Thienopyrimidine bisphosphonate (ThPBP) inhibitors of the human farnesyl pyrophosphate synthase: Optimization and characterization of the mode of inhibition

Article abstract of DOI:10.1021/jm400946f

Human farnesyl pyrophosphate synthase (hFPPS) controls the post-translational prenylation of small GTPase proteins that are essential for cell signaling, cell proliferation, and osteoclast-mediated bone resorption. Inhibition of hFPPS is a clinically validated mechanism for the treatment of lytic bone diseases, including osteoporosis and cancer related bone metastases. A new series of thienopyrimidine-based bisphosphonates (ThP-BPs) were identified that inhibit hFPPS with low nanomolar potency. Crystallographic evidence revealed binding of ThP-BP inhibitors in the allylic subpocket of hFPPS. Simultaneous binding of inorganic pyrophosphate in the IPP subpocket leads to conformational closing of the active site cavity. The ThP-BP analogues are significantly less hydrophilic yet exhibit higher affinity for the bone mineral hydroxyapatite than the current N-BP drug risedronic acid. The antiproliferation properties of a potent ThB-BP analogue was assessed in a multiple myeloma cell line and found to be equipotent to the best current N-BP drugs. Consequently, these compounds represent a new structural class of hFPPS inhibitors and a novel scaffold for the development of human therapeutics.

Full text of DOI:10.1021/jm400946f

Subscriber access provided by UQ Library
Article
Thienopyrimidine Bisphosphonate (ThPBP) Inhibitors
of the Human Farnesyl Pyrophosphate Synthase –
Optimization and Characterization of the Mode of Inhibition
Chun-Yuen Leung, Jaeok Park, Joris De Schutter, Michael
Sebag, Albert Marinus Berghuis, and Youla S. Tsantrizos
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400946f • Publication Date (Web): 02 Sep 2013
Downloaded from http://pubs.acs.org on September 10, 2013
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 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Thienopyrimidine Bisphosphonate (ThPBP) Inhibitors of the Human Farnesyl
Pyrophosphate Synthase – Optimization and Characterization of the Mode of Inhibition
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
‡
†
‡
§
Chun Yuen Leung, Jaeok Park, Joris W. De Schutter, Michael Sebag, Albert M.
Berghuis,
†
¶ ‡
‡†‡
Youla S. Tsantrizos*
‡
Department of Chemistry, McGill University, 801 Sherbrooke Street West, Montreal, QC,
Canada H3A 0B8
†
Department of Biochemistry, McGill University, 3649 Promenade Sir William Osler, Montreal,
QC, Canada H3G 0B1
¶
Department of Microbiology and Immunology, McGill University, 801 Sherbrooke Street West,
Montreal, QC, Canada
‡
Groupe de Recherche Axé sur la Structure des Protéines, McGill University, 3649 Promenade
Sir William Osler, Montreal, QC, Canada
§
Division of Haematology, McGill University Health Center, Royal Victoria Hospital, C6.80,
6
87 Pine Av W, Montreal, QC, Canada H3A 1A1
ABBREVIATIONS USED
hFPPS, human farnesyl pyrophosphate synthase; hGGPPS, human geranylgeranyl pyrophosphate
synthase; DMAPP, dimethylallyl pyrophosphate; IPP, isopentenyl pyrophosphate; GPP, geranyl
pyrophosphate; FPP farnesyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate; N-BPs,
nitrogen-containing bisphosphonates; MM, multiple myeloma; GTPases, small guanine
triphosphate binding proteins; ANT, mitochondrial adenine nucleotide translocase; IFN,
interferon-gamma; TNF, tumour necrosis factor-alpha; hSQS, human squalene synthase;
ApppI, derivative 1-adenosin-5’-yl ester 3-(3-methylbut-3-enyl) triphosphoric acid; HAP,
hydroxyapatite; structure-activity studies, SAR, structure-activity relationship; DLS, dynamic
light scattering.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 51
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
Abstract
Human farnesyl pyrophosphate synthase (hFPPS) controls the post-translational
prenylation of small GTPase proteins that are essential for cell signalling, cell proliferation and
osteoclast-mediated bone resorption. Inhibition of hFPPS is a clinically validated mechanism for
the treatment of lytic bone diseases, including osteoporosis and cancer related bone metastases.
A new series of thienopyrimidine-based bisphosphonates (ThP-BPs) were identified that inhibit
hFPPS with low nanomolar potency. Crystallographic evidence revealed binding of ThP-BP
inhibitors in the allylic sub-pocket of hFPPS. Simultaneous binding of inorganic pyrophosphate
in the IPP sub-pocket leads to conformational closing of the active site cavity. The ThP-BP
analogs are significantly less hydrophilic, yet exhibit higher affinity for the bone mineral
hydroxyapatite than the current N-BP drug risedronic acid. The anti-proliferation properties of a
potent ThB-BP analog was assessed in a multiple myeloma cell line and found to be equipotent
to the best current N-BP drugs. Consequently, these compounds represent a new structural class
of hFPPS inhibitors and a novel scaffold for the development of human therapeutics.
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
ACS Paragon Plus Environment

Page 3 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Human farnesyl pyrophosphate synthase (hFPPS) controls the first branching point of
the mevalonate pathway and catalyzes the biosynthesis of farnesyl pyrophosphate (FPP). FPP
is the key precursor for the biosynthesis of many metabolites, including geranylgeranyl
pyrophosphate (GGPP), squalene and cholesterol. FPP and GGPP are essential for the post-
translational prenylation of all small GTPase proteins that play a crucial role in cell signalling,
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
1
,2
cell proliferation and osteoclast-mediated bone resorption.
Inhibition of hFPPS can
downregulate the activity of mutated H-Ras, K-Ras and N-Ras proteins that function as major
drivers of tumor growth in many cancers. For example, whole genome sequencing of human
multiple myeloma (MM) tumors has revealed that 50% of MM patients harbored either K-Ras
3
or N-Ras activating mutations, underscoring the importance of hFPPS in this disease.
Metastatic bone disease is highly prevalent in patients with MM, breast and prostate cancers.
It is estimated that approximately 70% of all patients with advanced forms of these three types
4
of cancers will at some point develop bone metastases. Thus, the clinical benefits of hFPPS
inhibition include both decrease of prenylation of mutated Ras proteins, leading to a decrease
in cellular growth and/or survival, as well as alleviation of tumour-associated bone destruction
5
via inhibition of osteoclast activity. It is noteworthy that inhibitors of the farnesyltransferase
7
0
enzyme (FTase), which catalyze the prenylation of Ras (e.g. tipifarnib ), have also been
extensively investigated as an alternative mechanism for downregulating oncogenesis, but
failed to demonstrate significant clinical efficacy. It was subsequently realized that the
substrate specificity of the transferase enzymes (FTase and GGTase I and II) is not absolutely
stringent and cross-prenylation can occur, restoring the biological function of the mutated Ras
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 51
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
71
proteins. However, this redundancy mechanism cannot compromise the clinical effects of
drugs that directly down-regulates the biosynthesis of FPP by inhibiting hFPPS.
Additionally, blocking the catalytic activity of hFPPS impacts both the down-stream
and up-stream levels of isoprenoids in the mevalonate pathway, leading to numerous cellular
changes. Intracellular accumulation of the substrate isopentenyl pyrophosphate (IPP), as a
consequence of hFPPS inhibition, leads to accumulation of an ATP derivative (ApppI), which
induces cell apoptosis by inhibiting the mitochondrial adenine nucleotide translocase
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
,7
(ANT).
IPP is also a natural antigen that directly stimulates T cells that carry the
V2V2 T cell receptors, and is strongly implicated in the human innate immune response
8
against tumours. Interestingly, treatment of different types of cancer cells with the same
hFPPS inhibitor may result in dramatically different intracellular levels of IPP (from ~10-fold
9
to nearly 1,000-fold increase in intracellular concentrations have been reported),
highlighting the variability in regulating/disregulating metabolic pathways in different cells.
These observations suggest that selectivity in biochemical interventions may be possible, even
when targeting a critical step in an important metabolic pathway.
Currently, nitrogen-containing bisphosphonate drugs (N-BPs) are the only clinically
10,11
relevant compounds that inhibit hFPPS.
Bisphosphonates are chemically stable
bioisosteres of pyrophosphates and structurally characterized by two phosphonate groups
attached to a central carbon (C) instead of an oxygen atom. Additionally, one of the
Csubstituents (i.e. the N-BP side chain) is usually characterized by a basic nitrogen atom,
presumed to be protonated under physiological conditions and mimicking the interactions of
1
2
the putative allylic carbocation transition-state that forms during the hFPPS catalytic cycle.
The structure of the most potent N-BP drugs, such as zoledronic acid (1) and risedronic acid
4
ACS Paragon Plus Environment

Page 5 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
2a), is also characterized by a C-hydroxyl moiety, which maximizes their affinity for the
1
3,62
bone mineral hydroxyapatite (HAP).
The C-hydroxyl moiety also participates in
additional interaction with the active site of the enzyme, thus contributing to the potency 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
1
0,13
these compounds.
N-BP drugs are routinely used to treat osteoporosis and other lytic bone diseases,
14,15,16
including bone cancer metastasis and multiple myeloma (MM).
N-BPs bind so strongly
to bone that their half-life (in bone) can be months to several years, depending on the type of
1
7,18
drug and the degree of bone turnover.
In chronic diseases (e.g. osteoporosis), concerns
that prolonged use of N-BPs can lead to side effects, such as osteonecrosis of the jaw and
atypical femoral fractures, have led to the recommendation by physicians of a “drug holiday”.
However, this treatment can lead to uncertainty with respect to the type of drug that should be
1
9
used and the duration of treatment for different patients. Furthermore, the systemic half-life
of current N-BPs is extremely low; for example, after i.v. administration of 1, 50% of the dose
gets trapped in the bone mineral and the rest is rapidly cleared by the kidneys (the dose-
2
0, 21
limiting toxicity of 1 is based on nephrotoxicity).
In spite of their poor drug-like properties (including extremely low cell membrane
permeability and oral bioavailability), recent clinical investigations provide evidence that
some N-BP drugs (e.g. 1) are disease modifying agents that improve the survival of patients
with multiple myeloma (MM) via mechanisms that are both related, as well as unrelated to the
2
2, 23
skeletal benefits.
Similar results were reported for patients with premenopausal breast
2
4
25
cancer, although these findings seem to be more controversial. Nonetheless, the
identification of hFPPS inhibitors with superior oral bioavailability, half-life in plasma (i.e.
slower rates of elimination from the blood circulation) and higher non-skeletal tissue
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 51
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
distribution, may provide effective antiresorptive agents that are also more effective in cancer
chemotherapy than the current N-BP drugs. In this report, we describe our structure-activity
relationship (SAR) studies on thienopyrimidine-based bisphosphonate (ThP-BP) inhibitors of
hFPPS and the identification of analogs with low nanomolar potency. Co-crystal structures of
two ThP-BP analogs bound to the allylic sub-pocket of hFPPS (i.e. the DMAPP/GPP binding
sub-pocket of the active site) revealed details of the protein-inhibitor interactions and
conformational changes to the C-terminal region that lead to closing of the active site cavity.
These ThP-BP molecules are significantly less hydrophilic than the current N-BP drugs, but
exhibit high affinity for the bone mineral hydroxyapatite. The in vitro affinity of
bisphosphonates for hydroxyapatite is known to correlate well with the in vivo affinity of the
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
3,62
N-BP drugs for bone.
Thus, the ThP-BP compounds are promising new leads for
medicinal chemistry studies that aim to identify new inhibitors of hFPPS with better
biopharmaceutical properties than those of the current N-BP drugs.
6
ACS Paragon Plus Environment

Page 7 of 51
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
CHEMISTRY
In our search of new structural classes of hFPPS inhibitors, we recently identified
thienopyrimidine-based bisphosphonates (ThP-BPs; e.g. 6d) with modest in vitro potency in
26
inhibiting hFPPS.
The thienopyrimidine scaffold is considered to be privileged in drug
discovery, due to its inherently favourable biopharmaceutical profile. Thienopyrimidine-based
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 51
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
compounds are currently under investigations as therapeutics for the treatment of many
27
28
28,29,30,31,32,33,34,35,36
diseases, including fungal and viral infections, and cancer.
2
6
At the time of our initial report on ThP-BPs, the mechanism by which these compounds
inhibit hFPPS was unclear. Given the significantly larger molecular size of ThP-BPs (e.g. 6d) as
compared to the current N-BP drugs (e.g. 1 and 2a), we presumed that their binding interactions
with hFPPS may also be different. Mindful of the conformational plasticity of this enzyme that
permit the binding of fairly large N-BP molecules (e.g. 4) in the allylic sub-pocket, we prepared
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 focused library of ThP-BP derivatives guided by our previous SAR studies (e.g. SAR derived
37,38
from pyridine analogs, such as 4 and 5);
some representative examples are shown in Figure
1
.
The synthesis of the ThP-BP analogs was initiated from the key fragment 6-
bromothieno[2,3-d]pyrimidin-4-amine (10), which was prepared via the trimethylsilyl ylidine
2
6
9
as previously described (Scheme 1). Intermediate 10 was also prepared starting from 2,5-
dihydroxy-1,4-dithiane (12), via the unsubstituted 2-amino-thiophene-3-carbonitrile core,
39,40
following earlier literature procedures.
Although the two synthetic protocols (Path A and
B) are equivalent in the overall number of steps leading to 10, the average isolated yield via
Pathway A was significantly higher (Scheme 1; the average overall yield of 10 was 18% vs
8
% obtained via Pathway A and B, respectively).
Conversion of 10 to the bisphosphonate tetraester 11 was easily achieved upon
treatment with triethoxymethane and diethylphosphite. Cross coupling of 11 with a variety of
boronate esters, under typical Suzuki conditions, followed by hydrolysis of the tetraethyl
bisphosphonate esters 13 with TMSBr/MeOH resulted in the formation of the final
thienopyrimidine bisphosphonic acids of general structure 6 (Scheme 1).
8
ACS Paragon Plus Environment

Page 9 of 51
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
Scheme 1. Synthesis of C-6 substitted thieno[2,3-d]pyrimidin-4-amine-based
o
bisphosphonate inhibitors of hFPPS. Conditions: (a) CH(OEt) , diethylphosphite, 130 C,
3
o
2
4 h (75%); (b) 10 mol. % Pd(PPh ) , KF, MeOH, 120 C, W, 20 min; (c) TMSBr,
3
4
MeOH, RT, 72h [overall isolated yield for steps (b) and (c) ranged from ~20-80%].
RESULTS AND DISCUSSION
Metabolic disregulation of hFPPS has been implicated in many human diseases,
including various cancers (e.g. breast, prostate and MM) and neurodegenerative diseases, such as
4
1,42
the Alzheimer’s disease.
However, in vivo investigation of hFPPS as a therapeutic target is
hampered by the lack of molecular tools that can selectively inhibit this enzyme and can exhibit
significant in vivo distribution into soft tissues. Past efforts towards improving the clinical use of
4
3,44
45
N-BPs have included investigations of pro-drugs,
improved drug formulations, and
structural modifications that may increase oral bioavailability and decrease the rate of compound
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 51
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
46,47
clearance from circulation. N-BPs with more lipophilic side chains (e.g. 3,
4, 5) than the
clinical drugs 1 and 2a, as well as bicyclic heterocyclic side chains have been explored,
48
49,50
including imidazopyridines, benzimidazoles and azaindoles.
To the best of our knowledge
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
none of these efforts have produced an hFPPS inhibitor with superior clinical profile than that of
analogs 1 or 2a. We recently identified thienopyrimidine-based bisphosphonates (ThP-BPs) with
IC potency in inhibiting hFPPS in the 0.5-1 M range. These hits were used for further SAR
5
0
studies and optimized into low nanomolar inhibitors of hFPPS.
It is noteworthy that several different assays have been reported in the literature for
evaluating in vitro hFPPS inhibition. The most commonly used method was originally developed
5
1
by Reed and Rilling and has been adopted by many researchers with only minor
13,37,38,53
modifications;
we refer to these conditions as method 1 (M1). The IC values of hFPPS
50
inhibitors can vary considerably, depending on the assay method; for example, the reported IC50
1
1,52,53
54
values of 1 range from approximately 4 nM
to 200 nM, and can be as high as 475 nM if
53
the compound is tested without pre-incubation with the enzyme. Experimental artifacts in a
functional assay can further contribute to the observed potency of a compound and mislead SAR
studies. Lipophilicity-dependent aggregation of moderately (or poorly) hydrophilic compounds is
the most common factor responsible for false experimental data and can affect both in vitro and
5
5
cell-based assays. The ability of bisphosphonates to form stable, high-order complexes in
solution, particularly in the presence of metal ions, is a well-documented phenomenon, exploited
56,57,58
in a number of biomedical applications.
The high charge density of the magnesium cation
may also play a role in augmenting the solvent-induced interaction between the hydrophobic side
chains of the ThP-BP inhibitors (as compared to smaller and more polar N-BPs; CLogP values
7
2
are shown in Table 1). A combination of these effects appear to induce significant aggregation
1
0
ACS Paragon Plus Environment

Page 11 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
under the M1 assay conditions and lead to artifacts in characterizing hFPPS inhibition. A number
2
+
of assay modifications were investigated, including lowering the concentration of Mg ions and
adding Triton X-100 (0.01%) in the assay buffer. The addition of Triton X-100 in the buffer was
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
59,60
previously reported for testing non-bisphosphonate inhibitors of hFPPS.
The ratio of hFPPS
2
+
protein, IPP/GPP substrates and Mg ion concentrations were also optimized in order to achieve
maximum catalytic turnover and highest sensitivity in detecting weakly active compounds; we
refer to these conditions as assay method 2 (M2).
The IC values of several known hFPPS inhibitors were determined using both the M1
50
and M2 assay conditions. Little variability (≤2-fold) was observed in the observed potency of
pyridine-based N-BP molecules (Table 1; analog 2a and 5). In contrast, the IC of the hFPPS
5
0
allosteric inhibitor 7 was approximately 22-fold lower in the M2 assay than the M1 assay and
6
1
~
100-fold lower than the value reported using yet another assay protocol; the IC values of 7
50
were determined to be ~20 M and 0.92 M in the M1 and M2 assay, respectively (Table 1).
The IC values for our initial hits, 6e and 6f, were also lower in the M2 assay as compared to the
5
0
M1 assay (Fig. 1a, Table 1). A defining characteristic of molecular aggregation is the formation
of higher-order soluble particles on the submicrometer scale that are disrupted (to some degree)
5
5
by the presence of a detergent, such as Triton X-100, Tween-20 or Tween-80. The higher
propensity of ThP-BPs to aggregate in the assay buffer, as compared to the N-BP analog 2a was
confirmed by dynamic light scattering (DLS); aggregation was particularly pronounced in the
2+
presence of high Mg concentrations (an example is shown in SI Figure 1).
The optimized M2 assay was subsequently adopted for our routine biological screening
of all ThP-BP inhibitors. All analogs were initially tested at a concentration of 1.0 M and 0.1

M; representative examples are shown in Figure 1. The general potency profile for these
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 51
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
compounds was similar to the SAR previously observed with 2-amino- and 3-aminopyridine
3
7,38
libraties.
For example, analogs with polar heterocyclic side chains, such as 6a, 6b and 6c,
exhibited weak activity in inhibiting hFPPS (<20% inhibition was observed at 100 nM), and
meta-substituted phenyls were less potent than their corresponding para-analogs (e.g. analogs 3h
and 3i exhibited 73% vs 30% inhibition at 100 nM, respectively). Full dose-response inhibition
curves were obtained only for analogs deemed critical to our SAR studies and analogs exhibiting
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
>50% inhibition at 100 nM. A number of novel thienopyrimidine-based inhibitors of hFPPS
were identified with IC50 values in the 10-50 nanomolar range; representative examples are
shown in Table 1. In addition, all analogs with an IC <100 nM were tested in our in vitro
5
0
3
7
hGGPPS inhibition assay, using compound 8 as the positive control, as previously reported;
none of these compounds were active at concentration up to 10 M; the buffer of our hGGPPS
inhibition assay contained 0.2% Tween-20 (i.e. 20-fold higher concentration of detergent than
our M2 hFPPS inhibition assay), which should minimize or eliminate any effects due to
aggregation.
1
2
ACS Paragon Plus Environment

Page 13 of 51
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: Thienopyrimidine Inhibitors of the human FPPS. (a) Representative examples of
analogs. (b) Inhibition of hFPPS at 100 nM concentration of each compound using the M2 assay
(average of three determinations; standard deviation ≤10%).
Table 1: Inhibition data for key compounds

Compound
hFPPS IC (nM)
CLogP of
side chain
moiety
5
0
M1 assay
M2 assay

1
4.1
11
16
nd
5.2
18
-0.84
-0.16
3.6
nd
5.0
4.3
3.7
4.7
4.6
4.3
4.8
4.1
2a
5
7
~20,000
250
390
440
nd
920
63
22
15
21
200
14
39
6
e
6
6
f
j
6h
6
k
nd
nd
115
140
6
l
6
m
6n
11

Average IC50 values of a minimum of three determinations (standard deviation ≤10). All
compounds were also tested in our hGGPPS inhibition assay and none were active at

53
concentration up to 10 M. IC value reported by Kavanagh et al. CLogP values were
5
0
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 51
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
estimated without the two phosphonate groups in order to simplify the calculations. Value was
not determined (nd)
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
It has been shown that N-BPs such as 1, that bind very strongly to bone and get rapidly
21
eliminated from plasma, achieve extremely low concentrations in most non-calcified tissues.
Consequently, their therapeutic efficacy as antitumor agents for non-bone-related cancers is
greatly compromised. The C-hydroxy bisphosphonate moiety of the current drugs (1 and 2a) is
2
+
generally believed to enhance bone affinity by allowing a tridentate interaction with Ca ions
and the bone mineral hydroxyapatite. Consistent with this assumption, replacement of the C-
hydroxy group with a proton (e.g. 2b) or a halogen (e.g. 2c) was shown to decrease both the
1
3
affinity for the bone mineral hydroxyapatite and the ability of the compounds to inhibit hFPPS.
The in vitro affinity of bisphosphonates for hydroxyapatite is known to correlate well with the in
1
3,62
vivo affinity of the N-BP drugs for bone.
We compared the relative binding affinity of ThP-
BPs 6m and N-BP 2a for hydroxyapatite (HAP) using the NMR protocol reported by Jahnke and
62
Henry. After incubation with HAP, changes in the relative intensity of the aromatic proton
signals were observed for both compounds. However, a much more significant decrease in the
intensity of the aromatic signals of analog 6m was observed as compared to 2a, consistent with a
higher portion of 6m binding to HAP and removed from the solution (Fig. 2). Similar results
were also obtained with analogs 6h and 6l, suggesting that this high affinity for hydroxyapatite is
likely structure-dependent. These results are somewhat surprising, since it is generally believed
that the C-hydroxyl substituent of N-BPs provides the optimal “bone hook” and contributes to
the anti-resorptive efficacy of the current N-BP drugs. Differences in lipophilicity and bone
1
4
ACS Paragon Plus Environment

Page 15 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
affinity can potentially affect the pharmacokinetic and pharmacodynamics properties of these
compounds, thus contributing to a unique therapeutic profile.
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: Differentiation of 2a and 6m in their competitive binding to HAP (for clarity, only
select regions of the H NMR spectra are shown). (a) H NMR spectra of 2a and 6m at
1
1
approximately 1:1 molar ratio (~50 M of each compound) in Tris buffer before any treatment.
1
(
b) H NMR spectrum of the same solution of 2a and 6m as that used to acquire the spectrum
shown in (a) after incubation with 0.8 mg of HAP at room temperature for 5 min.
Clinical evidence is rapidly accumulating which implicates inhibitors of hFPPS as anti-
2
2,23
neoplastic and specifically anti-myeloma therapeutic agents.
Anti-proliferation and cytotoxic
effects were observed with inhibitor 6m that compared favorably[;;p to the effects of 1 and 2a
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 51
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
(
Table 1). Treatment of human MM cell line RPMI-8226 with each of the three compounds
resulted in reductions of cell proliferation with median effective concentrations for 50% growth
inhibition (EC ) of 11 M, 13 M and 8.5 M, for 1, 2a and 6m, respectively (Table 2); values
5
0
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
represent the average of n≥8 determinations with R in the range of 0.90-0.98.
Table 2: Anti-proliferation effects in the multiple myeloma cells
Compound hFPPS IC50
nM)
EC (M)
50
(
M2 assay
Cell line
RPMI-8226
1
nd
5.2
39
11
13
8.5
2a
6
m
2
Average EC values of n ≥8 determinations; R values in the range of 0.90-0.98.
Value was not determined (nd)
5
0
Finally, two representative ThP-BP inhibitors, 6f and 6m, were also co-crystallized with
hFPPS (Table 3). Both compounds were found to bind in the allylic sub-pocket of the active site
in the same manner, without causing major conformational strain to the protein, despite their
bulky and rigid side chains. However, unlike 1 and 2a (which bind to only a small portion of the
allylic sub-pocket), the interactions of 6f and 6m with the active site cavity extend through the
capping phenyls (Phe 98/99) and towards the hFPPS dimer interface. Some of the key features of
these binding interactions are illustrated with the hFPPS-6m-PPi ternary complex (PDB ID:
4L2X) in Figure 3. The simulated annealing omit map clearly demonstrates the presence of
bound 6m, as well as three magnesium ions and water molecules, which mediate the interaction
between the bisphosphonate moiety and the two DDXXD motifs of the protein (Fig. 3a). Some
1
6
ACS Paragon Plus Environment

Page 17 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of the metal-mediated interactions between the bisphosphonate and aspartic acid residues of the
DDXXD motifs are direct, whereas others are water mediated. There are three additional
interactions between the positively charged residues Arg 112, Lys 200, and Lys 257, which make
direct contacts with the bisphosphonate moiety of the inhibitor, which are not shown in Figure 3a
for clarity. The side chain of 6m fills the lipophilic cavity of the allylic sub-pocket completely,
with its cyclopropyl tail extending to the end of the cavity at the dimerization interface (Fig. 3b).
The phenyl ring of the 6m side chain is engaged in stacking interactions with the side chains 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
3
7,63
Phe 99 and Gln 171, similarly to what was previously observed with inhibitor 4 (Fig. 3c).
The thienopyrimidine scaffold also displaces the side chain hydroxyl of Thr 201 and the
main chain carbonyl of Lys 200 by ~0.65Å; these residues are presumed to participate in a
bifurcated H-bond interaction with the protonated side chain of 1 (Fig. 3c). The hFPPS-6m-PPi
complex adopts a fully closed conformation, similar to that observed in the hFPPS-1-IPP ternary
1
0
complex (PDB ID: 1ZW5). Superposition of our hFPPS-6m-PPi structure with the structure of
350
353
the hFPPS-1-IPP ternary complex (Fig. 4d) revealed the same folding of the KRRK C-
terminal tail over the IPP sub-pocket. These data are consistent with our earlier observations
indicating that following occupancy of the allylic sub-pocket, inorganic pyrophosphate (PPi) can
6
3
play the same role as IPP in inducing the closing of the hFPPS active site cavity. Previous
investigations suggested that this secondary ligand-induced conformational change of the C-
3
50
353
terminal basic residues ( KRRK ) leads to a nearly irreversible inhibition of the enzyme and
64
is partly responsible for the excellent in vivo efficacy of the bisphosphonate compounds.
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 51
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. Crystal structure of hFPPS in complex with analog 6m (PDB: 4L2X); key interactions
are indicated with dashed yellow lines. (a) The blue mesh represents a simulated annealing omit
map (F -F , contoured at 3 sigma) for the bisphosphonate ligand, magnesium ions (yellow
o
c
2
+
spheres), and coordinated water molecules (red spheres). In addition to the Mg -mediated
bisphosphonate interactions with the aspartic acid residues (some direct and others via water
molecules), there are three direct interactions between the bisphosphonate and Arg 112, Lys 200,
and Lys 257, which are not shown for clarity. (b) A cross section view of the allylic (GPP) sub-
pocket rendered in a surface representation. The allylic pocket is blocked at one end by the
adjacent monomer (hFPPS exists as a homodimer with an active site in each monomer), which is
shown in grey. (c) Superimposition of the N-BP inhibitors 1, 4, and the ThPBP analog 6m in
three co-crystal structures. Carbon atoms are represented in green, cyan, and magenta, for the
hFPPS-1-IPP (PDB: 1ZW5), hFPPS-4-PPi (PDB: 4H5D), and hFPPS-6m-PPi (PDB: 4L2X)
ternary complexes, respectively. The color representation for heteroatoms is: blue for nitrogen;
red for oxygen; orange for phosphorous; and yellow for sulfur. The secondary ligands (i.e. PPi or
IPP) are omitted for clarity. (d) Closing of the active site in the presence of bound IPP or PPi.
3
50
353
The KRRK tail and Lys 57 are rigidified in both the hFPPS-1-IPP and hFPPS-6m-PPi
complexes. The side chains of Arg 352 and Lys 353 are not shown for clarity. The potential
steric clash between the thienopyrimidine core of 6m and the prenyl side chain of IPP is evident
(closest carbon-to-carbon contact is less than 2.7Å).
1
8
ACS Paragon Plus Environment

Page 19 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In addition, our crystal structure suggested that co-binding of IPP and inhibitor 6m was
unlikely due to a potential steric clash between the thienopyrimidine core of 6m and the prenyl
side chain of IPP (Fig. 4b). This prediction was confirmed by determining the thermal stability 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
65
the hFPPS complexes using differential scanning fluorimetry (DSF). A substantial difference
in thermal stability was observed for the hFPPS-2a and hFPPS-4 complexes in the presence or
o
o
absence of IPP (T ) of approximately +8 C and +14 C, respectively (Fig. 5a,b). The thermal
m
stability difference observed between the hFPPS-2a and hFPPS-2a-IPP complexes is consistent
o
with a T of +8.3 C measured using differential scanning calorimetry (DSC) under the same
m
1
0
ratio of ligands (i.e. 2a:IPP in 1:1 ratio). In contrast, parallel titration of hFPPS with 6m and
IPP, showed no significant difference in the thermal stability of these complexes (Fig. 5c; Tm
o
≤
1 C), suggesting that co-binding of 6m and IPP was prohibited.
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 51
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: Differential Scanning Fluorimetry (DSF) surface plots showing concentration-
dependent increase in thermal stability of hFPPS-inhibitor-IPP complexes; hFPPS concentration
fixed at 4 M; the x/y axis indicate IPP:inhibitor molar ratios relative to the concentration of
hFPPS. (a) Co-binding of inhibitor 2a and IPP to hFPPS. (b) Co-binding of inhibitor 4 and IPP
to hFPPS. (c) Co-titration of inhibitor 6m and IPP to a solution of hFPPS; plot suggests the
simultaneous binding of 6m and IPP in the hFPPS active site does not occur.
CONCLUSIONS:
Herein we disclose the hit-to-lead optimization of a new series of thienopyrimidine-based
bisphosphonate (ThP-BP) inhibitors of hFPPS, which led to the identification of compounds with
low nanomolar potency. The crystal structures of the hFPPS-6f-SO and hFPPS-6m-PPi ternary
4
complexes revealed that the inhibitors bind to the allylic sub-pocket of the active site and their
side chain extends through the capping phenyls and towards the protein dimer interface. The side
chain of these inhibitors engages in stacking interactions with both Phe 99 and Gln 171, as we
2
0
ACS Paragon Plus Environment

Page 21 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
37,63
previously observed with inhibitor 4.
A number of other conformational changes within the
active site were also observed, including the closing of the C-terminal tail over the IPP sub-
pocket, which was induced by inorganic pyrophosphate (PPi). Our X-ray structures and the DSF
thermal stability data of the hFPPS-6m complex, also suggest that the ThP-BP inhibitors
compete for binding with both natural substrates of hFPPS (i.e. DMAPP/GPP and IPP) and
occupy (in part) both sub-pockets of the active site cavity. Significant physicochemical
differences between the ThP-BP compounds and the current N-BP drugs include higher
lipophilicity and higher affinity for the bone mineral hydroxyapatite. Analog 6m, which exhibits
good in vitro potency and a high lipophilicity, was also shown to inhibit proliferation of human
multiple myeloma cells (RPMI 8226) with equivalent potency to N-BPs 1 and 2a. Thus further
lead optimization of the ThP-BP analogs is warranted and currently in progress. The higher
lipophilicity of these compounds could result in lower rate of elimination from plasma, higher
volume of distribution into non-skeletal tissue and better therapeutic value for treating multiple
myeloma and other non-skeletal cancers.
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
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 51
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
Experimental
General Procedures for Characterization of 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
All analogs 13 were purified by normal phase flash column chromatography using a
CombiFlash instrument on silica gel and a solvent gradient from 5% EtOAc in hexanes to
1
00% EtOAc and then to 20% MeOH in EtOAc, unless otherwise indicated. The homogeneity
of all analogs 13 was confirmed by reverse-phase HPLC. Only bisphosphonate esters 13 with
homogeneity ≥95% were processed further though the hydrolysis step to the final
bisphosphonic acid inhibitors 6a-n. IC values were determined only for final compounds 6a-
5
0
®
n with ≥95% homogeneity. HPLC Analysis was performed using a Waters ALLIANCE
instrument (e2695 with 2489 UV detector and 3100 mass spectrometer). Each analog of 13
1
13
31
and 6 was fully characterized by H, C and P NMR, and MS. Chemical shifts () are
1
13
reported in ppm relative to the internal deuterated solvent ( H, C) or external H PO ( 0.00
3
4
3
1
P), unless indicated otherwise. The high-Resolution MS spectra of final products 6a-n were
+/-
recorded using electrospray ionization (ESI ) and Fourier transform ion cyclotron resonance
mass analyzer (FTMS).
Method (homogeneity analysis using a Waters Atlantis T3 C18 5 m column):
Solvent A: H O, 0.1% formic acid
2
Solvent B: CH CN, 0.1% formic acid
3
Mobile phase: linear gradient from 95%A and 5%B to 5%A and 95%B in 13 min, then 2 min at
1
00% B
Flow rate: 1 mL/min
2
2
ACS Paragon Plus Environment

Page 23 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis of tetraethyl (((6-bromothieno[2,3-d]pyrimidin-4-yl)amino)methylene)bis
(phosphonate) (11)
A solution of 6-bromothieno[2,3-d]pyrimidin-4-amine (10, 500 mg, 2.173 mmol, 1 eq.) in
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
anhydrous toluene (20 mL) was flushed with argon in a pressure vessel. Diethylphosphite
(1.96mL, 15.2mmol, 7 eq.) and triethyl orthoformate (0.61 mL, 3.69 mmol, 1.7 eq.) were added
to the reaction mixture via syringe and the reaction mixture was argon flushed, sealed and stirred
o
at 130 C in the dark for 48 hours. The crude mixture was concentrated to dryness under
vacuum. The crude product was purified by normal phase flash column chromatography on
silica gel using a CombiFlash instrument and a solvent gradient from 20% EtOAc/hexanes to
1
00% EtOAc and 100% EtOAc to 20% methanol/EtOAc to give intermediate 11 as a pale yellow
solid (834.2 mg, 74% isolated yield).
1
H NMR (400 MHz, CD OD): δ 8.44 (s, 1H), 7.85(s, 1H), 5.96 (t, J=23.7Hz, 1H), 4.24-4.17 (m,
3
8
H), 1.31-1.25 (m, 12H)
1
3
C NMR (75 MHz, CD OD): δ 168.7, 156.2, 154.5, 123.1, 119.1, 113.4, 65.2-65.0 (m), 45.6
3
(t, J=150.2Hz), 16.7-16.6 (m)
3
1
P NMR (81 MHz, D O): δ 17.2
2
+
MS (ESI-) m/z [M + H] : 514.1
General protocol for the Suzuki cross-coupling reactions using fragment 11:
Suzuki coupling reactions were run in parallel using aliquots of fragment 11 (~40 mg) and
various boronate esters (1.4 eq.); examples are shown in Figure 1. All reactions were carried
out in MeOH (purged with argon, both before and after the addition of reagents), in an argon
o
flushed microwave reactor vial, at 120 C for 20 min (120W), catalyzed by Pd(PPh ) (0.1 eq.)
3
4
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 51
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
and KF (2.5 eq.). The crude mixtures were filtered through celite, concentrated to dryness
under vacuum and purified using automation as described in the general procedures.
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
General protocol for the hydrolysis of the esters 13 to give the final inhibitors 6a-n:
o
A solution of the tetraethyl bisphosphonate ester (1 eq.) in CH Cl was cooled to 0 C and
2
2
trimethylsilyl bromide (15 eq.) was added via syringe. The reaction mixture was stirred at
31
room temperature for 3-5 days; completion of conversion was monitored by P NMR. The
mixture was then concentrated under vacuum, diluted with HPLC grade MeOH (~5 mL), and
the solvent was evaporated to dryness; this step was repeated four times. The organic solvents
were evaporated under vacuum, the residue was suspended in 0.5 mL MeOH, H O (~5 mL
2
Milli-Q grade) was added to induce precipitation of the thienopyrimidine bisphosphonic acid
6
. The amorphous powder was collected by filtration, washed with de-ionized water (2x),
HPLC-grade CH CN (2x), and with distilled Et O or toluene (2x). The residue was then dried
3
2
under vacuum to give the final compound 6a-n as a solid; it should be noted that although the
13
conversion of 13 to 6 was quantitative (as determined by HPLC and P NMR), the isolated
yields varied, depending on the solubility of each final products.
(((6-(1H-indazol-4-yl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid (6a)
Isolated as a yellow solid, 18.5mg (66% overall isolated yield).
1
H NMR (500 MHz, D O): δ 8.30 (overlapping singlets, 2H), 7.31-7.29 (m, 2H), 7.15-7.12 (m,
2
2
H); central methylene proton obscured by solvent signal (confirmed by HSQC).
1
3
C NMR (126 MHz, D O): δ 162.8, 156.0, 153.3, 140.6, 137.1, 132.5, 126.7, 125.7, 119.7,
2
118.7, 118.0, 115.6, 110.9, 50.4 (t, J=124.9Hz)
2
4
ACS Paragon Plus Environment

Page 25 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
13
1
13
HSQC ( H – C): H δ 4.57 correlates with C δ 50.4.
3
1
P NMR (81 MHz, D O): δ 13.8
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 439.99890, found m/z 439.99914.
1
4
12
5
2
6
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
(((6-(1H-indazol-5-yl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid (6b)
Isolated as a pale pink solid, 14.9mg (17% overall isolated yield).
1
H NMR (500 MHz, D O): δ 8.02 (s, 1H), 7.65 (d, J=8.2Hz, 1H), 7.59 (s, 1H), 7.55 (s, 1H), 7.50
2
(d, J=8.2Hz, 1H), 6.94 (s, 1H) ; central methylene proton obscured by solvent signal.
13
C NMR (126 MHz, D O): δ 161.7, 155.4, 152.8, 139.2, 139.1, 133.3, 125.8, 124.3, 122.1,
2
1
18.1, 117.9, 112.9, 111.8, C observed by HSQC.
1
13
1
13
HSQC ( H – C): H δ 4.56 correlates with C δ 50.8.
3
1
P NMR (81 MHz, D O): δ 13.9
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 439.99890, found m/z 439.99912.
1
4
12
5
2
6
(((6-(3,5-dimethylisoxazol-4-yl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid
(6c)
Isolated as a white solid, 18.2mg (39% overall isolated yield).
1
H NMR (500 MHz, D O): δ 8.29 (s, 1H), 7.56 (s, 1H), 4.63 (t, J=18.9Hz, 1H), 2.56 (s, 3H),
2
2.40 (s, 3H)
13
C NMR (126 MHz, D O): δ 167.9, 163.5, 159.8, 156.1, 153.5, 126.6, 118.5, 117.7, 110.1, 50.8
2
(t, J=125.9Hz), 11.4, 10.2
3
1
P NMR (81 MHz, D O): δ 13.6
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 418.9986, found m/z 418.9985.
1
2
13
4
2
7
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 51
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-(m-tolyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid (6g)
Isolated as a pale yellow solid, 25.5mg (55% overall isolated yield).
1
H NMR (500 MHz, D O): δ 8.28 (s, 1H), 7.87 (s, 1H), 7.65 (s, 1H), 7.60 (d, J=7.8Hz, 1H), 7.40
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
(t, J=7.8Hz, 1H), 7.27 (d, J=7.8Hz, 1H), 4.63 (t, J=18.8Hz, 1H), 2.39 (s, 3H)
1
3
C NMR (126 MHz, D O): δ 163.2, 156.2, 153.5, 139.5, 139.4, 133.1, 129.3, 129.2, 126.5,
2
1
23.0, 118.6, 114.6, 20.3, C observed by HSQC
1
13
1
13
HSQC ( H – C): H δ 4.63 correlates with C δ 51.0.
3
1
P NMR (81 MHz, D O): δ 13.7
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 414.00840, found m/z 414.00848.
1
4
14
3
2
6
(((6-(4-(trifluoromethyl)phenyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid
6h)
(
Isolated as a yellow solid, 25.4mg (52% overall isolated yield).
1
H NMR (500 MHz, D O): δ 8.29 (s, 1H), 8.00 (s, 1H), 7.93 (d, J=8.2Hz, 2H), 7.79 (d, J=8.2Hz,
2
2
H), 4.63 (t, J=18.9Hz, 1H)
1
3
C NMR (126 MHz, D O): δ 163.6, 156.3, 153.9, 137.5, 136.7, 129.2 (q, J=32.4Hz), 126.2,
2
1
26.0 (q, J=3.8Hz), 124.1 (q, J=270.1Hz), 118.5, 116.5, C observed by HSQC
1
13
1
13
HSQC ( H – C): H δ 4.63 correlates with C δ 51.1.
3
1
P NMR (81 MHz, D O): δ 13.6
2
-
HRMS (ESI-) calculated for C H F N P O S m/z [M - H] : 467.97959, found m/z 467.9783
1
4
11
3
3
2
6
2
6
ACS Paragon Plus Environment

Page 27 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
((6-(3-(trifluoromethyl)phenyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid
(6i)
Isolated as a white solid, 23.4mg (44% overall isolated yield).
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
1
H NMR (500 MHz, D O): δ 8.29 (s, 1H), 8.08 (s, 1H), 8.00 (d, J=7.8Hz, 1H), 7.95 (s, 1H), 7.71
2
(d, J=7.8Hz, 1H), 7.65 (t, J=7.8Hz, 1H), 4.64 (t, J=18.8Hz, 1H)
1
3
C NMR (126 MHz, D O): δ 163.5, 156.3, 153.8, 137.7, 134.0, 130.8 (q, J=32.9Hz), 129.7,
2
1
29.5, 124.9 (q, J=3.7Hz), 123.9 (q, J=272.3Hz), 122.7 (q, J=3.8Hz), 118.5, 115.9, 50.9
3
1
P NMR (81 MHz, D O): δ 13.7
2
-
HRMS (ESI-) calculated for C H F N P O S m/z [M - H] : 467.98014, found m/z
1
4
11
3
3
2
6
4
67.978033
(((6-(4-methoxyphenyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid (6j)
Isolated as a pale yellow powder; 30.8 mg (71% overall isolated yield).
1
H NMR (400 MHz, D O): δ 8.11 (s, 1H), 7.58(s, 1H), 7.57 (d, J=8.0Hz, 2H), 6.91 (d, J=8.0Hz,
2
2
H), 3.70 (s, 3H); central methylene proton obscured by solvent signal.
1
3
C NMR (126 MHz, D O): δ 162.7, 159.2, 156.0, 153.2, 139.1, 127.5, 126.3, 118.7, 114.6,
2
1
13.4, 55.3, C observed by HSQC
1
13
1
13
HSQC ( H – C): H δ 4.74 correlates with C δ 49.0.
3
1
P NMR (81 MHz, D O): δ 13.7
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 430.00332, found m/z 430.00442
1
4
14
3
2
7
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 51
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-(4-isopropoxyphenyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid (6k)
Isolated as a white powder; 21.9 mg (54% overall isolated yield).
1
H NMR (500 MHz, D O): δ 8.17 (s, 1H), 7.69 (s, 1H), 7.67 (d, J=8.8Hz, 2H), 7.02 (d, J=8.8Hz,
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
2
H), 1.27 (d, J=6.1, 6H); central methylene proton and i-Pr-CH obscured by solvent signal.
1
3
C NMR (126 MHz, D O): δ 164.4, 158.9, 157.7, 155.0, 140.7, 129.2, 128.2, 120.4, 118.6,
2
1
15.3, 73.3, 22.7, C observed by HSQC
1
13
1
13
1
13
HSQC ( H – C): H δ 4.60 correlates with C δ 73.3 and H δ 4.47 correlates with C δ
5
0.4.
31
P NMR (81 MHz, D O): δ 13.7
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 458.03462, found m/z 458.03532
1
6
18
3
2
7
(((6-(4-cyclopropoxyphenyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid
(6l)
Isolated as a beige powder; 16.2 mg (38% overall isolated yield)
1
H NMR (500 MHz, D O): δ 8.12 (s, 1H), 7.63 (s, 1H), 7.62 (d, J=8.7Hz, 2H), 7.10 (d, J=8.7Hz,
2
2H), 4.42 (br s, 1H), 3.82-3.80 (m, 1H), 0.74-0.71 (m, 2H), 0.65-0.62 (m, 2H) ; central
methylene proton obscured by solvent signal.
1
3
C NMR (126 MHz, D O): δ 162.7, 158.5, 156.0, 153.3, 139.1, 127.4, 126.7, 118.7, 115.7,
2
113.6, 51.3, 5.45, C observed by HSQC at ~50
31
P NMR (81 MHz, D O): δ13.6
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 456.0190, found m/z 456.0189
1
6
16
3
2
7
2
8
ACS Paragon Plus Environment

Page 29 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
((6-(4-cyclopropylphenyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)diphosphonic acid
(6m)
Isolated as a light brown powder 18.8 mg (47% overall isolated yield).
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
1
H NMR (400 MHz, D O): δ 8.13(s, 1H), 7.70 (s, 1H), 7.58(d, J=8.3Hz, 2H), 7.11(d, J=8.3Hz,
2
2H), 4.46 (t, J=18.7Hz, 1H), 1.88-1.84 (m, 1H), 0.93-0.86 (m, 2H), 0.66-0.62 (m, 2H)
1
3
C NMR (75 MHz, D O): δ 165.4, 158.7, 156.0, 148.0, 141.8, 132.9, 128.7, 128.6, 121.2,
2
1
16.6, 17.1, 11.9, C observed by HSQC
1
13
1
13
HSQC ( H – C): H δ 4.46 correlates with C δ 50.9.
31
P NMR (81 MHz, D O): δ 13.8
2
-
HRMS (ESI-) calculated for C H N P O S m/z [M - H] : 440.02405, found m/z 440.02414
1
6
16
3
2
6
(((6-(4-(2,2-difluorocyclopropyl)phenyl)thieno[2,3-d]pyrimidin-4-yl)amino)methylene)
diphosphonic acid (6n)
Isolated as a pale yellow solid 26.7 mg (50% overall isolated yield).
1
H NMR (500 MHz, D O): δ 8.27 (s, 1H), 7.86 (s, 1H), 7.75 (d, J=8.4Hz, 2H), 7.38 (d, J=8.4Hz,
2
2H), 4.63 (t, J=18.9Hz, 1H), 2.98-2.91 (m, 1H), 1.96-1.90 (m, 1H), 1.86-1.78 (m, 1H)
1
3
C NMR (126 MHz, D O): δ 163.2, 156.2, 153.4, 138.9, 134.3, 132.0, 128.7, 126.0, 118.5,
2
1
15.6, 113.3 (t, J=285Hz), 26.38 (t, J=11.2Hz), 16.1 (t, J=10.3Hz), C observed by HSQC
1
13
1
13
HSQC ( H – C): H δ 4.63 correlates with C δ 50.5.
31
P NMR (81 MHz, D O): δ 13.7
2
-
HRMS (ESI-) calculated for C H F N P O S m/z [M - H] : 476.0052, found m/z 476.0046
1
6
14
2
3
2
6
2
9
ACS Paragon Plus Environment