The identification of novel 5′-amino gemcitabine analogs as potent RRM1 inhibitors

Article abstract of DOI:10.1016/j.bmc.2014.02.007

The ribonucleotide reductase (RNR) enzyme is a heteromer of RRM1 and RRM2 subunits. The active enzyme catalyzes de novo reduction of ribonucleotides to generate deoxyribonucleotides (dNTPs), which are required for DNA replication and DNA repair processes.

Full text of DOI:10.1016/j.bmc.2014.02.007

Bioorganic & Medicinal Chemistry 22 (2014) 2303–2310
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
The identification of novel 50-amino gemcitabine analogs as potent
RRM1 inhibitors
b
c
c
c
Marc A. Labroli a, , Michael P. Dwyer , Ruichao Shen , Janeta Popovici-Muller , Qinglin Pu ,
Daniel Wyss a, Mark McCoy a, Dianah Barrett a, Nicole Davis d, Wolfgang Seghezzi d, Frances Shanahan d,
Lorena Taricani d, Maribel Beaumont d, Maria-Christina Malinao d, David Parry d, Timothy J. Guzi c
⇑
a Merck Research Laboratories, 2000 Galloping Hill Road, Kenilworth, NJ 07033, USA
b Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
c Merck Research Laboratories, 33 Avenue Louis Pasteur, BMB-3, Boston, MA 02115, USA
d Merck Research Laboratories, 901 South California Avenue, Palo Alto, CA 94304, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The ribonucleotide reductase (RNR) enzyme is a heteromer of RRM1 and RRM2 subunits. The active
enzyme catalyzes de novo reduction of ribonucleotides to generate deoxyribonucleotides (dNTPs), which
are required for DNA replication and DNA repair processes. Complexity in the generation of physiologi-
cally relevant, active RRM1/RRM2 heterodimers was perceived as limiting to the identification of selec-
tive RRM1 inhibitors by high-throughput screening of compound libraries and led us to seek alternative
methods to identify lead series. In short, we found that gemcitabine, as its diphosphate metabolite, rep-
resents one of the few described active site inhibitors of RRM1. We herein describe the identification of
novel 50-amino gemcitabine analogs as potent RRM1 inhibitors through in-cell phenotypic screening.
Ó 2014 Elsevier Ltd. All rights reserved.
Received 17 October 2013
Revised 29 January 2014
Accepted 7 February 2014
Available online 15 February 2014
Keywords:
Ribonucleotide reductase
Gemcitabine
CHK1
Deoxyribonucleotides
Oncology
tert-
p53R2 can substitute for RRM2 to form an active enzyme with
RRM1.8
The key role of RNR in DNA synthesis and cell growth has made
1. Introduction
it an important target for anticancer therapy.9–11 Non-selective
inhibitors of RNR activity such as hydroxurea (HU), clofarabine
(CAFdA), gemcitabine (GEM), Trimidox, and Didox have been
investigated for the treatment of a wide variety of solid tumors
and hematologic malignancies.12 Many of these agents suppress
dNTP levels and inhibit DNA replication.13–15 Thus, exposure to
multiple mixed-function non-selective DNA antimetabolites can
induce a coordinated series of intra-S events that leads to replica-
tion fork stalling likely to participate in S-phase checkpoint
activation.16,17
Ribonucleotide reductase (RNR) catalyzes the reduction of ribo-
nucleotides to deoxyribonucleotides, the essential precursors of
DNA synthesis. RNR is the rate limiting enzyme during de novo
synthesis of deoxyribonucleotides (dNTPs).1 RNR enzyme plays a
central role in the maintaining a balanced supply of dNTPs re-
quired for DNA synthesis and repair. Thus, RNR plays an important
role in genetic fidelity and cell viability.2,3 Of note, failure to ade-
quately control dNTP levels leads to cell death and genetic
abnormalities.4,5
20-Deoxy-20,20-difluorocytidine (gemcitabine) is an antimetabo-
lite nucleoside approved for the treatment of pancreatic, breast,
and non-small cell lung cancers either as a single agent or in com-
bination with other chemotherapeutic agents.18 Gemcitabine is a
prodrug which is intracellularly phosphorylated to active diphos-
phate and triphosphate forms.19 Gemcitabine diphosphate inhibits
ribonucleotide reductase through binding to the active site of
RRM1 leading to depletion of the deoxynucleotide pool and halt
of DNA synthesis.20 Gemcitabine triphosphate competes with
deoxycytidine triphosphate for incorporation into DNA resulting
The classical ribonucleotide reductase of the de novo pathway
comprises of two subunits, RRM1 and RRM2.1 The large subunit,
RRM1, contains the catalytic site, the substrate-specificity site
and the activity site.3 The RRM2 subunit contains an iron center-
generated tyrosyl free radical that can be scavenged by hydroxy-
urea.6 An additional RRM2 subunit, p53R2 has been identified.7
⇑
Corresponding author. Tel.: +1 908 740 3168.
E-mail address: marc.labroli@merck.com (M.A. Labroli).
http://dx.doi.org/10.1016/j.bmc.2014.02.007
0968-0896/Ó 2014 Elsevier Ltd. All rights reserved.

2304
M. A. Labroli et al. / Bioorg. Med. Chem. 22 (2014) 2303–2310
in termination of DNA polymerization.21 Together the effects elic-
ited by individual gemcitabine metabolites result in the observed
anticancer response.
Table 1
Preparation of 6a–d by SN2 substitution
Entry Amines
Product 6
Yield
(%)
Since the discovery of gemcitabine by Hertel et al.22 numerous
gemcitabine derivatives have been synthesized in the search for
new anticancer or antiviral agents.23 For example, the base-modi-
fied gemcitabine derivatives include adenosine, guanine and uracil
analogs.24 The ribose-modified derivatives include 40-azido ana-
logs, 40-allene substituted analogs, 30-deoxy analogs, and thio/
aza/carbocyclic analogs.25–28 However, due to the perceived
requirement for phosphorylation of the 50-OH to initiate the pro-
duction of active NDP and NTP metabolites, replacement of the
50-OH by other heteroatoms has been left virtually unexplored.
Crystal structures of RRM1 bound to the diphosphate metabolite
revealed a discreet diphosphate binding pocket with phosphate
binding mediated by a helix dipole interaction and several electro-
static interactions.29 As our interest (vide supra) was in identifica-
tion of selective RRM1 inhibitors, we hypothesized that
gemcitabine diphosphate might serve as a template for the identi-
fication of such molecules. Thus, identification of 50-substitution in
replacement of the diphosphate with limited other structural
changes may lead to effective competitors with the natural ligand.
Likewise, the inability to form the triphosphate metabolite in situ
would make these analogs devoid of competing incorporation into
DNA. We focused our initial efforts on amine replacements for sev-
eral reasons including the inability to revert back to gemcitabine
(esp. in vivo) and ease of synthesis to rapidly test the hypothesis
that something other than a diphosphate (or diphosphate mimic)
would work.
O
NH2
NH2
O
O
N
N
N
H
6a
32a
O
N
F
HO
F
NH2
N
NH2
O
N
O
N
53a
O
N
H
6b
O
F
HO
F
Me
NH2
Me
NH
Me
N
Me
N
N
NH2
O
N
N
53b
N
6c
N
H
O
F
HO
O
F
NH2
N
N
N
N
N
6d
48b
N
O
N
F
HO
F
a
Heating at 100 °C.
Heating at 90 °C.
b
Our initial route to the 50-amino-20,50-dideoxy-20,20-dif-
luorocytidine analogs is outlined in Scheme 1. The synthesis began
with protection of the 50-hydroxy group of gemcitabine hydrochlo-
ride (1) as the butyldimethylsilyl (TBDMS) ether followed by global
benzoyl protection to afford compound 2. Desilylation of the 50-
TBDMS ether with tetrabutylammonium fluoride (TBAF) at 0 °C
provided alcohol 3 followed by conversion to tosylate 4 under
standard conditions. Treatment of intermediate 4 with 7 M NH3
in MeOH cleaved the benzoate groups and the resulting tosylate
was treated with the requisite amine with heating to afford prod-
ucts 6a,b (Table 1). Alternatively, alcohol 3 was converted to the io-
dide intermediate 5 using methyltriphenoxyphosphonium iodide
followed by benzoyl deprotection with 7 M NH3 in MeOH. Treat-
ment of iodide 5 with excess amine (3 equiv) at 90 °C provided
50-amino-20,50-dideoxy-20,20-difluorocytidine analogs 6c,d (Table 1).
Utilization of iodide 5 allowed less amount of amine to be used in
the final displacement reaction which made the purification of the
final products more straightforward. The examples listed in Table 1
represent a limited set of compounds using the methodology out-
lined in Scheme 1 and further studies outlining the scope of this
work has recently been published.30
2. Chemistry
The 50-amino derivatives of gemcitabine, a class of synthetically
accessible analogs capable of displaying diverse functionality that
might catalyze binding to RRM1 have rarely been reported to date.
We therefore embarked on an effort to prepare 50-amino substi-
tuted gemcitabine analogs as novel RRM1 inhibitors. A variety of
50-amino gemcitabine analogs were prepared according to the gen-
eral scheme shown in Scheme 1 with several representative exam-
ples shown in Table 1.
NH2
HCl
NHBz
c
O
O
N
N
N
N
a, b
HO
TBSO
O
O
O
F
F
HO
BzO
F
F
2
1
NHBz
NHBz
O
O
N
N
N
N
HO
TsO
d
3. Biological results
3.1. H2AX assay
O
F
F
BzO
BzO
F
F
3
4
g, e
e, f
Complexity in the generation of physiologically relevant active
RRM1/RRM2 complexes as a method for robust in vitro calibration
of activity led us to seek alternative methods to characterize the
novel compounds. Activation of the DNA replication checkpoint
is one of the hallmarks of disruption of dNTP production by RNR.
Previously, we had established a functional readout for override
of the DNA replication checkpoint by monitoring gamma H2AX
induction following treatment with the replication inhibitor
hydroxyurea and CHK1 inhibitor SCH 900776.31 Similar induction
of gamma H2AX was observed following knockdown of RRM1 with
siRNA and subsequent treatment with SCH 900776 (unpublished
NH2
N
NH2
R1
O
N
O
N
F
N
N
R2
f
I
O
O
F
HO
HO
F
F
6a-d
5
Scheme 1. Reagents and conditions: (a) TBSCl, imidazole, DMF, rt, 12 h, 92%; (b)
BzCl, DMAP, pyridine, rt, 12 h, 90%; (c) TBAF, THF, 0 °C, 3.5 h then AcOH, 85%; (d)
TsCl, pyridine, Et3N, 85%; (e) NH3, MeOH, 92%; (f) R1R2NH, DMF, 32-55%; (g)
MeP(OPh)3I, DMF, 60 °C, 84%.

M. A. Labroli et al. / Bioorg. Med. Chem. 22 (2014) 2303–2310
2305
results). Hence, adaptation of this approach was reasoned to be
3.2. NMR binding studies
effective in evaluating compounds for biological activity. Hence,
treating cells with 50-amino analogs 6a–d overnight followed by
exposure to SCH 900776 led to compound and dose dependent
induction of H2AX (Fig. 1). These phenotypes matched those ob-
served with the RRM1 siRNA knockdown (unpublished results) as
well as that observed with gemcitabine shown in Figure 1. This as-
say was also useful for tracking SAR trends in a high throughput
fashion. As depicted in Figure 1, compounds 6c,d demonstrated a
more pronounced response both in the presence and absence of
SCH 900776 compared to 6a,b. Moreover, comparison of 6d to
RRM1 inhibitor gemcitabine in this assay demonstrated compara-
ble responses (Fig. 1). With in-cell demonstration of checkpoint
activation in hand, it remained to prove that this series of com-
The large subunit of ribonucleotide reductase, RRM1, is a 90 kDa
protein that binds nucleotides but has no enzymatic activity. In
solution, RRM1 is a 180 kDa homodimer that is difficult study
structurally but is well-suited to ligand-detected NMR experi-
ments. We probed ligand binding to RRM1 with 3 different NMR
methods. Each method relies on different effects (line broadening,
saturation transfer and NOE) that are transferred to the ligand and
detected in its solution NMR spectrum.
Small molecule line broadening was detected in 1H NMR spec-
tra of 6d acquired in the presence of RRM1 (Fig. 2). Ligand signals
are reduced up to 27% by the addition of 1.8 lM RRM1 as com-
pared to reference sample with the ligand and buffer but no pro-
pounds was acting in
gemcitabine.
a mechanistically similar way to
tein. Signal reductions are a result of protein binding and are
consistent with an approximate Kd of 200 l
M.32 The DMSO peak
Anti-phospho H2A.X---U2OS cells Compound 6a o/n then exposed
to +/- 0.5uM 900776
Anti-phospho H2A.X---U2OS cells Compound 6b o/n then exposed
to +/- 0.5uM 900776
100
100
+SCH 900776
-SCH 900776
+SCH 900776
-SCH 900776
80
80
60
40
20
0
60
40
20
0
µM
µM
Anti-phospho H2A.X---U2OS cells Compound 6d o/n then exposed
to +/- 0.5uM 900776
Anti-phospho H2A.X---U2OS cells Compound 6c o/n then exposed
to +/- 0.5uM 900776
100
100
+SCH 900776
-SCH 900776
+SCH 900776
-SCH 900776
80
80
60
40
20
0
60
40
20
0
µM
µM
Anti-phospho H2A.X---U2OS cells gemcitabine o/n then exposed
to +/- 0.5uM 900776
100
80
60
40
20
0
+SCH 900776
-SCH 900776
µM
Figure 1. H2AX response for 6a–d and gemcitabine in the presence and absence of SCH 900776.

2306
M. A. Labroli et al. / Bioorg. Med. Chem. 22 (2014) 2303–2310
6d
6d
6d
DMSO
ATP
Figure 2. Compound binding to RRM1. Line-broadening in 1H NMR spectra was
used to assess 6d binding to RRM1. The intensity of 6d NMR peaks is reduced with
increasing amounts of RRM1 protein (0, 0.9 and 1.8 lM), while the DMSO peak is
unaffected. The decrease in peak intensity is due to line-broadening of 6d while it is
bound to RRM1.
Figure 4. Structural data from transferred NOE experiments. A transferred NOE
data set for a complex of RRM1, ATP and 6d is shown. Both ATP and 6d showed
many NOE cross peaks that report on their RRM1-bound conformation.
CDP
6d
6d
yielded a number of ligand NOEs (Fig. 4) that are opposite in sign
to NOEs of the free ligands. This data confirms RRM1 binding of
both ATP and 6d and provides 1H–1H distance restraints on
RRM1-bound 6d.
CDP
3.3. NMR Activity Studies
1H NMR measurements can also be used as a convenient read-
out of RNR activity. RNR reactions require active enzyme, ATP
and Mg2+ and, when initiated with CDP, yield dCDP as a product.
ATP, CDP and dCDP each have distinctive NMR peaks making it
possible to monitor the progress of a RNR reaction real-time.
Figure 5 shows the extent of dCDP production and CDP consump-
tion when the RNR reaction was quenched after 1 h. Addition of
ATP
dCDP
6d
CDP
Figure 3. Binding of 6d to the active site of RRM1. STD-NMR spectroscopy was used
to test the ability of 6d to compete with substrate. CDP binds to the active site of
RRM1 and gives a large STD-NMR signal in the presence of 0.9 lM RRM1 (black).
Addition of 6d results in the appearance of new STD peaks for 6d and a reduction of
CDP STD-NMR peaks (red). 1H NMR spectra for CDP (blue) and the CDP, 6d mixture
(green) are shown for reference.
was unaffected in these measurements and can be used as a refer-
ence. The binding of 6d to RRM1 was also detected by saturation
transfer difference NMR (STD-NMR).33 Reference STD-NMR data
was first acquired on CDP, which binds to the RRM1 active site
and is a substrate of the active enzyme (Fig. 3). Addition of 6d re-
sults in a new set of STD-NMR peaks for 6d, from binding to RRM1,
and a decrease of the CDP STD-NMR signal due to competition
Figure 5. Inhibition of active RNR. Reaction products from active RNR are detected
by 1H NMR. 61
M of dCDP is produced by RNR 1 h after the reaction was initiated
with 100 M CDP (black). Addition of 100 M 6d inhibits RNR resulting in a 32%
reduction of dCDP production.
(Ki ꢀ 150
carried out to obtain structural information on the RRM1-bound
conformation of 6d.35 A single experiment with 9
M RRM1
l
M) with CDP.34 Transferred NOESY experiments were
l
l
l
l

M. A. Labroli et al. / Bioorg. Med. Chem. 22 (2014) 2303–2310
2307
Table 2
In order to evaluate these derivatives for biological activity, we
established the viability of using a previously established func-
tional readout by monitoring gamma H2AX induction following
treatment of replication inhibitor hydroxyurea and the CHK1
inhibitor, SCH 900776. Using this in-cell assay, we noted dose-
Quantitation (AUC) of dNTP & NTP pools in the CEM cells
dNTPs & NTPs
Untreated
6d (1 lM)
dCTP
dGTP
dATP
dTTP
CTP
GTP
ATP
UTP
0.11 0.01
0.09 0.03
0.60 0.11
0.25 0.09
2.68 0.09
4.00 0.13
38.89 2.50
10.05 0.56
0.03 0.01
0.03 0.01
0.14 0.02
0.11 0.03
1.48 0.14
1.25 0.04
17.83 0.86
3.59 0.31
dependent induction of
c-H2AX for this series of compounds
(Fig. 1) and could rank order the derivatives based upon how ro-
bust the activity was observed. Based upon the response in this as-
say, 6d was selected as a representative example for further
characterization in order to establish that this series of compounds
was behaving in a mechanistically similar way to gemcitabine
diphosphate. Towards this end, a series of NMR experiments were
conducted to establish ligand binding to RRM1. Small molecule
line broadening was detected in the 1H NMR spectra of 6d
(Fig. 2) in the presence of RRM1 and the resultant ligand signal
reduction (27%) are consistent with protein binding at an approx-
Analyses performed in triplicate.
Denotes standard deviation.
Units of measurement are AUC (area under the curve).
100 lM 6d yielded 32% inhibition of dCDP production and 32%
reduction of CDP consumption when compared with control
imate Kd of 200 lM. Furthermore, saturation transfer difference
spectra.
NMR (STD-MNR) analysis revealed a new set of STD-NMR signals
for 6d which could be attributed to binding to RRM1. In addition,
there was a decrease of the CDP STD-NMR signal due to competi-
The combination of binding and functional NMR data estab-
lishes that 6d binds to the active site of RRM1, competes with
the substrate and products for binding to the inactive enzyme
and inhibits active RNR in vitro.
tion by the ligand (Ki ꢀ 150
lM) with CDP indicative of the binding
of ligand to RRM1. The RRM1-bound conformation of 6d was mea-
sured by transferred NOESY experiments. A single experiment was
conducted with ligand and RRM1 which demonstrated ligand NOEs
that were of the opposite sign of the free ligand NOE’s which con-
firms RRM1 binding to both ATP and ligand. Finally, 1H NMR mea-
surements demonstrated 32% inhibition of dCDP production and
32% reduction of CDP consumption compared to control spectra
giving a convenient readout of RNR activity. The combination of
binding and functional NMR established that 6d binds to the active
site of RRM1, effectively competes with both substrate and product
binding to inactive enzyme, and inhibits active RNR in vivo. Finally,
6d was shown to reduce dNTP pools in U2OS cells consistent with
observations made for gemcitabine using an HPLC method.
3.4. dNTP suppression studies
Studies have shown that GEM treatment suppresses dCTP,
dGTP, and dATP pools.36–38 To confirm effects on dNTP levels, we
compared 6d with gemcitabine, using an HPLC-based assay.39 In
CEM control cells, dGTP was the smallest dNTP pool on the basis
of AUC (area under the curve) (0.09 AUC 0.03), compared with
dCTP (0.11 AUC 0.01), dATP (0.6 AUC 0.11), dTTP (0.25
AUC 0.09) (Table 2). Treatment with 6d resulted in depletion of
dATP, dCTP, dGTP (Table 2). Thus, treatment of cells with 6d for
2 h leads to suppression of discrete dNTP pools (see Table 2).
4. Discussion
5. Conclusions
The ribonucleotide reductase (RNR) enzyme is a heteromer of
RRM1 and RRM2 subunits where the active enzyme catalyzes de
novo reduction of ribonucleotides to generate deoxyribonucleo-
tides (dNTPs), which are required for DNA replication and DNA re-
pair processes. Complexity in the generation of physiologically
relevant, active RRM1/RRM2 heterodimers was perceived as limit-
ing to the identification of selective RRM1 inhibitors by high-
throughput screening of compound libraries and led us to seek
alternative methods to identify lead series. We found that gemcit-
abine, as its diphosphate metabolite, represents one of the few de-
scribed active site inhibitors of RRM1. However, due to the
perceived requirement for phosphorylation of the 50-OH of gemcit-
abine to initiate the production of active NDP and NTP metabolites,
replacement of the 50-OH has been left virtually unexplored. To-
wards the identification of selective RRM1 inhibitors, we hypothe-
sized that gemcitabine diphosphate might serve as an effective
template for the identification of such molecules. Thus, identifica-
tion of 50-substitution in replacement of the diphosphate with lim-
ited other structural changes may lead to effective competitors
with the natural ligand. Likewise, the inability to form the triphos-
phate metabolite in situ would make these 50-analogs devoid of
competing incorporation into DNA.
In summary, we have demonstrated that a series of novel
50-amino gemcitabine analogs are potent RRM1 inhibitors through
a combination of in-cell phenotypic screening, direct binding NMR
studies, and in vivo dNTP reduction studies. These analogs offer
great promise as novel therapeutics for cancer and additional work
in this vain will be described in due course.
6. Experimentals
6.1. Biology experimentals
6.1.1. c-H2AX assay
Briefly, analogs were titrated onto cells to induce the activation
of CHK1. Control populations were left untreated. SCH 900776 was
then added to cells over a two hour exposure window (in the pres-
ence of the analog). Following the 2 h co-exposure to SCH 900776,
cells were fixed and permeabilized (70% ethanol) before staining
with a FITC-conjugated anti-c-H2AX monoclonal antibody (Milli-
pore). Cells were counterstained with propidium iodide and subse-
quently analyzed using flow cytometry (Becton Dickinson LSR II) or
the Discovery 1 immunofluorescence platform (Molecular De-
vices). Data are presented as the percentage of
c-H2AX positive
Towards this end, we prepared a novel series of 50-amino deriv-
atives of gemcitabine that could catalyze binding to RRM1 using
the synthetic methods described in Scheme 1. This series of 50-ami-
no gemcitabine analogs was demonstrated to be potent RRM1
inhibitors thru a combination of in-cell phenotypic screening,
NMR studies, as well as functional readouts to confirm this finding.
cells, and thus reflect the overall penetrance of the
c-H2AX
phenotype.
6.1.2. NMR spectroscopy
All data was collected at 27 °C on a Bruker Avance DRX 500
spectrometer using a 5 mm TXI cryoprobe. Binding and structural

2308
M. A. Labroli et al. / Bioorg. Med. Chem. 22 (2014) 2303–2310
studies were carried out on yeast-RRM1 protein in 50 mM NaiPO4
D2O buffer, pH 7.4 with 5 mM d-DTT and 5 mM MgCl2. Line broad-
ening of 6d was measured in 1H-NMR spectra with and without
protein. Competition STD NMR measurements were carried out
6.2. Chemistry experimentals
Mass spectra were recorded using either an EXTREL 401 (CI),
JEOL or MAT-90 (FAB), VG, ZAB-SE (SIMS) or Finnigan MAT-CH-%
(EI) spectrometer. NMR structure determinations of the com-
pounds were made using chemical shifts, coupling constants, cou-
pling information from COSY spectra, and 1D NOE experiments. 1H
NMR and 13C NMR spectra were obtained on Varian XL 400
(400 MHz, 1H; 100 MHz, 13C) and are reported as ppm down field
from Me4Si with number of protons, multiplicities, and coupling
constants in Hertz indicated parenthetically. For 13C NMR, a Nal-
orac Quad nuclei probe was used. Compound purity was checked
by TLC and LC/MS analysis using an Applied Biosystems API-100
mass spectrometer and Shimadzu SCL-10A LC column: Altech plat-
inum C18, 3 micron, 33 mm ꢁ 7 mm ID; gradient flow: 0 min—10%
CH3CN, 5 min—95% CH3CN, 7 min—95% CH3CN, 7.5 min—10% CH3-
CN, 9 min stop. Chromatography was performed with Selecto Sci-
using 0.9
l
M RRM1 with 200
M 6d. Transferred NOESY data were collected on a sam-
M RRM1, 200 M ATP and 200 M 6d using a mix
lM ATP, 200 lM CDP with and with-
out 200
l
ple of 9
l
l
l
times ranging from 75 to 200 ms.
6.1.3. Functional RNR assay with NMR readout
Reactions were performed using human RNR proteins at 27 °C.
The reaction buffer consisted of 50 mM d-Tris (D2O), pH 7.6,
10 mM d-DDT, 70
6d was incubated for 30 min with active RNR enzyme (0.8
RRM1, 5.6 M RNR2) and 1 mM ATP. Reactions were initiated with
the addition of 200 M CDP and quenched by heating after 1 h.
CDP consumption and dCDP production were measured by 1H
NMR with 25 M TSP added as an internal standard. Percent inhi-
lM Fe(NH4)(SO4)2 and 6 mM MgCl2. Compound
lM
l
l
l
entific flash silica gel, 32–63 lM.
bition with 6d was compared to a reference spectrum where active
RNR was incubated with d-DMSO. Reaction controls include dATP,
didox and gemcitabine diphosphate.
NH2
NH2
HCl
O
O
N
N
N
N
TBSO
HO
6.1.4. Cell lines
O
O
F
F
HO
HO
F
Tumor cell lines were obtained from the American Type Culture
Collection (ATCC) and cultured under the suggested growth
conditions.
F
1
1a
6.1.5. NTP/dNTP extraction
Compound 1a: A solution of gemcitabine-hydrochloride (10.9 g,
36.6 mmol) in DMF (80 mL) at 25 °C was treated successively with
imidazole (7.4 g, 109 mmol) and tert-butyldimethylsilyl chloride
(6.0 g, 40 mmol). The solution was stirred at 25 °C for 15 h
whereupon the reaction mixture was concentrated under reduced
pressure. The resultant material was purified by silica gel chroma-
tography using a gradient of 5–20% MeOH in CH2Cl2 to provide 1a
(13.4 g, 98%) as a clear oil. 1H NMR (DMSO-d6) d 7.64 (1H, d,
J = 7.7 Hz), 7.41 (2H, br s), 6.14 (1H, dd, J = 7.9, 7.6 Hz), 5.76 (1H, d,
J = 7.7 Hz), 4.12 (1H, m), 3.94 (1H, br d, J = 12.4 Hz), 3.86 (1H, m),
3.80 (1H, dd, J = 12.4, 3.0 Hz), 0.89 (9H, s), 0.086 (6H, s). Mass
calculated for formula C15H25F2N3O4Si 377.2, observed LCMS m/z
378.3 (M+H).
NTP/dNTP Extraction was performed as described previously.39
The cell pellets were resuspended in ice-cold ultrapure water and
deproteinized with equal volume of 6% TCA. Acid cell extracts (cor-
responding to 1 ꢁ 107 cells/ 80
lL) were vortexed for 20 s, ice
bathed for 10 min, vortexed again for 20 s, and then centrifuged at
13,000 rpm for 10 min at 4 °C. The resulting supernatants were fro-
zen on dry-ice and stored at ꢂ80 °C prior to analysis. Before HPLC
analysis, samples were thawed and aliquots of 60
lL were neutral-
ized with 0.5 lL of 0.67 % NH4OH. 50 lL of this solution was injected
onto the HPLC column. All experiments were done in triplicate.
6.1.6. Instrument and chromatographic conditions and
measurement of NTP/dNTP levels
NHBz
NH2
The chromatographic system consisted of an HPLC Ettan LC (GE
Healthcare) with a Halo column from MacMod Analytical 2.7 lm
O
O
N
N
N
N
TBSO
TBSO
O
O
(4.6 ꢁ 150 mm). The column was kept at room temperature. The
mobile phase was delivered using the following linear gradient
elution program (rinsing and reequilibration steps included): (%
Buffer A/% Buffer B), flow rate of 1 mL/min (100:0) to (30:70) in
20 min ? flow rate of 0.9 mL/min (30:70) to (0:100) in 12 min ? -
flow rate of 0.9 mL/min hold at (0:100) for 10 min ? flow rate of
0.5 mL/min (0:100) to (50:50) in 5 min ? flow rate of 1 mL/min
(100:0) for 10 column volumes.
Buffer A contained 10 mM tetrabuylammonium hydroxide,
10 mM KH2PO4 and 0.25% MeOH, and adjusted to pH 7.0 with
1 M HCl. Buffer B contained 5.6 mM tetrabuylammonium hydrox-
ide, 50 mM KH2PO4 and 30% MeOH, and adjusted to pH 7.0 with
1 M NaOH. Both Buffer A and B were made fresh for experiment
F
F
BzO
HO
F
F
2
1a
Compound 2: To a solution of silyl ether 1a (5.88 g, 15.6 mmol) in
pyridine (78 mL) at 25 °C was added 4-dimethylaminopyridine
(3.81 g, 31.2 mmol) portionwise followed by dropwise addition of
benzoyl chloride (7.2 mL, 62.2 mmol). The resultant solution was
stirred at 25 °C for 15 h whereupon the reaction was quenched with
water. The crude reaction mixture was extracted with EtOAc and the
organic layers were combined. The organic layer was washed with
5% aqueous NaHCO3 solution, brine solution, dried over Na2SO4,
and concentrated in vacuo. The resultant residue was purified by sil-
ica gel chromatography with a gradient of 10–100% EtOAc/hexane
provided compound 2 (7.8 g, 85%). 1H NMR (CDCl3) d 8.74 (1H, br
s), 8.17 (1H, br d, J = 7.0 Hz), 8.09 (2H, dd, J = 8.3, 1.5 Hz), 7.91 (2H,
br d, J = 7.1 Hz), 7.63 (2H, m), 7.54–7.48 (5H, m), 6.58 (1H, t,
J = 8.0 Hz), 5.74 (1H, m), 4.34 (1H, m), 4.08 (1H, dd, J = 12.0,
2.3 Hz), 3.95 (1H, dd, J = 12.0, 2.5 Hz), 0.96 (9H, s), 0.16 (6H, s). Mass
calculated for formula C29H33F2N3O6Si 585.2, observed LCMS m/z
586.2 (M+H).
and passed through 0.22
lm filters to avoid microbial contamina-
tion. Both solvents were degassed. The injection volume was 50
l
L.
Detection was UV absorption at 260 nm. Chromatograms are ana-
lyzed for peak integration on Unicorn version 5.2 (GE Healthcare)
to calculate the area under the curve to quantitate the levels of
dNTPs (dCTP, dGTP, dATP, dTTP) and for the NTPs (CTP, GTP, ATP,
UTP) in U20S cells.

M. A. Labroli et al. / Bioorg. Med. Chem. 22 (2014) 2303–2310
2309
32%) was obtained. 1H NMR (CD3OD) d 7.67 (d, J = 8.0 Hz, 1H), 7.44
(d, J = 2.2 Hz, 1H), 6.25–6.36 (m, 1H), 6.27 (d, J = 3.6 Hz, 1H), 6.18
(t, J = 8.0 Hz, 1H), 5.90 (d, J = 7.3 Hz, 1H), 4.07–4.14 (m, 1H), 3.91–
3.96 (m, 1H), 3.83 (s, 2H), 3.03 (dd, J = 2.9 Hz, J = 13.9 Hz, 1H), 2.94
(dd, J = 6.6 Hz, J = 13.2 Hz, 1H). Mass calculated for formula C14H16-
F2N4O4 342.1, observed LCMS m/z 343.0 (M+H).
NHBz
NHBz
O
O
N
N
N
N
TBSO
HO
O
O
F
F
BzO
BzO
F
F
2
3
O
NH2
Compound 3: To a solution of the silyl ether 2 (7.9 g, 13.5 mmol) in
THF (270 mL) at 0 °C was added a 1.0 M solution of tetrabutylammo-
nium fluoride in THF (14.9 mL, 14.9 mmol). The solution was stirred
at 0 °C for 3.5 h and then quenched with AcOH (1.16 mL, 20.3 mmol)
and water (600 mL). EtOAc (200 mL) was added and the organic
layer was collected. The aqueous layer was extracted with EtOAc
(2 ꢁ 50 mL). The organic layers were combined, washed with brine,
dried over Na2SO4 and concentrated in vacuo. Trituration of the res-
idue with CH2Cl2 followed by filtration provided alcohol 3 (5.4 g,
85%) as a white solid. 1H NMR (DMSO-d6) d 11.50 (1H, br s), 8.36
(1H, br d, J = 7.3 Hz), 8.12 (2H, dd, J = 8.2, 1.2 Hz), 8.06 (2H, dd,
J = 8.2, 1.2 Hz), 7.80 (1H, m), 7.69 (1H, m), 7.65 (2H, m), 7.57 (2H,
m), 7.50 (1H, br d, J = 6.8 Hz), 6.49 (1H, t, J = 8.7 Hz), 5.68 (1H, m),
5.43 (1H, br s), 4.55 (1H, m), 3.91 (1H, br d, J = 12.4 Hz), 3.81 (1H,
br d, J = 12.4 Hz). Mass calculated for formula C23H19F2N3O6 471.1,
observed LCMS m/z 472.0 (M+H).
O
N
N
N
N
H
O
F
HO
F
Compound 6b: Using the procedure outlined for the preparation of
Compound 6a, treatment of tosylate 4 (67 mg, 0.19 mmol) with 1-
(2-aminoethyl)pyrrolidin-2-one (0.20 g, 1.6 mmol) afforded 6b
(32 mg, 53%) as a white solid. 1H NMR (400 MHz, CD3OD) d 7.67
(d, J = 7.5 Hz, 1H), 6.20 (t, J = 8.4 Hz, 1H), 5.94 (d, J = 7.5 Hz, 1H),
4.08–4.14 (m, 1H), 3.91–3.95 (m, 1H), 3.49 (t, J = 7.0 Hz, 2H), 3.34–
3.44 (m, 2H), 3.07 (dd, J = 3.3, 13.4 Hz, 1H), 2.95 (dd, J = 6.6,
13.4 Hz, 1H), 2.79–2.88 (m, 2H), 2.34–2.40 (m, 2H), 2.01–2.08 (m,
2H). mass calculated for C15H21F2N5O4 373.2, observed LCMS m/z
374.2 (M+H).
NHBz
NHBz
O
O
N
N
N
N
NHBz
NHBz
HO
I
O
O
O
O
N
N
N
N
F
F
HO
TsO
BzO
BzO
F
F
O
O
3
3a
F
F
BzO
BzO
F
3
F
4
Compound 3a: Into a round-bottom flask was added compound 3
(5.3 g, 11.24 mmol), methyltriphenoxyphosphonium iodide
(12.0 g, 26.54 mmol) and DMF (80 mL). The reaction mixture was
stirred at 60 °C overnight and quenched with methanol. The crude
mixture was concentrated, diluted with dichloromethane
(200 mL), and washed with 5% sodium thiosulfate solution
(2 ꢁ 100 mL) and brine (3 ꢁ 100 mL), dried over Na2SO4, filtered
and concentrated. Theresidue was purified by column chromatogra-
phy (SiO2, 12–16% ethyl acetate/DCM) to afford 3a (5.47 g; 84%) as a
white solid. Mass calculated for formula C23H18F2IN3O5 581.03, ob-
served LCMS m/z 582.1 (M+H).
Compound 4: A solution of alcohol 3 (2.65 g, 5.62 mmol) in pyridine
(56 mL) at 0 °C was treated successively with p-toluenesulfonyl
chloride (4.27 g, 31.6 equiv) and triethylamine (1.55 mL,
15.8 equiv). The solution was stirred at 0 °C for 3 h and the ice bath
was removed and stirring was continued at 25 °C for 15 h. The solu-
tion was concentrated and the residue was purified using silica gel
chromatography with a gradient of 0–25% acetone in CH2Cl2 to af-
ford tosylate 4 as a white solid (3.06 g, 87%). Mass calculated for for-
mula C30H25F2N3O8S 625.6, observed LCMS m/z 626.1 (M+H).
NHBz
NH2
NH2
NHBz
O
O
N
N
N
N
TsO
O
TsO
O
N
N
N
N
I
I
O
O
F
F
O
BzO
O
HO
F
F
F
F
HO
BzO
F
F
4
4a
3a
5
Compound 4a: A solution of tosylate 4 (2.06 g, 3.29 mmol) in 7 N
NH3 in MeOH solution(100 mL) was stirredat 25 °C for 2 h. The solu-
tion was concentrated and the residue was purified using silica gel
chromatography with a gradient 10–30% MeOH in CH2Cl2 to afford
compound 4a as white solid (0.80 g, 58%) after concentration. Mass
calculated for formula C16H17F2N3O6S 417.1, observed LCMS m/z
418.0 (M+H). This material was used crude in the next transforma-
tion without purification.
Compound 5: To a pressure bottle was added compound 3a (5.40 g,
9.29 mmol) followed by 7 M NH3 in methanol (210 mL, 1470 mmol).
The heterogeneous mixture became clear 30 min later, and was kept
stirring at rt for 2 h. The solution was concentrated and purified by
column chromatography (SiO2, 10–15% methanol/DCM) to afford 5
(3.19 g; 92%) as a white solid. Mass calculated for formula C9H10F2-
IN3O3 373.0, observed LCMS m/z 374.0 (M+H).
NH2
NH2
NH2
NH2
Me
N
Me
O
O
N
N
N
N
O
O
O
I
N
N
N
N
N
H
N
TsO
N
H
O
O
F
O
O
F
HO
HO
F
F
F
5
F
HO
HO
F
F
6c
4a
6a
Compound 6c: Compound 5 (3.04 g, 8.15 mmol) in DMF (3 mL) was
treated with (S)-1-(1-methyl-1H-pyrazol-5-yl)ethanamine (3.43 g,
27.4 mmol) in a 20 mL vial, and was heated to 90 °C for 3.5 h. The
mixture was concentrated and purified by column chromatography
(SiO2, 15% methanol/DCM), followed by lyophilization to afford 6c
(1.6 g, 53%) as a white solid. 1H NMR (400 MHz, CD3OD) d 7.58
Compound 6a: A solution of tosylate 4a (80 mg, 0.19 mmol) and 2-
furylmethylamine (0.085 mL, 0.99 mmol) in DMF (0.2 mL), in a
sealed tube, was heated at 100 °C for 3 h. The solution was cooled
to room temperature and concentrated under reduced pressure.
The residue was purified using silica gel chromatography with a gra-
dient of 0–40% MeOH in CH2Cl2 whereupon an white solid (21 mg,

2310
M. A. Labroli et al. / Bioorg. Med. Chem. 22 (2014) 2303–2310
18. (a) Burris, H. A., III; Moore, M. J.; Andersen, J.; Green, M. R.; Rothenberg, M. L.;
Modiano, M. R.; Cripps, M. C.; Portenoy, R. K.; Storniolo, A. M.; Tarassoff, P.;
Nelson, R.; Dorr, F. A.; Stephens, C. D.; Von Hoff, D. D. J. Clin. Oncol. 1997, 15,
2403; (b) Manegold, C. Expert Rev. Anticancer Ther. 2004, 4, 345; (c) Heinemann,
V. Expert Rev. Anticancer Ther. 2005, 5, 429.
(1H, d, J = 7.4 Hz), 7.38 (1H, d, J = 2.0 Hz), 6.25 (1H, d, J = 2.0 Hz), 6.18
(1H, t, J = 8.6 Hz), 5.86 (1H, d, J = 7.4 Hz), 4.13 (1H, q, J = 6.8 Hz), 4.06
(1H, m), 3.91 (1H, m), 3.82 (3H, s), 2.95 (1H, dd, J = 13.2, 3.2 Hz), 2.86
(1H, dd, J = 13.2, 6.4 Hz), 1.40 (3H, d, J = 6.8 Hz). Mass calculated for
formula C15H20F2N6O3 370.2, observed LCMS m/z 371.2 (M+H).
19. Heinemann, V.; Hertel, L. W.; Grindey, G. B.; Plunkett, W. Cancer Res. 1988, 48,
4024.
20. (a) van der Donk, W. A.; Yu, G.; Pérez, L.; Sanchez, R. J.; Stubbe, J.; Samano, V.;
Robins, M. J. Biochemistry 1998, 37, 6419; (b) Artin, E.; Wang, J.; Lohman, G. J.
S.; Yokoyama, K.; Yu, G.; Griffin, R. G.; Bar, G.; Stubbe, J. Biochemistry 2009, 48,
11622.
21. Huang, P.; Chubb, S.; Hertel, L. W.; Grindey, G. B.; Plunkett, W. Cancer Res. 1991,
51, 6110.
NH2
O
N
N
N
N
O
N
F
HO
F
22. Hertel, L. W.; Kroin, J. S.; Misner, J. W.; Tustin, J. M. J. Org. Chem. 1988, 53, 2406.
23. (a) Hertel, L. W.; Grossman, C. S.; Kroin, J. S.; Mineishib, S.; Chubb, S.; Nowak,
B.; Plunkett, W. Nucleosides Nucleotides 1989, 8, 951; (b) Fahrig, R.; Lohmann,
D.; Rolfs, A.; Dieks, H.; Teubner, J.; Heinrich, J.-C. WO 2008017515, 2008; Chem.
Abstr. 2008, 148, 239457.
24. Smith, D. B.; Kalayanov, G.; Sund, C.; Winqvist, A.; Maltseva, T.; Leveque, V. J.-
P.; Rajyaguru, S.; Le Pogram, S.; Najera, I.; Benkestock, K.; Zhou, Z.; Kaiser, A. C.;
Maag, H.; Cammack, N.; Martin, J. A.; Swallow, S.; Johansson, N. G.; Klumpp, K.;
Smith, M. J. Med. Chem. 2009, 52, 2971.
25. Qiu, Y.-L.; Wang, C.; Peng, X.; Ying, L.; Or. Y. S. WO 2010030858, 2010; Chem.
Abstr. 2010, 152, 335423.
26. Hertel, L. W.; Grossman, C. S.; Kroin, J. S. EP 329348, 1989; Chem. Abstr. 1989,
112, 56592.
Compound 6d: Using the procedure outlined for the preparation of
Compound 6c, treatment of compound 5 (98 mg, 0.26 mmol) with
4,5,6,7-tetrahydropyrazolo[1,5-a]pyrazine (0.10 g, 83 mmol) affor-
ded 6d (46 mg, 48%) as a white solid. 1H NMR (400 MHz, CD3OD) d
8.14 (d, J = 8.2 Hz, 1H), 7.62 (d, J = 2.1 Hz, 1H), 6.34 (d, J = 2.1 Hz,
1H), 6.27 (t, J = 7.9 Hz, 1H), 6.22 (d, J = 8.2 Hz, 1H), 4.80 (bm, 2H),
4.58–4.47 (m, 3H), 4.42–4.36 (m, 1H), 4.05 (t, J = 5.4 Hz, 2H), 4.01–
3.88 (m, 2H). Mass calculated for formula C15H18F2N6O3 368.14, ob-
served LCMS m/z 369.2 (M+H).
27. (a) Qiu, X.-L.; Xu, X.-H.; Qing, F.-L. Tetrahedron 2010, 66, 789; (b) Devos, R.;
Dymock, B. W.; Hobbs, C. J.; Jiang, W.-R.; Martin, J. A.; Merrett, J. H.; Najera, I.;
Shimma, N.; Tsukuda, T. WO 2002018404, 2002; Chem. Abstr. 2002, 136,
217007.
References and notes
28. Guzi, T. J.; Parry, D. A.; Labroli, M. A.; Dwyer, M. P.; Paruch, K.; Rosner, K. E.;
Shen, R.; Popovici-Muller, J. WO 2009061781, 2009; Chem. Abstr. 2009, 150,
515402.
29. Eriksson, M.; Uuhlin, U.; Ramaswamy, S.; Ekberg, M.; Regnström, K. Structure
1997, 5, 1077.
30. Labroli, M. A.; Dwyer, M. P.; Shen, R.; Popovici-Muller, J.; Pu, Q.; Richard, J.;
Rosner, K.; Paruch, K.; Guzi, T. J. Tetrahedron Lett. 2014, 55, 598.
31. Guzi, T. J.; Paruch, K.; Dwyer, M. D.; Labroli, M. A.; Shanahan, F.; Davis, N.;
Taricani, L.; Wiswell, D.; Seghezzi, W.; Penaflor, E.; Bhagwat, B.; Wang, W.; Gu,
D.; Hsieh, Y.; Lee, S.; Liu, M.; Parry, D. Mol. Cancer Ther. 2011, 10, 591.
32. Shortridge, M. D.; Hage, D. S.; Harbison, G. S.; Powers, R. J. Comb. Chem. 2008,
10, 948.
33. Mayer, M.; Meyer, B. A. Angew. Chem., Int. Ed. 1998, 38, 1784.
34. McCoy, M. A.; Senior, M. M.; Wyss, D. F. J. Am. Chem. Soc. 2005, 127, 7978.
35. Fisher, A.; Laub, P. B.; Cooperman, B. S. Nat. Struct. Biol. 1995, 11, 951.
36. Cooper, T.; Ayres, M.; Nowak, B.; Gandhi, V. Cancer Chemother. Pharmacol. 2005,
55, 361.
1. Nordlund, P.; Reichard, P. Annu. Rev. Biochem. 2006, 75, 681.
2. Cory, J. G.; Sato, A. Mol. Cell. Biochem. 1983, 53–54, 257.
3. Thelander, L.; Reichard, P. Annu. Rev. Biochem. 1979, 48, 133.
4. Chabes, A.; Georgieva, B.; Domkin, V.; Zhao, X.; Rothstein, R.; Thelander, L. Cell
2003, 112, 391.
5. Reichard, P. Annu. Rev. Biochem. 1988, 57, 349.
6. Thelander, M.; Gräslund, A.; Thelander, L. J. Biol. Chem. 1985, 260, 2737.
7. Tanaka, H.; Arakawa, H.; Yamaguchi, T.; Shiraishi, K.; Fukuda, S. A. Nature 2000,
404, 42.
8. Guittet, O.; Håkansson, P.; Voevodskaya, N.; Fridd, S.; Gräslund, A.; Arakawa,
H.; Nakamura, Y.; Thelander, L. J. Biol. Chem. 2001, 274, 40637.
9. Szekeres, T.; Fritzer-Szekeres, M.; Elford, H. L. Crit. Rev. Clin. Lab. Sci. 1997, 34,
503.
10. Nocentini, G. Crit. Rev. Oncol. Hematol. 1996, 22, 89.
11. Lien, E. J. Prog. Drug Res. 1987, 31, 101.
12. Shao, J.; Zhou, B.; Chu, B.; Yen, Y. Curr. Cancer Drug Targets 2006, 6, 409.
13. Gandhi, V.; Plunkett, W. Biochem. Pharmacol. 1989, 38, 3551.
14. Xie, K. C.; Plunkett, W. Cancer Res. 1996, 56, 3030.
15. Sampath, D.; Rao, V. A.; Plunkett, W. Oncogene 2003, 22, 9063.
16. Ewald, B.; Sampath, D.; Plunkett, W. Mol. Cancer Ther. 2007, 6, 1239.
17. Lopes, M.; Cotta-Ramusino, C.; Pellicioli, A.; Liberi, G.; Plevani, P.; Muzi-Falconi,
M.; Newion, C. S.; Folani, M. Nature 2001, 412, 557.
37. Gandhi, V.; Mineishi, S.; Huang, P.; Yang, Y.; Chubb, S.; Chapman, A. J.; Nowak,
B. J.; Hertel, L. W.; Plunkett, W. Semin. Oncol. 1995, 22, 61.
38. Heinemann, V.; Schulz, L.; Issels, R. D.; Plunkett, W. Semin. Oncol. 1995, 22, 11.
39. Decosterd, L. A.; Cottin, E.; Chen, X.; Lejeune, F.; Mirimanoff, R. O.; Biollaz, J.;
Coucke, P. A. Anal. Biochem. 1999, 270, 59.