superior to symmetrical N⁶,N⁶-disubstitution. The most potent inhibitors were N⁶-(4-chlorobenzyl)- (10l, PSB-12441, K_i 7.23 nM), N⁶-phenylethyl- (10h, PSB-12425, K_i 8.04 nM), and N⁶-benzyl-adenosine-5′-O-[(phosphonomethyl)phosphonic acid] (10g, PSB-12379, K_i 9.03 nM). Replacement of the 6-NH group in 10g by O (10q, PSB-12431) or S (10r, PSB-12553) yielded equally potent inhibitors (10q, 9.20 nM; 10r, 9.50 nM). Selected compounds investigated at the human enzyme did not show species differences; they displayed high selectivity versus other ecto-nucleotidases and ADP-activated P2Y receptors. Moreover, high metabolic stability was observed. These compounds represent the most potent eN inhibitors described to date.

Full text of DOI:10.1021/acs.jmedchem.5b00802

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
#,#-Methylene-ADP derivatives and analogs: development of
potent and selective ecto-5#-nucleotidase (CD73) inhibitors
Sanjay Bhattarai, Marianne Freundlieb, Jan Pippel, Anne Meyer, Aliaa Abdelrahman, Amelie Fiene,
Sangyong Lee, Herbert Zimmermann, Gennady Yegutkia, Norbert Sträter, Ali El-Tayeb, and Christa E Müller
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00802 • Publication Date (Web): 06 Jul 2015
Downloaded from http://pubs.acs.org on July 7, 2015
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 56
Journal of Medicinal Chemistry
1
2
3
4
α,βꢀMethyleneꢀADP (AOPCP) derivatives and analogs:
development of potent and selective ectoꢀ5′ꢀnucleotidase
(CD73) inhibitors
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
§
§
‡
§
§
Sanjay Bhattarai, Marianne Freundlieb, Jan Pippel, Anne Meyer, Aliaa Abdelrahman, Amelie
§
§
#
x
‡
Fiene, SangꢀYong Lee, Herbert Zimmermann, Gennady G.Yegutkin, Norbert Sträter, Ali Elꢀ
§
,†
*,§
Tayeb, and Christa E. Müller
§
PharmaCenter Bonn, Pharmaceutical Institute, Pharmaceutical Chemistry I, University of Bonn, An der
Immenburg 4, Dꢀ53121 Bonn, Germany
‡
Institute of Bioanalytical Chemistry, Center for Biotechnology and Biomedicine, University of
Leipzig, Deutscher Platz 5, 04103 Leipzig, Germany
#
Institute of Cell Biology and Neuroscience, GoetheꢀUniversity, Frankfurt am Main, Germany
x
Medicity Research Laboratory, University of Turku, 20520 Turku, Finland
*Address correspondence to:
Dr. Christa E. Müller
Pharmazeutisches Institut
Pharmazeutische Chemie
An der Immenburg 4, Dꢀ53121 Bonn, Germany
Phone:
+49ꢀ228ꢀ73ꢀ2301
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 56
Fax:
+49ꢀ228ꢀ73ꢀ2567
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
Eꢀmail: christa.mueller@uniꢀbonn.de
†
On leave from the University of AlꢀAzhar, Assiut, Egypt
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
Abbreviations
ACN, acetonitrile; ADP, Adenosine diphosphate; AMP, Adenosine monophosphate; ATP, Adenosine
triphosphate; AOPCP, α,βꢀmethyleneꢀADP, adenosineꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid], [{5ꢀ
(6ꢀaminopurinꢀ9ꢀyl)ꢀ3,4ꢀdihydroxyoxolanꢀ2ꢀyl}methoxyhydroxyphosphoryl]methylphosphonic
acid;
CD73, cluster of differentiation 73; CE, capillary electrophoresis; DEAE, diethylaminoethyl; DEPT,
distortionless enhancement by polarization transfer; DMAP, 4ꢀdimethylaminopyridine; DMF, N,Nꢀ
dimethylformamide; DMSOꢀd , deuterated dimethyl sulfoxide; D O, deuterated water; EC, enzyme
6
2
commission; ESI, electrospray ionization; FPLC, fast protein liquid chromatography; GPI,
glycosylphosphatidylinositol; HPLC, high performance liquid chromatography; LCꢀMS, liquid
chromatographyꢀmass spectrometry; MeODꢀd , deuterated methanol; MOE, Molecular Operating
4
Environment; NMR, nuclear magnetic resonance; eN, ectoꢀ5′ꢀnucleotidase; eNPPs, ectoꢀnucleoside
pyrophosphatases/phosphodiesterases; eNTPDases, ectoꢀnucleoside triphosphate diphosphohydrolases;
pdb, protein data bank; POMs, polyoxometalates; PSB, Pharmaceutical Sciences Bonn; rt, room
temperature; SARs, structureꢀactivity relationships; SEM, standard error of the mean; SF9, Spodoptera
frugiperda 9; SI, supporting information, TEA, triethylamine; TEAC, triethylammonium
hydrogencarbonate; THF, tetrahydrofuran, TLC, thin layer chromatography, TNBC, tripleꢀnegative
breast cancers.
2
ACS Paragon Plus Environment

Page 3 of 56
Journal of Medicinal Chemistry
Abstract
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
Ectoꢀ5′ꢀnucleotidase (eN, CD73) catalyzes the hydrolysis of extracellular AMP to adenosine. eN
inhibitors have potential as cancer therapeutics. The eN inhibitor α,βꢀmethyleneꢀADP (AOPCP,
adenosineꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid]) was used as a lead structure, and derivatives
modified in various positions were prepared. Products were tested at rat recombinant eN.
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
6
6ꢀ(Ar)alkylaminoꢀsubstitution led to the largest improvement in potency. N ꢀMonoꢀsubstitution was
6
6
6
superior to symmetrical N ,N ꢀdiꢀsubstitution. The most potent inhibitors were N ꢀ(4ꢀchlorobenzyl)ꢀ
6
6
(10l, PSBꢀ12441, K 7.23 nM), N ꢀphenylethylꢀ (10h, PSBꢀ12425, K 8.04 nM), and N ꢀbenzylꢀ
i
i
adenosineꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] (10g, PSBꢀ12379, K 9.03 nM). Replacement of
i
the 6ꢀNH group in 10g by O (10q, PSBꢀ12431) or S (10r, PSBꢀ12553) yielded equally potent inhibitors
(10q, 9.20 nM; 10r, 9.50 nM). Selected compounds investigated at the human enzyme did not show
species differences; they displayed high selectivity versus other ectoꢀnucleotidases and ADPꢀactivated
P2Y receptors. Moreover, high metabolic stability was observed. These compounds represent the most
potent eN inhibitors described to date.
Keywords
AOPCP, blood serum, ectoꢀ5′ꢀnucleotidase, ectoꢀnucleotidase inhibitors, native enzyme, phosphonic
acids, recombinant enzymes, stability, synthesis, structureꢀactivity relationships
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 56
Introduction
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
Ectoꢀ5′ꢀnucleotidase (ectoꢀ5′ꢀNT, eN, CD73, EC 3.1.3.5) is a member of the group of ectoꢀnucleotidases
1
which dephosphorylate extracellular nucleotides. eN catalyzes the dephosphorylation of nucleoside
1
, 2
monophosphates, and its main substrate is AMP. Other members of the membraneꢀbound group of
ectoꢀnucleotidases are the nucleoside triphosphate diphosphohydrolases (NTPDases; subtypes 1ꢀ3 and
8), the nucleotide pyrophosphatase/phosphodiesterases (NPPs 1ꢀ4) and the alkaline phosphatases (APs;
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
3ꢀ8
tissue nonꢀspecific, intestinal, placental and germ cell). NTPDases and NPPs are ATPꢀ and ADPꢀ
hydrolyzing ectoꢀnucleotidases, which reduce the concentration of extracellular ATP and/or ADP and
6
thus the activation of P2X and P2Y receptors by these nucleotides. eN elevates extracellular
7
concentrations of adenosine. The membraneꢀbound vertebrate enzyme hydrolyzes 5′ꢀAMP to adenosine
and phosphate but cannot hydrolyze adenosine 2′ꢀ and 3′ꢀmonophosphates. The hydrolysis of 5′ꢀAMP is
1
, 8
stereoselective, i.e., only the Dꢀenantiomer is a substrate.
2+
eN is a Zn ꢀbinding glycosylphosphatidylinositolꢀ (GPIꢀ) anchored protein, with its catalytic domain
7
facing the extracellular medium. It can be cleaved off the anchor and released as a soluble protein.
Eukaryotic eN functions as a noncovalent homodimer with two structural domains: the Nꢀterminal and
9, 10
the Cꢀterminal domain are connected by a small hinge region.
This hinge region enables the enzyme
9
to undergo large domain movements and switches between the open and closed conformations. Recent
studies on eN knockout mice and eN overexpressing tissues have provided evidence for possible
1
1ꢀ13
therapeutic applications of eN inhibitors in various pathological conditions.
eN and adenosine A2A
receptors have been found to be coexpressed in various tissues and can thereby activate the receptors in
9
a concerted action. 2ꢀSubstituted AMP derivatives, which serve as eN substrates and are converted to
A adenosine receptor agonists, can serve as siteꢀspecific A agonist prodrugs that are targeted to sites
2A
2A
1
2, 13
of elevated eN expression, e.g. under inflammatory conditions.
eN inhibitors have potential applications as novel therapeutics for melanomas, lung, prostate and
1
4ꢀ16
breast cancers.
eN favors tumor cell growth by inhibiting T cell activity and by promoting
4
ACS Paragon Plus Environment

Page 5 of 56
Journal of Medicinal Chemistry
1
7, 18
angiogenesis.
It has been shown that inhibition of eN with a monoclonal antibody, siRNA, or small
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
9, 20
molecules delays tumor growth and metastasis.
The same effects were observed in eN knockout
2
1
mice due to decreased adenosine production. High eN expression has been reported in tripleꢀnegative
breast cancers (TNBC), and it has also been demonstrated that targeted blockade of CD73 significantly
1
6
prolonged the survival in anthracyclineꢀresistant animal models of cancer.
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
1
5ꢀ18
Inhibitors of eN have therefore potential as novel drugs for cancer therapy.
eN inhibitors reduce
extracellular adenosine levels, resulting in an indirect blockade of adenosine (P ) receptor activation. In
1
contrast to direct receptor ligand interaction, enzyme inhibitors are indirect antagonists, which will
exhibit siteꢀ and eventꢀspecific effects since they are only active in the presence of enzyme, substrate,
and receptor. ADP and ATP are competitive inhibitors of the enzyme, with inhibition constants in the
1
, 6, 8
micromolar range.
enzymatic degradation by NTPDases, NPPs and alkaline phosphatases.
(phosphonomethyl)phosphonic acid] or [{5ꢀ(6ꢀaminopurinꢀ9ꢀyl)ꢀ3,4ꢀdihydroxyoxolanꢀ2ꢀ
yl}methoxyhydroxyphosphoryl]methylphosphonic acid (I, α,βꢀmethyleneꢀADP, AOPCP), a more stable
However these inhibitors are not ideal, since they themselves are subjected to
2
0, 21
Adenosineꢀ5′ꢀOꢀ
[
2
2
analog of ADP, is one of the most potent competitive inhibitors of eN known to date. In addition to
2
4
25
26
this nucleotide analog, only anthraquinones, sulfonamides, various polyphenols and some
2
7
polyoxometalates (POMs) are currently known to potently inhibit eN. Among them the most potent
24
inhibitor is the anthraquinone derivative IV (Figure 1). Anthraquinones and sulfonamides were found
to display a competitive mechanism of inhibition whereas polyphenols and POMs were shown to be
nonꢀcompetitive inhibitors.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 56
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 1. Selected inhibitors of ectoꢀ5′ꢀnucleotidase: nucleotide, anthraquinone, sulfonamide and
polyphenol derivatives. Compounds II (ADP) and III (ATP) are physiological inhibitors, α,βꢀ
methyleneꢀADP (I) is an ADP analogue with increased metabolic stability compared to ADP,
2
4
25
compound IV is an anthraquinone derivative, V and VI are sulfonamide and polyphenolic
26
a
flavonoid derivatives, respectively. Unpublished data from our laboratory. Results obtained by our
group in the same assay are depicted in red color, while results obtained in different assays are shown in
blue color.
The number of structureꢀactivity relationship studies (SARs) of eN inhibitors is very limited. We
therefore started a program to develop potent eN inhibitors based on nucleotide as well as nonꢀ
nucleotide scaffolds. However, our efforts to improve the potency of nonꢀnucleotide based eN inhibitors
24,25
have so far met with limited success.
Thus, in the present study, we chose the competitive inhibitor
AOPCP as a lead structure, for which, to the best of our knowledge, no SARs have been described.
2
2,23
AOPCP is the most potent and selective eN inhibitor known to date.
It has a much higher metabolic
stability as compared to the physiological eN inhibitors ADP and ATP due to replacement of the diꢀ (or
6
ACS Paragon Plus Environment

Page 7 of 56
Journal of Medicinal Chemistry
8
,24,25
triꢀ) phosphate by a methylenediphosphonate residue.
AOPCP derivatives and analogs will be
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
deprotonated at physiologic pH values and display high waterꢀsolubility making them suitable for in
vivo target validation studies, and also for parenteral application in cancer therapy. Moreover, the
crystal structure of AOPCP in complex with the human enzyme has been published (pdb: 4H2I) and
9
provides information for the design of derivatives and analogs with improved potency. Moreover, we
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
designed and synthesized compounds which were expected to exhibit high metabolic stability and will
not be metabolized to adenosine receptor agonists.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 56
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
Results and Discussion
Chemistry
6
6
6
For the preparation of AOPCP derivatives 9aꢀd and 10aꢀw substituted at N , O , S and Cꢀ8 a
convergent synthetic strategy was applied. It started with the synthesis of the intermediate nucleosides
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
(6aꢀn, 6pꢀr, 6w, 6v, 7aꢀd, 8b, 8c) which were subsequently reacted with methylenebis(phosphonic
dichloride) in trimethyl phosphate (see Schemes 1–3, for more details see Supporting Information). Diꢀ
substituted derivative 10x was synthesized from intermediate nucleoside 8d by reaction with
methylenebis(phosphonic dichloride) (see Scheme 4). Further derivatives (12aꢀc) with substitution in
the methylenebisphosphonate sideꢀchain of AOPCP were synthesized by nucleophilic displacement of
2
′,3′ꢀOꢀisopropylideneꢀ5′ꢀOꢀtosylꢀadenosine (11b) with trisꢀ(tetraꢀnꢀbutylammonium) salts of
substituted bisphosphonic acids (see Schemes 5).
6
For the preparation of N ꢀsubstituted intermediate nucleosides (6aꢀn), 6ꢀchloropurine riboside was used
28, 29
which was synthesized from inosine according to a reported procedure
with minor modifications
(see Schemes 1). The chlorination of the keto function at the 6ꢀposition of inosine requires protection at
the 2′ꢀ, 3′ꢀ, and 5′ꢀhydroxyl groups as they are all susceptible for chlorination. There are several reported
protecting groups such as acetyl or benzoyl residues. We chose acetyl groups for protection as they can
be conveniently removed under mild alkaline conditions. Protection of the 2′ꢀ, 3′ꢀ and 5′ꢀhydroxyl
groups of inosine (1) was carried out using acetic anhydride and pyridine. Chlorination of the protected
2
8
derivative 3 was achieved using phosphorus oxychloride in the presence of N,Nꢀdimethylaniline.
Subsequent deprotection was performed using a solution of 7N ammonia in methanol for 24 hours
yielding 4 in good yield of 70 %.
6
N ꢀsubstituted adenosine derivatives 6aꢀn were obtained by reaction of 6ꢀchloropurine riboside (4) with
NꢀmonoalkylamineꢀHCl, monoꢀar(alk)ylamine, N,NꢀdialkylamineꢀHCl, or N,Nꢀdiar(alk)ylamine,
30, 31
respectively.
Phosphorylation in the 5′ꢀposition of nucleosides proceeds without sideꢀproduct
formation at the 2′ꢀ and 3′ꢀpositions if the 2′ꢀ and 3′ꢀhydroxyl groups are protected. Frequently, the
8
ACS Paragon Plus Environment

Page 9 of 56
Journal of Medicinal Chemistry
nucleosides are treated with pꢀmethoxybenzaldehyde in tetrahydrofuran (THF) in the presence of zinc
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
3
2
6
chloride yielding acetalꢀprotected nucleosides. However, protection of the N ꢀsubstituted nucleosides
with pꢀmethoxybenzyldehyde did not work in our hands. Therefore we introduced the classical 2′,3′ꢀOꢀ
isopropylidene protecting group using acetone and 2,2ꢀdimethoxypropane under strong acidic condition
3
3
to give nucleosides 7aꢀd.
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
6
Scheme 1. Synthesis of inosineꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] and N ꢀmonoꢀar(alk)ylamine
6
6
a,b
or N ,N ꢀdiar(alk)ylꢀpurine ribosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid]
O
N
O
N
Cl
N
N
N
N
N
N
N
NH
NH
N
HO
AcO
a
b
AcO
O
O
O
OH OH
OAc OAc
OAc OAc
1
2
3
R1
R2
c
f
N
Cl
N
O
N
N
N
N
d
N
N
N
NH
N
O
N
O
N
HO
O
P
O
P O
HO
O
HO
OH OH
OH OH
OH
OH
OH OH
4
6
a-n
10o
f
e
R1
R1
R2
R2
R1
R2
N
N
N
N
N
N
N
N
N
N
N
N
O
P
O
N
g
N
HO
O
O
f
N
P O
O
O
HO
P
P O
OH OH
O
HO
OH OH
O
O
O
O
OH OH
0a-n
1
9a-d
7a-d
R1
R2
R1
R2
H
compd
compd
a
b
c
d
e
f
methyl
ethyl
H
H
h
i
2-phenylethyl
benzyl
ethyl
benzyl
H
methyl
ethyl
methyl
ethyl
methyl
H
j
benzyl
k
l
4-aminobenzyl
4-chlorobenzyl
4-nitrobenzyl
ethyl
H
phenyl
benzyl
m
n
H
g
H
4-sulfamoylbenzyl
H
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 56
a
o
Reagents and conditions: (a) acetic anhydride, dry pyridine, room temp, reflux, 80 C, 3 h, yield 98%;
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
o
(
b) POCl , N,Nꢀdimethylaniline, reflux,110 C, 20 min, yield 82%; (c) 7N ammonia, methanol, rt, 24 h,
3
o
yield 70%; (d) N,NꢀdialkylamineꢀHCl, triethyl amine, dry DMF, 0 C to rt, 20 h, yield 90ꢀ95%; or Nꢀ
o
monoꢀar(alk)ylamine or N,Nꢀdiar(alk)ylamine, triethyl amine, absolute alcohol, reflux, 60 C, 20 h,
yield 70ꢀ95%; (e) acetone, 2,2ꢀdimethoxypropane, H SO , rt, 30 min, yield 85ꢀ90%; (f) for compounds
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
2
4
o
9
aꢀd and 10eꢀo two steps: (i) methylenebis(phosphonic dichloride), trimethyl phosphate, 4 C, 30 min;
b
(
ii) triethylammonium hydrogencarbonate buffer pH 7.4ꢀ7.6, rt, 15 min, yield 40ꢀ68%; (g) for
b
compounds 10aꢀd: triflouroacetic acid (6ꢀ8%), CH Cl , H O, rt, 4 h, yield 55ꢀ75%. For details see
2
2
2
Supporting Information.
For the preparation of nucleosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] there are several
commonly used multiꢀstep methods: (i) the protected nucleoside may be reacted with activated
33, 34
bisphosphonates, e.g. (4ꢀnitrophenyl)ethyl phosphonates,
(ii) protected nucleoside may be reacted
3
5
with phosphonates via a Mitsunobu reaction, (iii) protected nucleoside may be esterified with
phosphonic acid in trichloroacetonitrile in the presence of dicyclohexylcarbodiimide, or (iv) the
protected nucleosideꢀ5′ꢀsulfonyl ester may be prepared followed by nucleophilic substitution with
3
6
alkylammonium salts of bisphosphonic acid. All of these methods utilize protected nucleosides and
afford products in only low yields. In the early 2000s two separate groups have reported the preparation
37, 38
of nucleosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid]
by phosphorylation of nucleosides with
methylenebis(phosphonic dichloride) followed by hydrolysis with TEAC buffer (Triethylammonium
3
9
hydrogencarbonate). The formation of the nucleoside phosphonate proceeds via intermediates (i.e.
nucleosideꢀOꢀ[{chloro(methyl)phosphoryl}phosphonic dichloride]); in addition to 5′ꢀphosphonylation
there is an equal risk of reaction at the 2′ꢀ and 3′ꢀposition. Apart from this, there is also a high
probability of the formation of nucleosideꢀ3′,5′ꢀcyclomethylenebis(phosphonic acid) and dinucleosideꢀ
bisphosphonic acid derivatives. The formation of many byꢀproducts will complicate the purification
process. Therefore we decided to start from the 2′,3′ꢀOꢀisopropylideneꢀprotected nucleosides to prevent
10
ACS Paragon Plus Environment

Page 11 of 56
Journal of Medicinal Chemistry
2
′ꢀ and 3′ꢀphosphonylation and the formation of cyclic phosphonates. To increase the yield of
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
nucleosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] formation using protected nucleosides we used
two different optimization strategies, (i) the amount of methylenebis(phosphonic dichloride) used during
the reaction, or (ii) the reaction time. Optimized reaction conditions included the use of 5 equivalents of
methylenebis(phosphonic dichloride) and a reaction time of 30 min, followed by hydrolysis with TEAC
producing only nucleosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] as a single product. (Scheme 1,
for detailed information on the optimization of the reaction see Supporting Information). The protected
nucleotides 9aꢀd were deprotected by trifluoroacetic acid to give nucleotides 10aꢀd in good yields.
While the initial phosphorylation reactions were carried out using protected nucleosides, we were able
to employ the same optimized reaction conditions with unprotected nucleosides 6eꢀn yielding
nucleosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] as the main product 10eꢀn. Nucleotide 10o was
synthesized from the inosine (1). All final compounds were purified by anion exchange chromatography
and/or reverse phase C18 HPLC.
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
For the synthesis of 6ꢀ(ar)alkoxyꢀsubstituted purine nucleoside derivatives (6p, 6q) 6ꢀchloropurineꢀβꢀDꢀ
riboside (4) was reacted with sodium (ar)alkoxide in the corresponding (ar)alkyl alcohol according to a
4
0, 41
reported procedure
(see Scheme 2). The 6ꢀbenzylthioꢀsubstituted derivative (6r) was also
2
9, 42
synthesized from 6ꢀchloropurine riboside (4), first converting it into 6ꢀthiopurineꢀriboside
followed
by reaction with benzyl chloride. 6ꢀChloro, 6ꢀ(ar)alkoxyꢀ and 6ꢀbenzylthioꢀpurine ribosides were
phosphonylated as described above by reaction with methylenebis(phosphonic dichloride) followed by
hydrolysis with TEAC (see Scheme 2). The optimized reaction conditions yielded mainly the desired
products which were purified by reverse phase C18 HPLC.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 56
Scheme 2. Synthesis of 6ꢀsubstitutedꢀpurineβꢀDꢀribofuranosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic
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
a,b
acid] derivatives
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
a
Reagents and conditions: (a) 1.0 M sodium (ar)alkoxide in the corresponding (ar)alkyl alcohol,
o
reflux,100 C, 20 h, yield 70ꢀ80%; (b) two steps: (i) methylenebis(phosphonic dichloride), trimethyl
o
phosphate, 4 C, 30 min; (ii)TEAC buffer pH 7.6, rt, 15 min, yield 40ꢀ65%;(c) thiourea, ethanol, 2 h, rt,
o
yield 90%; (d) benzyl chloride, K CO microwave, 100 C, 200W, 10 min, yield 55%.
2
3,
b
For details see Supporting Information.
For the synthesis of 8ꢀsubstituted derivatives (10tꢀw), commercially available adenosine (8a) was
brominated using bromineꢀwater solution according to a reported method to give 8ꢀbromoadenosine
43
(
8b). 8ꢀThioadenosine was prepared from 8ꢀbromoadenosine (8b) by reaction with thiourea and was
2
9, 42
used without further purification.
8ꢀChloroadenosine (8c) was prepared from 8ꢀthioadenosine by
reaction with Nꢀchlorosuccinimide (see Scheme 3). 8ꢀMethylaminoadenosine (6v) was synthesized
44,45
from 8ꢀchloradenosine by treating it with methylamine.
8ꢀThioadenosine was reacted with ethyl
4
6
iodide to give 8ꢀethylthioadenosine (6w). These 8ꢀsubstituted adenosine derivatives were subsequently
reacted applying the earlier optimized method using 5 equivalents of methylenebis(phosphonic
dichloride) to give nucleoside 5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] as the final products 10tꢀw
12
ACS Paragon Plus Environment

Page 13 of 56
Journal of Medicinal Chemistry
(
Scheme 3). The final phosphorylation reaction yielded mainly the desired products with very low
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
amounts of sideꢀproducts. The products were purified by anion exchange chromatography followed by
reverse phase C18 HPLC. The yields of 8ꢀsubstituted nucleotides were low compared with those of 6ꢀ
substituted nucleotides (for details see Supporting Information).
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
Scheme 3. Synthesis of 8ꢀsubstitutedꢀadenosineꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid]
derivatives
a,b
a
Reagents and conditions: (a) bromineꢀH O, sodiumꢀacetate buffer, pH 4.0, rt, 4 h, yield 96%; (b) two
2
steps: (i) thiourea, ethanol, 2 h, rt; (ii) Nꢀchlorosuccinimide, MeOH, rt, 4 h, yield 52%; (c) methylamine,
TEA DMF, 20 h, rt, yield 93%; (d) Two steps: (i) thiourea, ethanol, 2 h, rt; (ii) ethyl iodide, NaOH,
,
waterꢀethanol (1:1), 2 h, rt, yield 56%; (e) Two steps: (i) methylenebis(phosphonic dichloride), trimethyl
o
b
phosphate, 4 C, 30 min, (ii)TEAC buffer pH 7.6, rt, 15 min, yield 35ꢀ45% ( for details see Supporting
Information).
The disubstituted nucleotide derivative 10x was synthesized by reaction of nucleoside 8d with
methylenebis(phosphonic dichloride) applying the previously optimized reaction conditions.
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 56
6
a,b
Scheme 4. Synthesis of N ꢀbenzylꢀ8ꢀbromopurine ribosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid]
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
HN
N
HN
N
HN
N
N
N
N
N
N
N
N
N
N
Br
Br
O
P
O
a
b
HO
HO
P O
O
O
O
HO
OH OH
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
OH OH
OH OH
OH OH
6
g
8d
10x
a
Reagents and conditions: (a) bromineꢀH O, sodiumꢀacetate buffer, pH 4.0, rt, 4 h, yield 96%; (b) two
2
o
steps: (i) methylenebis(phosphonic dichloride), trimethyl phosphate, 4 C, 30 min, (ii)TEAC buffer pH
b
7
.6, rt, 15 min, yield 35% ( for details see Supporting Information).
The reactions for the synthesis of methylene diphosphonate sideꢀchain modified analogs (12aꢀc) were
also started from the commercially available adenosine (8a). Adenosine was protected on the 2′ꢀ and the
3
′ꢀhydroxyl group to give 2′ꢀ,3′ꢀOꢀisopropylideneꢀadenosine (11a) followed by tosylation at the 5′ꢀ
position to yield intermediate 11b. Substituted bisphosphonic acid derivatives were converted to the
corresponding tris(tetraꢀnꢀbutylammonium)salts by trituration with tetraꢀnꢀbutylammonium hydroxide.
Then
trisꢀ(tetraꢀnꢀbutylammonium)
dibromomethylenebis(phosphonic
acid),
trisꢀ(tetraꢀnꢀ
butylammonium) dichloromethylenebis(phosphonic acid), or trisꢀ(tetraꢀnꢀbutylammonium)ꢀ1ꢀ
hydroxyethaneꢀ1,1ꢀdiyldiphosphonic acid, respectively, was reacted with 2′ꢀ,3′ꢀOꢀisopropylideneꢀ
adenosine 5′ꢀOꢀtosylate (11b) providing the phosphorylated intermediates 11cꢀe. After purification of
the intermediates they were subsequently deprotected by treatment with 6ꢀ8% trifluoroacetic acid in
3
4ꢀ36
dichloromethane, to obtained desired products 12aꢀc.
All diphosphate sideꢀchainꢀsubstituted
nucleotides were purified by anion exchange chromatography followed by reverse phase C18 HPLC.
Product 12c was obtained as a mixture of diastereomers, which was not separated.
14
ACS Paragon Plus Environment

Page 15 of 56
Journal of Medicinal Chemistry
a,b
Scheme 5. Synthesis of α,βꢀsubstitutedꢀmethyleneꢀADP derivatives
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
a
Reagents and conditions: (a) acetone, 2,2ꢀdimethoxypropane, H SO , rt, 30 min, yield 85ꢀ90%; (b)
2
4
pꢀtolenesulfonyl chloride, DMAP, pyridine, rt, 14 h, yield 50%; (c) for compound 12a two steps: (i)
tris(tetraꢀnꢀbutylammonium) dibromomethylenebis(phosphonic acid), DMF, rt, 24 h, yield 40%; (ii)
6
ꢀ8% TFA, CH Cl , H O, 4 h, rt, yield 75%; (d) for compound 12b two steps: (i) tris(tetraꢀnꢀ
2 2 2
butylammonium) dichloromethylenebis(phosphonic acid), DMF, rt, 24 h, yield 40%; (ii) 6ꢀ8% TFA,
CH Cl , H O, 4 h, rt, yield 75%; (e) for compound 12c two steps: (i) tris(tetraꢀnꢀbutylammonium)ꢀ1ꢀ
2
2
2
hydroxyethaneꢀ1,1ꢀdiyldiphosphonic acid, DMF, 30 h, yield 35%; (iii) 6ꢀ8% TFA, CH Cl , H O, 4 h, rt,
2
2
2
b
yield 75%. For details see Supporting Information.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 56
Pharmacological Evaluation
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
3
The compounds’ potency to inhibit eN was determined by a radiometric eN assay using [ H]AMP as a
3
substrate, which was separated from the enzymatic product [ H]adenosine by precipitation with
22
lanthanum chloride followed by filtration through glass fiber filters. Recombinant rat eN expressed in
4
7
Sf9 insect cells was employed. For all compounds full concentrationꢀresponse curves were determined
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
in at least three separate experiments using 10 different concentrations. K values were calculated from
i
48, 49
the obtained IC values using the ChengꢀPrusoff equation.
Results are summarized in Tables 1ꢀ4,
5
0
and selected concentrationꢀinhibition curves are shown in Figure 2.
16
ACS Paragon Plus Environment

Page 17 of 56
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 2. Concentrationꢀinhibition curves of selected compounds at rat eN. (A) Compounds 10b,
1
0e,10f, 10g, 10h, compared to lead structure I (AOPCP); (B) Compounds 10l, 10m, 10q, 10s
compared to I. Rat enzyme K , 59 ꢁM; AMP concentration, 5 ꢁM. Data points are from three separate
m
experiments performed in duplicate. For K values see Tables 1ꢀ4.
i
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 56
For two of the most potent compounds inhibitory potency at human eN was additionally determined
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
using a radiometric assay based on separation of substrate and products by thin layer chromatography
5
0
(TLC). Measurements were performed on recombinant human enzyme and also on native eN present
51
in human blood serum (see below).
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
Structureꢀactivity relationships
A series of 31 derivatives (compounds 9aꢀd, 10aꢀx and 12aꢀc) was evaluated for inhibition of rat eN
(see Tables 1, 2, 3 and 4).
The lead compound AOPCP (I), a competitive inhibitor of rat eN, showed a K value of 197 nM
i
6
determined under the same conditions. N ꢀsubstitution generally increased potency of the lead structure
I. Monoꢀsubstitution displayed the following rank order of potency: 2ꢀphenylethyl (10h, K 8.04 nM) ~
i
benzyl (10g, 9.03 nM) > phenyl (10f, 36.8 nM) ≥ ethyl (10b, 43.8 nM) > methyl (10a, 104 nM) > H (I,
6
197 nM). These results showed that large, hydrophobic N ꢀsubstituents were beneficial for high
6
inhibitory potency at eN. N ꢀDisubstitution was also tolerated. Among the symmetrically disubstituted
derivatives the N,Nꢀdiethylꢀsubstituted compound (10d, 68.0 nM) was slightly superior to the dimethyl
6
derivative 10c (86.0 nM) and the dibenzylꢀsubstituted compound 10j (92.5 nM). However, the N ꢀ
monoꢀethyl derivative was superior to the diethylꢀsubstituted compound (compare 10b/10d), and the
6
difference was even more pronounced for the larger N ꢀmonoꢀbenzylꢀsubstituted derivative 10g in
6
comparison with the N ꢀdiꢀbenzyl derivative 10j (ca. 10ꢀfold difference). These results indicate that only
6
one large N ꢀsubstituent was well tolerated, but not two voluminous residues at the exocyclic N atom. In
6
6
contrast, in the case of N ꢀmethyl substitution, the N ꢀdisubstituted derivative 10c was even somewhat
more potent than the monoꢀmethylꢀsubstituted compound 10a indicating that a second small substituent
6
at N was well tolerated and might even increase potency. We therefore prepared compounds with two
6
6
different N ꢀsubstituents, 10e and 10i. In fact, combination of an ethyl and a methyl substituent at N
(10e) increased potency as compared to the monoꢀethylꢀsubstituted compound 10b by about 2ꢀfold (10e,
18
ACS Paragon Plus Environment

Page 19 of 56
Journal of Medicinal Chemistry
6
2
3.6 nM). In case of 10i, in which a benzyl residue was combined with an ethyl substituent at N , a large
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
drop in affinity was observed in comparison with the monoꢀbenzylꢀsubstituted derivative 10g (10i, 76.4
6
nM, 10g, 9.03 nM). These results indicate that disubstitution at N may increase the compounds’
6
potency, but while one of the N ꢀsubstituents may be large, the size of the second one is more limited
and should preferably be a methyl group.
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
6
Since the N ꢀbenzylꢀAOPCP 10g had so far been one of the most potent derivatives, we investigated
compounds with various paraꢀsubstituted benzyl residues (10k, 10l, 10m, 10n). The following rank
6
order of potency for N ꢀsubstituents was observed: pꢀchlorobenzyl (10l, 7.23 nM) ≥ benzyl (10g, 9.03
nM) > pꢀsulfamoylbenzyl (10n, 14.2 nM) ≥ 4ꢀnitrobenzyl (10m, 17.6 nM) > 4ꢀaminobenzyl (10k, 29.0
nM). pꢀSubstituents were tolerated but did not lead to a significant increase in inhibitory potency.
6
To rationalize the effects of the N ꢀsubstituents, we studied possible interactions of the substituents by
molecular modelling based on the crystal structure of human eN in complex with AOPCP in the closed
conformation (pdb: 4H2I). Rat and human eN display 90% sequence identity and their active sites differ
in only one amino acid: F500, which is involved in the nucleotide base stacking in human eN, is
replaced by a tyrosine residue in the rat enzyme. In the human eN crystal structure, the terminal
phosphate group of AOPCP is coordinated by two zinc ions and the nucleobase is sandwiched via πꢀ
stacking interactions between the side chains of F417 and F500. In addition, a hydrogen bonding
network with water molecules and side chains of the Nꢀterminal as well as the Cꢀterminal domain
9
further stabilizes the interaction.
6
The 6ꢀamino group (N ) of AOPCP points towards the solvent and is not involved in any interaction
9
with the protein or water molecules. Therefore, small as well as large substituents are not expected to
interfere with adjacent residues important for binding of AOPCP. Residues L184, S185, N186, D121,
6
and N122 of the Nꢀterminal domain of eN are at a distance of 4 to 11 Å to the N and thus constitute
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 56
6
potential interaction partners for the N ꢀsubstituents. Figure 3 shows the modeled binding mode of the
most potent compound 10l. While the overall binding mode of the AOPCP scaffold was stable in the
energy minimization, further hydrophobic interactions might be formed by the benzyl group with the
backbone atoms of L184 and S185 of the Nꢀterminal domain. This also applies to other voluminous, but
more flexible hydrophobic substituents that might interact additionally with D121, N122 or N186
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 3). Furthermore, paraꢀsubstituents at the aromatic ring (10kꢀn) may result in polar interactions
with the carbonyl oxygen of S185, the side chain of N186, or water molecules. Smaller hydrophobic
6
groups at N (compounds 9aꢀ10e) also result in increased binding affinity, probably via hydrophobic
interactions with F417 and F500. The modeled binding mode of 10j shows that both aromatic rings at
6
N have enough space without causing steric clashes, but only one benzyl group is involved in favorable
hydrophobic interactions with the protein (Figure 3).
Figure 3. Modeled binding modes of 10l (left) and 10j (right) in the substrate binding site of human eN.
Residues of the Nꢀterminal and Cꢀterminal domain of human eN are shown in blue and green,
respectively (pdb: 4H2I was used for modelling).
20
ACS Paragon Plus Environment

Page 21 of 56
Journal of Medicinal Chemistry
1
2
We also tested a few compounds that were modified at the ribose moiety: the 2′,3′ꢀisopropylideneꢀ
substituted nucleotides 9aꢀd (see Table 1), that had been obtained as intermediate products during the
preparation of 10aꢀd. The isopropylideneꢀsubstituted compounds were 5ꢀ to 37ꢀfold less potent than
their unsubstituted counterparts. Unsubstituted 2′,3′ꢀhydroxyl groups appear to be important, since they
form hydrogen bonds with aspartate and arginine residues in the substrate binding site of the enzyme as
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
1
0, 24, 25, 56
evidenced by crystal structures (pdb: 4H2I).
The isopropylideneꢀsubstituted nucleotides
6
showed a different rank order of potency with regard to the N ꢀsubstitution than that observed with the
compounds with free hydroxyl groups in the 2′ꢀ and 3′ꢀposition. This indicates different binding modes
of the 2′,3′ꢀisopropylideneꢀsubstituted as compared to the unsubstituted nucleotides. The most potent
6
compound among the isopropylideneꢀsubstituted nucleotides was the N ꢀethylꢀderivative 9c (365 nM),
which was 8ꢀfold weaker than its analog 10b (43.8 nM).
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 56
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
Table 1. Inhibitory potency of N ꢀsubstitutedꢀpurine ribosideꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic
acid] derivatives at rat ectoꢀ5′ꢀnucleotidase
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
Rat eN
1
2
a
Compd.
R
R
K
i
± SEM (nM)
I, AOPCP
H
methyl
H
H
197 ± 5
3870 ± 918
365 ± 31
9
a
9
9
c
c
ethyl
H
methyl
methyl
ethyl
H
415 ± 60
9
d
ethyl
679 ± 49
1
0a
0b
0c
0d
0e
0f
0g
0h
methyl
104 ± 13
1
ethyl
H
43.8 ± 0.3
86.0 ± 19.1
68.0 ± 5.1
23.6 ± 4.9
36.8 ± 4.7
9.03 ± 1.24
8.04 ± 2.24
76.4 ± 5.0
92.5 ± 8.5
29.0 ± 1.7
7.23 ± 0.78
17.6 ± 1.5
14.2 ± 1.7
1
methyl
methyl
ethyl
methyl
H
1
ethyl
1
ethyl
1
phenyl
1
benzyl
H
1
2ꢀphenylethyl
benzyl
H
1
0i
ethyl
benzyl
H
1
0j
benzyl
1
0k
0l
0m
0n
4ꢀaminobenzyl
4ꢀchlorobenzyl
4ꢀnitrobenzyl
4ꢀsulfamoylbenzyl
1
H
1
H
1
H
22
ACS Paragon Plus Environment

Page 23 of 56
Journal of Medicinal Chemistry
a 3
[
H]AMP (5 µM) was used a substrate (K value 59 µM).
m
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
Finally we replaced the amino group at C6 of the lead structure I by a chlorine atom (10s) or a keto
group leading to inosine derivative 10o. While the 6ꢀchloropurine nucleotide showed about the same
inhibitory potency as the corresponding adenine nucleotide I (compare 10s with I), the hypoxanthine
nucleotide 10o was about 14ꢀfold less potent (10o, 2830 nM). This could be due to the different, nonꢀ
aromatic tautomeric structure of 10o as compared to I and 10s.
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
Table 2. Inhibitory potency of 6ꢀClꢀ, Oꢀ and Sꢀsubstituted purineꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic
acid] derivatives at rat ectoꢀ5′ꢀnucleotidase
Rat eN
Compd.
0o
X
R
a
K ± SEM (nM)
i
1
see structure above
2830 ± 420
32.0 ± 4.1
9.20 ± 0.52
9.50 ± 1.79
157 ± 9
1
1
0p
0q
O
O
S
ethyl
benzyl
benzyl
1
0r
0s
1
see structure above
a 3
[
H]AMP (5 µM) was used a substrate (K value 59 µM).
m
6
Analogs of the very potent 6ꢀbenzylaminoꢀsubstituted derivative 10g in which the 6ꢀamino group (N H)
was replaced by an oxygen (10q) or a sulfur atom (10r) were also investigated. Both compounds were
6
equally potent as their N ꢀanalog. This offers the possibility to move away from adenine nucleotide
6
derivatives, which might be metabolized to (N ꢀsubstituted) adenosine derivatives after hydrolysis of the
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 56
6
ester bond between the ribose and the diphosphonate moiety. Certain N ꢀsubstituted adenosine
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
derivatives are known to potently activate adenosine A receptors and may thereby induce negative
1
5
2, 53
inotropic and chronotropic effects, which might even lead to cardiac arrest.
Table 3. Potency of 8ꢀsubstituted adenosineꢀ5′ꢀOꢀ[(phosphonomethyl)phosphonic acid] derivatives at
rat ectoꢀ5′ꢀnucleotidase
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
Rat eN
Compd.
R
a
K ± SEM (nM)
i
1
0t
bromo
chloro
491 ± 50
93.5 ± 4.5
1720 ± 380
3610 ± 580
20.4 ± 0.7
1
0u
10v
10w
10x
aminomethyl
thioethyl
see structure above
a 3
[ H]AMP (5 µM) was used a substrate (K value 59 µM).
m
Next, we investigated derivatives of lead structure I bearing substituents in the 8ꢀposition. The
following rank order of potency was observed: Cl (10u, 93.5 nM) > H (I, 197 nM) > Br (10t, 491 nM) >
NHMe (10v, 1720 nM) > SEt (10w, 3610 nM). These results showed that only small substituents were
tolerated at the purine 8ꢀposition. The smaller chlorine atom increased potency of I by 2ꢀfold, but the
larger bromine atom already led to a significant reduction.
To further investigate and explain the influence of substituents in the 8ꢀposition, a model for the binding
mode of 10w was generated (Figure 4). Substituents in the 8ꢀposition result in close contacts of the
substituent with the terminal phosphate group in the eNꢀAOPCP complex structure. The coordination of
the phosphate group by the zinc ions plays a decisive role for the inhibitory binding mode of AOPCP to
24
ACS Paragon Plus Environment

Page 25 of 56
Journal of Medicinal Chemistry
9
eN. Thus, small shifts in the position of the terminal phosphate probably result in a loss of inhibitory
potency for larger substituents such as thioethyl (10w). Only for the smaller Clꢀsubstituent, minor shifts
of the nucleobase might be sufficient to compensate for the close contact (10u). Possibly, halogen
bonding interactions of the Cl atom with the phosphate oxygens contribute to the improved affinity of
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
0u. Due to their larger size, the bromo (10t) and thioethyl substituents (10w) likely result in steric
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
clashes with the phosphate oxygens. Another possible explanation for our findings may be that larger 8ꢀ
substituents can induce a conformational change from the antiꢀ to the synꢀconformation around the
5
4, 5
nucleosidic bond, which is unfavourable for binding to the enzyme.
Figure 4. Modeled binding mode of 10w in the substrate binding site of human eN (pdb: 4H2I was used
for modelling).
In order to further study the effect of 8ꢀsubstitution we synthesized hybrid molecule 10x which
6
combines an N ꢀbenzyl substitution (one of the best substituents at the 6ꢀposition) with a bromoꢀ
substitution at the 8ꢀposition (a substituent that had decreased potency in AOPCP). For 10x an K value
i
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 56
6
of 20 nM was observed as compared to the corresponding 8ꢀunsubstituted N ꢀbenzyl derivative (10g,
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
9
.03 nM). A 2ꢀfold decrease in potency was observed for 10x as compared to 10g, which was similar to
the 2ꢀfold decrease for the 8ꢀbromo derivative 10t as compared to AOPCP (I).
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
Table 4. Potency of α,βꢀsubstitutedꢀmethyleneꢀADP derivatives
rat eN
5
6
Compd.
R
R
a
K ± SEM (nM)
i
1
2a
2b
2c
bromo
chloro
bromo
chloro
methyl
3930 ± 530
292 ± 23
1
1
hydroxyl
4050 ± 390
a 3
[
H]AMP (5 µM) was used a substrate (K value 59 µM).
m
Moreover, some derivatives with substitution in the methylenediphosphonate sideꢀchain of AOPCP
were synthesized and investigated. Adenosineꢀ5′ꢀdibromomethylenediphosphonic acid (12a),
adenosineꢀ5′ꢀdichloromethylenediphosphonic acid (12b), and adenosineꢀ5′ꢀ(1ꢀhydroxy)ethaneꢀ1,1ꢀ
diphosphonic acid (12c) were tested under same assay condition. All the substitutions resulted in a
reduction of potency as compared to AOPCP, however 12b was more potent than 12a and 12c. An
analysis of the environment of substituents at the methylene bridge of AOPCP shows that both positions
have a very polar environment (Figure 5). One position is in hydrogen bonding distance to a zincꢀ
coordinated water molecule, but at quite short distance to the carbonyl group of H243 (1.71 Å distance
when the substituent is OH, before energy minimization). This would require some rearrangements in
the positions of the phosphate groups. Substituents larger than OH cannot be accepted at this position.
The second substituent atom is within hydrogen bonding distance to water molecules. Thus, a methyl
26
ACS Paragon Plus Environment

Page 27 of 56
Journal of Medicinal Chemistry
group is not a suitable substituent at the methylene bridge of AOPCP. A single hydroxyl group or a
hydroxyl group and a second polar substituent may be more favorable.
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. Modeled binding mode of one enantiomer of 12c in the substrate binding site of human eN
(pdb: 4H2I was used for modelling).
Inhibition of human ectoꢀ5′ꢀnucleotidase
Two of the most potent inhibitors, 10g and 10r, and lead structure I were selected for further
investigation on human eN. The compounds were tested for inhibition of recombinantly produced
human enzyme, and also of soluble eN present in human blood serum. Results are presented in Table 5
and concentrationꢀinhibition curves are depicted in Figure 6.
Table 5. K values of selected inhibitors at recombinant human eN and at soluble eN in human serum
i
a
determined by radiometric TLC assay
Compd.
Recombinant rat eN
K ± SEM (nM)
Recombinant human eN
K ± SEM (nM)
Human serum eN
K ± SEM (nM)
i
i
i
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 56
I (AOPCP)
197 ± 5
88.4 ± 4.0
487 ± 186
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
10g
9.03 ± 1.24
9.50 ± 1.79
2.21 ± 0.40
4.58 ± 0.23
18.8 ± 12.5
1
0r
68.8 ± 23.2
a 3
[ H]AMP (40 µM) was used a substrate (K value 40 µM and 65 µM, respectively for recombinant
human eN and eN in human serum)
m
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
Determined inhibitory potencies at human recombinant eN were in the same range as those observed for
the rat enzyme, which is in agreement with the high similarity of both enzymes, especially in the active
56
site. The compounds appeared to be even somewhat more potent (2ꢀ to 4ꢀfold) at human as compared
to rat eN. However, in human blood serum the potency was reduced (5ꢀ15ꢀfold). This may be explained
by plasma protein binding of the compounds which reduces the free fraction of the drugs. Nevertheless,
the compounds, in particular 10g, showed very high inhibitory potency, and the rank order of potency of
the three compounds was identical for both species and in the different test systems (Figure 6).
A
B
Figure 6. Concentrationꢀinhibition curve of selected inhibitors (A) at recombinant human eN; (B) at eN
in human serum. AMP concentration, 40 µM. Results are plotted as the percentage of maximal activity
measured in the absence of inhibitors. K values are collected in Table 5.
i
28
ACS Paragon Plus Environment

Page 29 of 56
Journal of Medicinal Chemistry
Selectivity
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
Lead structure I and selected derivatives were investigated for inhibition of further ectoꢀnucleotidases,
including human NTPDase 1ꢀ3 and human NPP 1ꢀ3. Compound I showed a moderate inhibition of NPP
57
1
(K 16500 nM), but otherwise we observed no inhibition by I or any of the selected derivatives 10e,
i
1
0g, 10h, 10l, and 10r at a high test concentration of 10 µM at any of the ectoꢀnucleotidases (see
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
Supporting Information, page S20). Furthermore the ADP analogs were investigated on ADPꢀactivated
G proteinꢀcoupled receptors, namely P2Y and P2Y receptors. None of the selected compounds (10b,
1
12
10e, 10g, 10l) activated or inhibited P2Y or P2Y receptors at the high concentration of 10 µM (see
1
12
Supporting Information, page S22).
Metabolic stability
The most potent and selective eN inhibitors 10g and 10l as well as the lead compound AOPCP were
further investigated for their stability in rat liver microsomes in order to investigate potential metabolic
degradation by liver enzymes. Compound 10g and AOPCP (but not compound 10l owing to its lower
stability as compared to 10g) were further investigated for their stability in human blood plasma. The
experiments were performed as described in the Experimental Section. Samples were incubated at 37 °C
and subsequently analyzed by HPLC coupled to electrospray ionization mass spectroscopy (HPLCꢀESIꢀ
MS).
Incubation of the compounds with rat liver microsomes showed that 10g and 10l were quite stable. Only
a small percentage (<23%) of 10g was metabolized under the applied conditions (see Experimental
Section), while >77% of the compounds were recovered unchanged after a long incubation for 8 h
(Figure 7). Compound 10g was somewhat less stable, and 56% were metabolized within 8 h. Hydrolytic
cleavage of the glycosidic bond, which is a typical phase I reaction of nucleosides and nucleotides, was
observed as the main reaction. Under same conditions AOPCP was cleaved much faster, and hydrolysis
was complete within 15 min.
29
ACS Paragon Plus Environment