Adenosine-derived inhibitors of 78 kDa glucose regulated protein (Grp78) ATPase: Insights into isoform selectivity

Article abstract of DOI:10.1021/jm101625x

78 kDa glucose-regulated protein (Grp78) is a heat shock protein (HSP) involved in protein folding that plays a role in cancer cell proliferation. Binding of adenosine-derived inhibitors to Grp78 was characterized by surface plasmon resonance and isothermal titration calorimetry. The most potent compounds were 13 (VER-155008) with K_D = 80 nM and 14 with K _D = 60 nM. X-ray crystal structures of Grp78 bound to ATP, ADPnP, and adenosine derivative 10 revealed differences in the binding site between Grp78 and homologous proteins.

Full text of DOI:10.1021/jm101625x

ARTICLE
pubs.acs.org/jmc
Adenosine-Derived Inhibitors of 78 kDa Glucose Regulated Protein
(Grp78) ATPase: Insights into Isoform Selectivity†
Alba T. Macias,* Douglas S. Williamson, Nicola Allen, Jenifer Borgognoni, Alexandra Clay, Zoe Daniels,
Pawel Dokurno, Martin J. Drysdale, Geraint L. Francis, Christopher J. Graham, Rob Howes,
Natalia Matassova, James B. Murray, Rachel Parsons, Terry Shaw, Allan E. Surgenor, Lindsey Terry,
Yikang Wang, Mike Wood, and Andrew J. Massey
Vernalis (R&D) Ltd., Granta Park, Great Abington, Cambridge, CB21 6GB, U.K.
S Supporting Information
b
ABSTRACT: 78 kDa glucose-regulated protein (Grp78) is a heat shock protein (HSP)
involved in protein folding that plays a role in cancer cell proliferation. Binding of adenosine-
derived inhibitors to Grp78 was characterized by surface plasmon resonance and isothermal
titration calorimetry. The most potent compounds were 13 (VER-155008) with KD = 80 nM
and 14 with KD = 60 nM. X-ray crystal structures of Grp78 bound to ATP, ADPnP, and
adenosine derivative 10 revealed differences in the binding site between Grp78 and
homologous proteins.
’ INTRODUCTION
expression or its pharmacological activity could have important
therapeutic benefits in the fight against cancer.
The Hsp70 family of molecular chaperones represents one of
the most evolutionarily conserved groups of chaperone proteins
involved in protein folding.1,2 Hsp70 proteins have an N-term-
inal ATPase and a C-terminal substrate binding domain (SBD)
connected via a linker.3 The binding and release of peptide
substrates in the SBD are coupled to ATP hydrolysis, which
causes conformational changes in the protein.4 The ATP-bound
state has low substrate affinity, and the ADP-bound state has high
substrate affinity. Cochaperone proteins such as J domains stimu-
late ATP hydrolysis,5 while nucleotide exchange factors such as
Bag-16 enhance ATPase activity by facilitating ADP release.
Glucose regulated protein 78 (Grp78), also known as BiP or
HSP5A, is the Hsp70 isoform located predominantly in the
endoplasmic reticulum (ER),7 while Hsp70 and Hsc70 isoforms
are located in the cytosol. As a key component of the ER chaperon-
ing function, Grp78 ensures proper protein folding, preventing
aggregation and targeting misfolded proteins for degradation
within the ER as well as binding calcium and regulating ER stress
signaling.8
Grp78 has been shown to play a role in tumor proliferation,
survival, and angiogenesis.9,10 Grp78 overexpression in tumor
cells appears necessary to survive oncogenic stress. Elevated
glucose metabolism leads to glucose starvation, low pH, and
severe hypoxia, which are conditions under which cancer cells
must survive, and are all factors that induce ER stress and activation
of the Grp78 promoter. Grp78 has also been found on the cell
membrane of cancer cells,11,12 where it has been shown to interact
with proteins involved in oncogenic signaling.13,14 Grp78 over-
expression also provides resistance to chemotherapeutic agents
in multiple tumor cell lines.15,16 Therefore, inhibiting Grp78
Known inhibitors targeting Grp78 activity or its induction are
derived from natural products. Genistein and (ꢀ)-epigallocatechin
gallate (EGCG) inhibit the pharmacological activity of Grp78,
while the bacterial AB5 subtilase cytotoxin specifically cleaves
Grp78 at a single amino acid.17 These agents were able to
selectively inhibit the growth of cancer cells or sensitize them
to cytotoxic chemotherapy. Genistein and EGCG act on many
cellular pathways and components in addition to Grp78. The
macrocycle versipelostatin (VST) inhibits the transcriptional
activation of the unfolded protein response (UPR) target genes
Grp78 and Grp94 and inhibited MNK047 tumor xenograft
growth.18 Grp78 can also be found in the cell membranes of
cancer cells, and recent therapeutic approaches have targeted the
membrane-bound SBD. A prodrug that contains a cyclic 13-mer
peptide (Pep42) was shown to bind to Grp78 and was inter-
nalized into the cell.19,20 Antibodies that bind to the SBD
inhibited cell proliferation and induced apoptosis in prostate
cancer cells and melanoma cells.21
Interest in Grp78 as an oncology target, the lack of structural
data on inhibitors binding to the ATPase domain, and the
unexplored questions of isomer selectivity prompted the current
study. In-house studies comparing the effect of siRNA knock-
down of Hsp70, Hsc70, and Grp78 corroborate the role of Grp78
in cell proliferation. We recently reported potent and novel
nucleoside inhibitors for the ATPase domain of the Hsp70/
Hsc70 isoforms. Compounds from our Hsp70 program22,23
Received: December 22, 2010
Published: April 28, 2011
r
2011 American Chemical Society
4034
dx.doi.org/10.1021/jm101625x J. Med. Chem. 2011, 54, 4034–4041
|

Journal of Medicinal Chemistry
ARTICLE
Scheme 1a
Table 1. Binding Data for Inhibitors with Grp78 and Hsp70
a Reagents and conditions: (a) MeNH2, EtOH, 130 °C, microwave,
91%; (b) Cs2CO3, RCH2Cl, DMF, 40ꢀ50%; (c) CF3CO3H, 18ꢀ25%.
Scheme 2a
a Reagents and conditions: (a) 1-(isoquinolin-3-yl)methanamine,
EtOH, 170 °C, microwave, PS-benzaldehyde, 25%.
a All IC50, KD, and GI50 are the mean of at least two determinations.
b Value at the detection limit in the FP assay.
enabled us to study Grp78 inhibition and isomer selectivity.
Inhibitor binding was studied by surface plasmon resonance
(SPR) and isothermal titration calorimetry (ITC). We report the
first X-ray crystallographic structures of the ATPase domain of
apo Grp78 and in complex with adenosine 50-triphosphate
(ATP), 50-adenylyl β,γ-imidodiphosphate (ADPnP), and a
nucleoside inhibitor. The modest selectivity seen in the binding
data for Grp78 and Hsp70 is discussed in the context of structure.
synthesized as shown in Scheme 1. Treatment of 8-bromo-20-30-
O-(1-methylethylidine)adenosine 1526 with methylamine affor-
ded the 8-methylamino derivative 16.27 Reaction of compound
16 with a range of primary alkyl chlorides in the presence of
cesium carbonate afforded intermediates 17aꢀd and 18aꢀd as
inseparable mixtures. Further treatment of this mixture with trifluoro-
acetic acid resulted in removal of the acetonide protecting group
and decomposition of the undesired regioisomers 18aꢀd, enabling
isolation of 4ꢀ7 in 18ꢀ25% yield. Compound 10 was synthesized
using existing methodology22,28 by treatment of 8-bromoadenosine
19 with 1-(isoquinolin-3-yl)methanamine (Scheme 2).
’ RESULTS
Inhibitor Synthesis. The syntheses of 2,24 3,25 8, 9, and
11ꢀ1422 have been previously reported. Compounds 4ꢀ7 were
4035
dx.doi.org/10.1021/jm101625x |J. Med. Chem. 2011, 54, 4034–4041

Journal of Medicinal Chemistry
ARTICLE
Table 2. Thermodynamic Binding Data and Cell Assay
Results
a All ITC and GI50 are the mean of at least two determinations unless
otherwise stated. b ITC KD = 1/KA. c Reference 22. d n = 1 e Low GI50 due
to predominantly non-HSP70 mechanism.
Figure 1. X-ray crystallography structures of Grp78/Apo (cyan,
3LDN), Grp78/ADPnP (green, 3LDO), and Grp78/ATP (3LDL) are
shown superimposed with Grp78/ADP31 (blue, 3IUC). Detail of the
binding site shows residue positions near ligands.
Inhibitor Binding Studies. Binding of compounds to Grp78
(Table 1) was assessed using SPR to measure the dissociation
constant (KD) and off-rates (kd). Adenosine 50-diphosphate
(ADP, 1) had the slowest off rate (kd = 0.005 sꢀ1) but not the
smallest KD (2.17 μM). Compounds 2 and 3 contain an 8-amino
and 8-methylamino substituent on the adenine core and have
similar KD of 6.80 and 7.46 μM, indicating that the additional
methyl has little effect on binding affinity. When binding to the
Hsp70 isoform, the 8-amino group in 2 resulted in 60-fold
improvement over the hydrogen in adenosine, likely due to an
internal hydrogen bond.22
rank ordering for compounds 2ꢀ5 are similar in Grp78 and
Hsp70. However, changing the phenyl in 4 to more polar
substituents (pyridyl 6, primary carboxamide 7) lowers binding
in Grp78 but has little effect on Hsp70 binding. Binding of 7 is
12-fold lower than binding of 4 in Grp78 but only change by
2-fold in Hsp70. The addition of aromatic subsituents in R1
(8ꢀ11), compared to 2 and 3, increases potency more signifi-
cantly in Grp78 than Hsp70.
Compounds 4ꢀ7 explore the effect of substituents at the R2
position when R1 is 8-methylamino. Changing the phenyl in 4
(KD = 0.99 μM) to an amide in 7 (KD = 12.36 μM) resulted in a
12-fold loss of binding affinity in Grp78. Compounds 8ꢀ10
explore aromatics substituents in R1, keeping R2 as hydrogen.
The quinoline in 9 provided affinity gains of 4-fold and 10-fold
compared to aromatics in 8 and 10. Aromatic substituents at R1
in 8ꢀ10 result in slower off-rates (kd = 0.2ꢀ0.5 sꢀ1) than those
observed for compounds 4ꢀ7 (kd > 1 sꢀ1), even for compounds
with similar KD. Changing the hydrogen in 8 to a methyl (11)
improved affinity by 4-fold. Compounds 12ꢀ14 have aryl
substituents in R1 and R2. Comparing 12 and 13 reveals that
the addition of a nitrile increases affinity by 10-fold.
To allow relative comparisons between Grp78 and Hsp70
affinities, Ki values for Hsp70 were calculated from the FP IC50
using equations assuming a tight inhibitor and free concentra-
tions.29,30 For the more potent inhibitors, SPR was used to
measure KD, since the Hsp70 FP assay used a protein concentra-
tion of 400 nM, which means that the assay response is limited to
compounds with affinities greater than 300 nM. The values and
Binding was confirmed for several compounds by ITC from
which the dissociation constant and enthalpic and entropic
contributions were obtained (KD, ΔH, ΔS) as reported in
Table 2. KD values obtained with ITC were usually within 2-fold
of those obtained with SPR. The quinoline (9 and 14) provides
an enthalpic gain over the dichloro substituent (8 and 13), which
is balanced by a decrease in entropy loss resulting in only small
improvements in overall KD. Despite the quinoline's higher
affinity for the protein shown by quinolines 9 and 14, they did
not inhibit cell growth like the less potent dichloro compounds.
Structure of Grp78 Bound to ADPnP and ATP. Structures of
the crystallized Grp78 ATPase domain (residues 24ꢀ407) in the
apo form and with soaked ligands were solved by X-ray crystal-
lography. Figure 1 illustrates that there are no large conforma-
tional changes upon binding of ADPnP and ATP, which caused
the CR backbone RMSDs to change by only 0.67 and 0.54 Å,
respectively. The main change is the shift observed in domain
IIB. Comparison with the cocrystal of Grp78/ADP solved by
Wisniewska and co-workers (3IUC), which crystallized with a
different space group and packing,31 shows a very similar binding
4036
dx.doi.org/10.1021/jm101625x |J. Med. Chem. 2011, 54, 4034–4041

Journal of Medicinal Chemistry
ARTICLE
Figure 2. (A) Conserved residues among Hsp70 isoforms (Grp78, Hsp70, Hsc70, Hsp70B0, Hsp701-Hom, Hsp70-2) are shown in green on the ribbon
representation of the Grp78/ADPnP structure; nonconserved residues are in red. (B) Within 4.5 Å around ADPnP, only one residue differs between
Grp78 and Hsp70 in the binding site. (C) Overlay of X-ray crystal structures of 10 bound to Grp78 and Hsc70/Bag reveals similar binding mode for
inhibitor. (D) Crystal structure of Hsc70/Bag-1 with 7 reveals that the amide π-stacks with Tyr15 and interacts with Thr37 through a water mediated
H-bond. Overlay with apo Grp78 shows that the polar amide would be too close to the hydrophobic Ile61 in Grp78.
site. The CR backbone rmsd values of 0.74 and 0.95 Å between
Grp78/ADP and our Grp78 and Grp78/ATP structures are
mostly due to surface residues.
as a surrogate for Hsp70, since they were more amenable to
crystallography and there is high homology between Hsp70 and
Hsc70. Hsc70 was crystallized in complex with its cochaperone
Bag-1, which opens up the ATPase domain by separating
domains IB and IIB. The presence of Bag-1 results in a larger
binding cavity reflected in the increased distance between the
residues in domain IB (Tyr and Ile) and residues in domain IIB
which bind the adenine. Several attempts to soak in the potent
diaryl inhibitors into the closed forms of Hsp70 and Grp78 failed.
A homology model of an open conformation of Grp78, built
using the Hsc70/Bag-1/14 structure (3fzm)22 as template,
suggests that 14 could bind in a similar mode as it does in
Hsc70/Bag-1 (Figure S3 in Supporting Information).
Figure 2D illustrates the location of the amide substituent of 7
bound to Hsc70/Bag-1. In Hsc70, the amide π-stacks with Tyr15
and interacts with Thr37 via bridging water molecules. Super-
imposing the complex of Hsc70/Bag-1/7 with a structure of
Grp78 shows that the polar amide in 7 would be in proximity to
the hydrophobic residue Ile61, which should be detrimental to
potency.
Comparison of Grp78, Hsp70, and Hsc70 Structures.
Grp78 is the least conserved isoform within the Hsp70 family
members. The closest homologue of Grp78 based on ATPase
sequence is Hsc70 with 70% residue identity as analyzed with
BLAST,32 while the residue identity between Hsc70 and Hsp70
is 89%. Sequence alignments can be found in Supporting Infor-
mation. Figure 2A highlights the residues that differ between
Grp78 ATPase and other Hsp70 isoforms. The ATPase struc-
tures31 of these isoforms showed that they have the same
secondary fold and a highly conserved binding site, with most
differences in residues located at the protein surface.
Comparison of the X-ray crystallographic structures of Grp78
and Hsp70 (Figure 2B) showed that there is one residue that is
different within a radius of 4.5 Å around ADPnP. Grp78 has an
isoleucine (Ile61) where Hsp70 has a threonine (Thr37). The
remaining binding site residues are conserved and have similar
conformations except for the solvent exposed Arg297, Arg367,
and Glu293 which occupy slightly different positions in some of
the structures.
’ DISCUSSION AND CONCLUSIONS
Structures of Inhibitors Bound to Grp78 and Hsc70/Bag-1.
Superimposition of the X-ray crystallography structures of 10
bound to Grp78 and Hsc70/Bag-1 in Figure 2C shows a similar
binding mode in Grp78 and Hsc70/Bag-1. While 10 is not one of
the more potent compounds, it was one for which we had
obtained crystals in both proteins. Crystals of Hsc70 were used
We have studied the inhibition of Grp78 due to its potential
interest as an oncology target. In-house studies comparing the
effect of siRNA knockdown of Hsp70, Hsc70, and Grp78
corroborate the role of Grp78 in cell proliferation. siRNA
knockdown of Grp78 had a greater effect on tumor cell growth
than knockdown of the cytosolic Hsp70 or Hsc70 in a panel of 10
4037
dx.doi.org/10.1021/jm101625x |J. Med. Chem. 2011, 54, 4034–4041

Journal of Medicinal Chemistry
ARTICLE
human cancer cells (Figure S2 in Supporting Information). Bind-
ing of compounds to Grp78 was investigated by SPR, ITC, and
X-ray crystallography. SPR and ITC produced highly correlated
KD, providing additional information on the off-rates and ther-
modynamic energy terms. The slow off-rate of ADP (1) (kd =
0.005 sꢀ1) is consistent with the need for nuclear exchange
cofactor proteins to help release ADP from the binding site of
Hsp70 proteins. The nucleoside inhibitors tested in ITC showed
a high enthalpic contribution which was counterbalanced by
disfavorable entropic contribution. The quinoline containing
ligands 9 and 14 were slightly more potent, had better enthalpic
contributions, and slower off-rates than the dichlorophenyl
inhibitors 8 and 13. However, previously published data obtained
for HCT116 cells indicated that only the dichloro compounds
were able to stop cell growth in cell-based assays,22 likely due
to physicochemical properties affecting the ability of the com-
pounds to reach the target protein.
The most potent compound with cellular activity, VER-
155008 (13, HCT116 GI50 = 5 μM), has similar potencies in
both Grp78 and Hsp70, thus this compound is likely a non-
selective pan-inhibitor for similar protein isoforms of the Hsp70
family. Crystal structures of Hsp70 and Grp78 reveal that while
most residues are conserved within the binding site, one residue
differs between Grp78 and Hsp70 and has the potential to be
used for designing in selectivity between Grp78 and the other
Hsp70 isoforms. Selectivity between Hsp70 and Grp78 was
observed for two compounds (6, 7) that contained polar subs-
tituents in R2. The change from a polar Thr37 in Hsp70 to
nonpolar Ile61 in Grp78 creates a more hydrophobic binding site
in Grp78 which could explain why polar substitutents at the R2
position were more detrimental to affinity in Grp78 than Hsp70.
The increased hydrophobicity in Grp78 could be exploited in
ligand design to achieve further potency; our library of nucleo-
side inhibitors was initially designed for Hsp70 and contained
few hydrophobic and saturated substituents in R2. Nevertheless,
the polarity, size and flexibility of the overall ATP binding pocket
still pose a challenge to achieve high affinity binding by small
druglike molecules.
LCꢀMS was performed using a Hewlett-Packard 1100 series instrument
linked to a quadrupole detector. The column was a Phenomenex Luna
3 μm C18(2), 30 mm ꢁ 4.6 mm i.d. Buffer A was prepared by dissolving
1.93 g of ammonium acetate in 2.5 L of HPLC grade water and adding
2 mL of formic acid. Buffer B was prepared by adding 132 mL of buffer A
to 2.5 L of HPLC grade acetonitrile and adding 2 mL of formic acid. The
elution gradient was buffer Aꢀbuffer B (95:5 to 5:95) over 3.75 min.
Flow rate was 2.0 mL minꢀ1. Retention times (tR) are reported in
minutes. Ionization is positive, unless otherwise stated. Nuclear mag-
netic resonance (NMR) analysis was performed with a Bruker DPX400
spectrometer, and proton NMR spectra were measured at 400 MHz.
The spectral reference was the known chemical shift of the solvent.
Proton NMR data are reported as follows: chemical shift (δ) in ppm,
followed by the integration, the multiplicity (s = singlet, d = doublet, t =
triplet, q = quartet, p = pentet, m = multiplet, dd = doublet of doublets,
and br = broad), and the coupling constant rounded to the nearest
0.1 Hz.
Compound 4 was purified by preparative HPLC at pH 4, performed
on a Waters FractionLynx MS autopurification system, with a Gemini
5 μm C18(2), 100 mm ꢁ 20 mm i.d. column from Phenomenex,
running at a flow rate of 20 mL minꢀ1 with UV diode array detection
(210ꢀ400 nm) and mass-directed collection. Solvent A was 10 mM
ammonium acetate in HPLC grade water þ 0.08% v/v formic acid.
Solvent B was 95% v/v HPLC grade acetonitrile þ 5% v/v solvent A þ
0.08% v/v formic acid. The gradient was (time, % solvent B) 0 min, 5%;
1 min, 8%; 7 min, 28%; 7.5 min, 95%; 9.5 min, 95%; 10 min, 5%. The
mass spectrometer was a Waters Micromass ZQ2000 instrument,
operating in positive or negative ion electrospray ionization modes,
with a molecular weight scan range of 150ꢀ1000. IUPAC chemical
names were generated using AutoNom Standard.
Compounds 1 (Sigma-Aldrich Ltd.) and 2 (Toronto Research Chem-
icals) are commercially available, and 3,25 8, 9, and 11ꢀ14,22 and 1526
have been previously described. Compounds 4ꢀ7 were prepared as
shown in Scheme 1; 10 was prepared as shown in Scheme 2.
(2R,3R,4S,5R)-2-(6-Amino-8-methylaminopurin-9-yl)-5-
benzyloxymethyltetrahydrofuran-3,4-diol (4). A solution of
17a and 18a (0.044 g, 0.10 mmol) in trifluoroacetic acid (2 mL) was
stirred at room temperature overnight. The solvent was then removed in
vacuo followed by repeat coevaporation with methanol. The resulting
crude product was purified by preparative LCꢀMS at pH 4, which
afforded a clear glass, identified as 4 (0.007 g, 0.019 mmol, 18%).
LCꢀMS: m/z 387 [M þ H]þ; tR = 1.64 min; total run time 3.75 min. 1H
NMR: δH (DMSO-d6) 7.89 (1 H, s, Ar-H), 7.35 (5 H, m, Ph), 6.50 (2 H,
s, NH2), 6.41 (1 H, q, J 4.6 Hz, MeNH), 5.81 (1 H, d, J 6.2 Hz, 2-H), 5.35
(1 H, br s, 3-OH), 5.25 (1 H, br s, 4-OH), 4.76 (1 H, t, J 5.9 Hz, 3-H),
4.57 (2 H, s, CH2Ph), 4.23 (1 H, dd, J 5.4 and 4.0 Hz, 4-H), 4.01 (1 H, q,
J 3.5 Hz, 5-H), 3.75 (1 H, dd, J 10.6 and 3.0 Hz, 5-CHH), 3.63 (1 H, dd,
J 10.7 and 4.0 Hz, 5-CHH), 2.66 (3 H, d, J 4.6 Hz, MeNH).
Using a combination of surface plasmon resonance and
isothermal titration calorimetry, we have identified submicromo-
lar inhibitors of Grp78 that bind to the ATPase domain of this
molecular chaperone. X-ray crystallographic studies of a nucleo-
side inhibitor bound to Grp78 revealed key structural differences
between Grp78 and the closely homologous proteins Hsc70 and
Hsp70. This has allowed us to rationalize the modest selectivity
for Grp78 observed in certain nucleoside analogues. These
molecules, which to our knowledge are the first inhibitors known
to bind to the ATPase domain of Grp78, may serve as useful
chemical tools to further probe the biology of Grp78.
(2R,3R,4S,5R)-2-(6-Amino-8-methylaminopurin-9-yl)-5-
cyclohexylmethoxymethyltetrahydrofuran-3,4-diol (5).
Synthesis was as described for 4. LCꢀMS: m/z 393 [M þ H]þ; tR =
1.84 min; total run time 3.75 min. 1H NMR: δH (DMSO-d6) 7.91 (1 H,
s, ArH), 6.48 (2 H, s, NH2), 6.43 (1 H, q, J 4.4 Hz, MeNH), 5.79 (1 H, d,
J 6.0 Hz, 2-H), 5.30 (1 H, br s, 3-OH), 5.15 (1 H, br s, 4-OH), 4.76 (1 H,
t, J 5.7 Hz, 3-H), 4.19 (1 H, t, J 4.4 Hz, 4-H), 3.96 (1 H, q, J 3.7 Hz, 5-H),
3.66 (1 H, dd, J 10.8 and 2.9 Hz, 5-CHHO), 3.51 (1H, dd, J 10.8 and 4.13
Hz, 5-CHHO), 3.31 (1 H, dd, J 9.6 and 6.3 Hz, OCHHcycloalkyl), 3.20
(1 H, dd, J 9.5 and 7.2 Hz, OCHHcycloalkyl), 2.92 (3 H, d, J 4.6 Hz,
MeNH), 1.50ꢀ1.71 (6 H, m, cycloalkyl), 1.10ꢀ1.20 (3 H, m, cycloal-
kyl), 0.82ꢀ0.92 (2 H, m, cycloalkyl).
’ EXPERIMENTAL SECTION
General Methods. All commercial reagents were used without
further purification. Anhydrous solvents were obtained from commer-
cial sources and used without further drying. Flash chromatography was
performed with prepacked silica gel cartridges (IST Flash II, 54 Å,
Biotage Ltd., Hengoed, U.K.). Thin layer chromatography was con-
ducted with 5 cm ꢁ 10 cm plates coated with Merck Type 60 F254 silica
gel. Microwave heating was performed with a Biotage Initiator 2.0
instrument.
(2R,3R,4S,5R)-2-(6-Amino-8-methylaminopurin-9-yl)-5-
(pyridin-3-ylmethoxymethyl)tetrahydrofuran-3,4-diol (6).
Synthesis was as described for 4. LCꢀMS: m/z 388 [M þ H]þ; tR = 1.19
min; total run time 3.75 min. 1H NMR: δH (DMSO-d6) 8.56 (1 H, d,
Compound purity was assigned by liquid chromatographyꢀmass
spectrometry (LCꢀMS) at pH 4; all compounds had g95% purity.
4038
dx.doi.org/10.1021/jm101625x |J. Med. Chem. 2011, 54, 4034–4041

Journal of Medicinal Chemistry
ARTICLE
J 1.7 Hz, pyridyl 2-H), 8.52 (1 H, dd, J 4.8 and 1.6 Hz, pyridyl 6-H), 7.90
(1 H, s, ArH), 7.70 (1 H, dt, J 7.9, 1.7, and 1.7 Hz, pyridyl 5-H), 7.38 (1
H, dd, 7.9 and 4.8 Hz, pyridyl 4-H), 6.47 (2 H, s, NH2), 6.45 (1 H, q,
J 4.6 Hz, MeNH), 5.79 (1 H, d, J 5.9 Hz, 2-H), 5.34 (1 H, br s, 3-OH),
5.19 (1 H, br s, 4-OH), 4.81 (1 H, t, J 5.6 Hz, 3-H), 4.60 (2 H, s,
OCH2Ar), 4.26 (1 H, t, J 4.5 Hz, 4-H), 4.01 (1 H, q, 4.1 Hz, 5-H), 3.78
(1 H, dd, J 10.7 and 3.1 Hz, 5-CHH), 3.65 (1 H, dd, J 10.7 and 4.3 Hz,
5-CHH), 2.69 (3 H, d, J 4.6 Hz, MeNH),
extract was dried (MgSO4) and the solvent removed in vacuo. The
resultant crude product was purified by silica gel (10 g) flash chroma-
tography, eluting with dichloromethaneꢀmethanol (98:2 to 95:55
gradient) to afford the desired products 17a and 18b as a cream glass
(0.044 g, 0.10 mmol, 42%). LCꢀMS: m/z 427 [M þ H]þ; tR = 2.04 and
2.13 min; total run time 3.75 min.
Protein Expression and Purification. Expression trials of the
GST tagged, ATPase plasmid pRT860 was initiated in the E. coli BL21
DE3 strain. Expression was achieved overnight after induction with
0.6 mM IPTG and temperature reduction to 18 °C. The cells were
subsequently harvested after 18 h of growth and yielded approximately
6 mg/mL protein. The cell paste was frozen immediately for later use or
lysed using an Emulsiflex cell disrupter. Lysed cells were clarified for 1 h
at 30 000 rpm. Cleared lysates were loaded onto a 5 mL glutathione S-
transferase column. Pooled peak fractions were eluted with 50 mM Tris-
HCl, pH 8, 100 mM NaCl, 1 mM DTT, and 10 mM glutathione. To
achieve crystallizable material, the tagged protein was cleaved. A final
polishing step using a HiLoad 16-60 SD200pg column eluted with
20 mM Tris-HCl, pH 7.5, 200 mM NaCl, and 1 mM DTT gave very pure
material. The Grp78 protein included the residues 24ꢀ407. The Hsc70
protein included residues 5ꢀ381. Hsc70 and Bag-1 were prepared as
previously described.22
Crystallization and Structure Determination. Very thin
plates of apo Grp78 crystals were obtained from purified protein mixed
with 0.1 M Tris buffer, pH 8.4ꢀ8.6, 20ꢀ25% Peg3350, 0.1ꢀ0.2 M Na, K
tartrate, in hanging drop vapor diffusion experiments, carried out at 293
K. Because of merohedral twinning, single crystals had to be selected by
initial X-ray diffraction examination. Crystallization of Hsc70/Bag-1 has
been described in Sondermann et al.6 Diffraction data were acquired on
Rigaku RUH3 rotating anode X-ray source equipped with Raxis IVþþ
image plate detector (Hsc70/Bag-1/7) or at ESRF synchrotron beam-
line ID29 (Grp78 apo and complexes and ID23-1 (Hsc70/Bag-1/10).
Data were processed with Rigaku/MSC's CrystalClear or HKL200034
software. Structures were solved by molecular replacement using the
Hsc70 3fzf22 structure as the starting model for Grp78 and different and/
or missing fragments/side chains built in, and the structure was refined.
Grp78 ATPase domain (residues 24-407) crystallized in the mono-
clinic P21 space group with two independent molecules in the unit cell.
Apo crystals of Grp78 were soaked in a DMSO solution of the ligands or
in an aqueous solution of ATP or ADPnP at pH 7.5. Ligands were fitted
into difference Fourier electron density maps and subsequently included
in refinement. Model corrections and structure refinement were itera-
tively carried out with COOT35 and Refmac36 (as implemented in the
CCP4i suite37). Structures of Hsc70/Bag-1 were solved as previously
described.22
2-[(2R,3S,4R,5R)-5-(6-Amino-8-methylaminopurin-9-yl)-
3,4-dihydroxytetrahydrofuran-2-ylmethoxy]acetamide (7).
Synthesis was as described for 4. LCꢀMS: m/z 354 [M þ H]þ; tR =
0.86 min; total run time 3.75 min. 1H NMR: δH (DMSO-d6) 7.90 (1 H,
s, ArH), 7.28 (1 H, br s, CONHH) 7.22 (1 H, br s, CONHH), 6.58 (1 H,
q, J 4.7 Hz, MeNH), 6.46 (2 H, s, NH2), 5.75 (1 H, d, J 6.2 Hz, 2-H), 5.28
(2 H, br s, 3- and 4-OH), 4.90 (1 H, t, J 5.9 Hz, 3-H), 4.27 (1 H, dd, J 5.6
and 3.9 Hz, 4-H), 3.99 (1 H, q, J 3.9 Hz, 5-H), 3.87 [2 H, d, J 1.5 Hz,
CH2C(O)NH2], 3.78 (1 H, dd, J 10.7 and 3.0 Hz, 5-CHH), 3.63 (1 H,
dd, J 10.7 and 4.4 Hz, 5-CHH), 2.89 (3 H, d, J 4.5 Hz, MeNH).
(2R,3R,4S,5R)-2-{6-Amino-8-[(quinolin-2-ylmethyl)amino]-
purin-9-yl}-5-hydroxymethyltetrahydrofuran-3,4-diol (10).
A suspension of (2R,3R,4S,5R)-2-(6-amino-8-bromopurin-9-yl)-5-hydro-
xymethyltetrahydrofuran-3,4-diol (8-bromoadenine-9-β-D-ribofuranoside)
(19) (0.50 g, 1.45 mmol) and 3,4-dichlorobenzylamine (1.93 mL, 14.4
mmol) in ethanol (15 mL) was heated with microwaves in a sealed tube at
170 °C for 30 min.28 The solvent was removed in vacuo. The resultant
residue was dissolved in methanolꢀdichloromethane (1:4, 75 mL), and
then PS-benzaldehyde (13.0 g, 13.0 mmol) was added. The suspension was
stirred at room temperature overnight. The PS-benzaldehyde was removed
by filtration and the solvent removed in vacuo from the filtrate. The
resultant crude product was purified by silica gel (25 g) flash column
chromatography [methanolꢀdichloromethane (5ꢀ10% gradient)] to
afford the desired product 10 as a white solid (0.116 g, 0.26 mmol,
25%). LCꢀMS: m/z 424 [M þ H]þ; tR = 1.44 min; total run time 3.75
min. 1H NMR: δH (DMSO-d6) 8.32 (1 H, d, J 8.5 Hz, Ar-H), 7.93ꢀ7.99
(2 H, m, 2 ꢁ Ar-H), 7.91 (1 H, s, Ar-H), 7.75ꢀ7.77 (2 H, m, 2 ꢁ Ar-H ),
7.54ꢀ7.59 (2 H, m, NHMe and Ar-H), 6.53 (2 H, s, NH2), 6.00 (1 H, d,
J 7.4 Hz, 2-H), 5.87 (1 H, dd, J 6.2 and 4.1 Hz, 5-CH2OH), 5.36 (1 H, d,
J 6.7 Hz, 3-OH), 5.19 (1 H, d, J 4.0 Hz, 4-OH), 4.80ꢀ4.81 (3 H, m, 3-H
and CH2Ar), 4.14ꢀ4.17 (1 H, m, 4-H), 4.02 (1 H, m, 5-H), 3.62ꢀ3.70
(2 H, m, 5-CH2OH).
(3aR,4R,6R,6aR)-[6-(6-Amino-8-methylaminopurin-9-yl)-2,2-
dimethyltetrahydrofuro[3,4-d][1,3]dioxol-4-yl]methanol
(16). A suspension of 1526,33 (0.50 g, 1.3 mmol) in methylamine (33%
in ethanol, 5 mL) was heated in a sealed tube with microwaves at 130 °C
for 45 min. The solvent was then removed in vacuo. The residual crude
product was purified by silica gel (50 g) flash chromatography, eluting
with dichloromethaneꢀmethanol (95:5 to 90:10 gradient) to afford the
desired product 16 as a white foam (0.40 g, 1.19 mmol, 91%). LCꢀMS:
m/z 337 [M þ H]þ; tR = 1.46 min; total run time 3.75 min. 1H NMR:
δH (DMSO-d6) 7.91 (1 H, s, Ar-H), 6.96 (1 H, q, J 4.4 Hz, MeNH), 6.59
(2 H, s, NH2), 5.99 (1 H, d, J 3.6 Hz, 2-H), 5.49 (1 H, t, J 5.3 Hz,
CH2OH), 5.42 (1 H, m, 3-H), 4.97 (1 H, dd, J 2.8 and 3.5 Hz, 4-H), 4.15
(1 H, m, 5-H), 3.55 (2 H, m, CH2OH), 2.88 (3 H, d, J 4.6 Hz, MeNH),
1.54 (3 H, S, MeMeC), 1.30 (3 H, s, MeMeC).
Surface Plasmon Resonance (SPR). SPR measurements were
performed on BIAcore T100 instrument (BIAcore GE Healthcare), at
25 °C on series S NTA chips (certified) according to provider's proto-
cols with 10 mM HEPES, pH 7.4, 150 mM NaCl, 500 μM EDTA, 0.05%
Tween-20, and 1% DMSO as a running buffer. Histidine-tagged Grp78
was immobilized on the sensor surface; reference surfaces without
immobilized Ni2þ served as controls for nonspecific binding and refrac-
tive index changes. Concentrations of inhibitors (0ꢀ200 μM) were
typically injected over the sensor chip at 35 μL/min. Zero concentration
samples were used as blanks. The sensor surface was regenerated between
experiments by injections of 0.1 mg/mL trypsin and 50% DMSO. Data
processing was performed using BIAevaluation 2.1 software (BIAcore GE
Healthcare Bio-Sciences Corp.) by globally fitting the entire inhibitor
concentration series data set to the steady state affinity model.
9-{(3aR,4R,6R,6aR)-6-Benzyloxymethyl-2,2-dimethyltetra-
hydrofuro[3,4-d][1,3]dioxol-4-yl}-N8-methyl-9H-purine-
6,8-diamine (17a) and (3aR,4R,6R,6aR)-{6-[6-Amino-8-
(benzylmethylamino)purin-9-yl]-2,2-dimethyltetrahydrofuro-
[3,4-d][1,3]dioxol-4-yl}methanol (18a). Cesium carbonate
(0.156 g, 0.48 mmol) and benzyl chloride (0.033 mL, 0.28 mmol) were
added sequentially to a solution of 1627 (0.080 g, 0.24 mmol) in DMF
(5 mL). The mixture was stirred at room temperature for ∼72 h. The
solvent was removed in vacuo. Then the resulting residue was parti-
tioned between ethyl acetate (20 mL) and water (10 mL). The organic
Isothermal Titration Calorimetry. ITC measurements were
performed using an iTC200 instrument (Microcal, GE Healthcare),
with 11 μM protein at 25 °C, 10 mM HEPES, pH7.4, 150 mM NaCl,
500 μM EDTA, 5 mM monothioglycerol, 0.05% Tween-20, and 1%
DMSO. All data were fitted to a one site model using the provided software.
4039
dx.doi.org/10.1021/jm101625x |J. Med. Chem. 2011, 54, 4034–4041

Journal of Medicinal Chemistry
ARTICLE
’ ASSOCIATED CONTENT
(9) Dong, D.; Ni, M.; Li, J.; Xiong, S.; Ye, W.; Virrey, J. J.; Mao, C.;
Ye, R.; Wang, M.; Pen, L.; Dubeau, L.; Groshen, S.; Hofman, F. M.; Lee,
A. S. Critical role of the stress chaperone GRP78/BiP in tumor
proliferation, survival, and tumor angiogenesis in transgene-induced
mammary tumor development. Cancer Res. 2008, 68, 498–505.
(10) Li, J.; Lee, A. S. Stress induction of GRP78/BiP and its role in
cancer. Curr. Mol. Med. 2006, 6, 45–54.
(11) Misra, U. K.; Gonzalez-Gronow, M.; Gawdi, G.; Pizzo, S. V. The
role of MTJ-1 in cell surface translocation of GRP78, a receptor for alpha
2-macroglobulin-dependent signaling. J. Immunol. 2005, 174, 2092–2097.
(12) Delpino, A.; Castelli, M. The 78 kDa glucose-regulated protein
(GRP78/BIP) is expressed on the cell membrane, is released into cell
culture medium and is also present in human peripheral circulation.
Biosci. Rep. 2002, 22, 407–420.
S
Supporting Information.
Crystallographic and refine-
b
ment data, sample ITC data, siRNA data and methods, model of
Grp78/14, and sequence alignment of protein isoforms. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Accession Codes
†Crystal structures of Grp78 (3ldn), Grp78/ATP (3ldl), Grp78/
ADPnP (3ldo), Grp78/10 (3ldp), Hsc70/Bag-1/10 (3ldq), and
Hsc70/Bag-1/7 (3m3z) have been deposited in the RCSB Protein
Data Bank.
(13) Kelber, J. A.; Panopoulos, A. D.; Shani, G.; Booker, E. C.;
Belmonte, J. C.; Vale, W. W.; Gray, P. C. Blockade of Cripto binding to
cell surface GRP78 inhibits oncogenic Cripto signaling via MAPK/PI3K
and Smad2/3 pathways. Oncogene 2009, 28, 2324–2336.
(14) Misra, U. K.; Deedwania, R.; Pizzo, S. V. Activation and cross-
talk between Akt, NF-kappaB, and unfolded protein response signaling
in 1-LN prostate cancer cells consequent to ligation of cell surface-
associated GRP78. J. Biol. Chem. 2006, 281, 13694–13707.
(15) Ranganathan, A. C.; Zhang, L.; Adam, A. P.; Aguirre-Ghiso, J. A.
Functional coupling of p38-induced up-regulation of BiP and activation of
RNA-dependent protein kinase-like endoplasmic reticulum kinase to drug
resistance of dormant carcinoma cells. Cancer Res. 2006, 66, 1702–1711.
(16) Reddy, R. K.; Mao, C.; Baumeister, P.; Austin, R. C.; Kaufman,
R. J.; Lee, A. S. Endoplasmic reticulum chaperone protein GRP78
protects cells from apoptosis induced by topoisomerase inhibitors: role
of ATP binding site in suppression of caspase-7 activation. J. Biol. Chem.
2003, 278, 20915–20924.
(17) Lee, A. S. GRP78 induction in cancer: therapeutic and prog-
nostic implications. Cancer Res. 2007, 67, 3496–3499.
(18) Park, H. R.; Furihata, K.; Hayakawa, Y.; Shin-Ya, K. Versipe-
lostatin, a novel GRP78/Bip molecular chaperone down-regulator of
microbial origin. Tetrahedron Lett. 2010, 43, 6941–6945.
(19) Yoneda, Y.; Steiniger, S. C.; Capkova, K.; Mee, J. M.; Liu, Y.;
Kaufmann, G. F.; Janda, K. D. A cell-penetrating peptidic GRP78 ligand
for tumor cell-specific prodrug therapy. Bioorg. Med. Chem. Lett. 2008,
18, 1632–1636.
(20) Kim, Y.; Lillo, A. M.; Steiniger, S. C.; Liu, Y.; Ballatore, C.;
Anichini, A.; Mortarini, R.; Kaufmann, G. F.; Zhou, B.; Felding-
Habermann, B.; Janda, K. D. Targeting heat shock proteins on cancer
cells: selection, characterization, and cell-penetrating properties of a
peptidic GRP78 ligand. Biochemistry 2006, 45, 9434–9444.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: þ44 (0)1223 895555. Fax: þ44 (0)1223 895556.
E-mail: a.macias@vernalis.com.
’ ACKNOWLEDGMENT
We thank Heather Simmonite for NMR spectroscopy and
Loic Le Strat for chromatographic support.
’ ABBREVIATIONS USED
ADP, adenosine 50-diphosphate; ADPnP, 50-adenylyl β,γ-imido-
diphosphate; ATP, adenosine 50-triphosphate; Bag-1, BCL-2-as-
sociated athanolgene 1; BiP, immunoglobulin heavy-binding
protein; CR, R carbon; DMSO, dimethylsulfoxide; EGCG, (ꢀ)-
epigallocatechin gallate; ER, endoplasmic reticulum; FP, fluores-
cence polarization; Grp78, 78 kDa glucose regulated protein;
ΔH, enthalpy change; ΔS, entropy change; KD, dissociative con-
stant; kd, off-rate; HCT116, human colon tumor 116; HEPES,
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; HSC70,
70 kDa heat shock cognate protein; HSP, heat shock protein;
HSP70, 70 kDa heat shock protein; ITC, isothermal titration
calorimetry; rmsd, root-mean-square deviation; SBD, substrate
binding domain; siRNA, short interfering ribonucleic acid; SPR, sur-
face plasmon resonance; Tris, (hydroxymethyl)aminomethane; UPR,
unfolded protein response; VST, versipelostatin
(21) Misra, U. K.; Mowery, Y.; Kaczowka, S.; Pizzo, S. V. Ligation of
cancer cell surface GRP78 with antibodies directed against its COOH-
terminal domain up-regulates p53 activity and promotes apoptosis. Mol.
Cancer Ther 2009, 8, 1350–1362.
’ REFERENCES
(1) Mayer, M. P.; Bukau, B. Hsp70 chaperones: cellular functions
and molecular mechanism. Cell. Mol. Life Sci. 2005, 62, 670–684.
(2) Bukau, B.; Weissman, J.; Horwich, A. Molecular chaperones and
protein quality control. Cell 2006, 125, 443–451.
(3) Jiang, J.; Prasad, K.; Lafer, E. M.; Sousa, R. Structural basis of
interdomain communication in the Hsc70 chaperone. Mol. Cell 2005,
20, 513–524.
(4) Vogel, M.; Bukau, B.; Mayer, M. P. Allosteric regulation of
Hsp70 chaperones by a proline switch. Mol. Cell 2006, 21, 359–367.
(5) Amor-Mahjoub, M.; Gomez-Vrielyunck, N.; Suppini, J. P.; Fouchaq,
B.; Benarouj, N.; Ladjimi, M. Involvement of the interdomain hydro-
phobic linker and the C-terminal helices in self-association of the
molecular chaperone HSC70. Arch. Inst. Pasteur Tunis 2006, 83, 53–62.
(6) Sondermann, H.; Scheufler, C.; Schneider, C.; Hohfeld, J.; Hartl,
F. U.; Moarefi, I. Structure of a Bag/Hsc70 complex: convergent
functional evolution of Hsp70 nucleotide exchange factors. Science
2001, 291, 1553–1557.
(22) Williamson, D. S.; Borgognoni, J.; Clay, A.; Daniels, Z.;
Dokurno, P.; Drysdale, M. J.; Foloppe, N.; Francis, G. L.; Graham,
C. J.; Howes, R.; Macias, A. T.; Murray, J. B.; Parsons, R.; Shaw, T.;
Surgenor, A. E.; Terry, L.; Wang, Y.; Wood, M.; Massey, A. J. Novel
adenosine-derived inhibitors of 70 kDa heat shock protein, discovered
through structure-based design. J. Med. Chem. 2009, 52, 1510–1513.
(23) Massey, A. J.; Williamson, D. S.; Browne, H.; Murray, J. B.;
Dokurno, P.; Shaw, T.; Macias, A. T.; Daniels, Z.; Geoffroy, S.; Dopson,
M.; Lavan, P.; Matassova, N.; Francis, G. L.; Graham, C. J.; Parsons, R.;
Wang, Y.; Padfield, A.; Comer, M.; Drysdale, M. J.; Wood, M. A novel,
small molecule inhibitor of Hsc70/Hsp70 potentiates Hsp90 inhibitor
induced apoptosis in HCT116 colon carcinoma cells. Cancer Chemother.
Pharmacol. 2010, 66, 535–545.
(24) Holmes, R. E.; Robins, R. K. Purine Nucleosides. IX. The
synthesis of 9-beta-D-ribofuranosyl uric acid and other related 8-sub-
stituted purine ribonucleosides. J. Am. Chem. Soc. 1965, 87, 1772–1776.
(25) Long, R. A.; Robinds, R. K.; Townsend, L. B. Purine nucleo-
sides. XV. The synthesis of 8-aminoand 8-substituted aminopurine
nucleosides. J. Org. Chem. 1967, 32, 2751–2756.
(7) Haas, I. G. BiP (GRP78), an essential hsp70 resident protein in
the endoplasmic reticulum. Experientia 1994, 50, 1012–1020.
(8) Lee, A. S. The ER chaperone and signaling regulator GRP78/BiP
as a monitor of endoplasmic reticulum stress. Methods 2005, 35, 373–381.
4040
dx.doi.org/10.1021/jm101625x |J. Med. Chem. 2011, 54, 4034–4041

Journal of Medicinal Chemistry
ARTICLE
(26) Ikehara, M.; Tada, H.; Muneyama, K.; Kaneko, M. Synthesis of
purine cyclonucleoside having an 8,20-O-anhydro linkage. J. Am. Chem.
Soc. 1966, 88, 3165–3167.
(27) Sasaki, T.; Minamoto, K.; Fujiki, Y. Systematic synthesis of
purine 8,50-imino and substituted imino cyclonucleosides. Chem. Lett.
1983, 7, 1017–1020.
(28) Chattopadhyaya, J. B.; Reese, C. B. Reaction between 8-bro-
moadenosine and amines. Chemistry of 8-hydrazinoadenosine. Synthesis
1977, 10, 725–726.
(29) Huang, X. Fluorescence polarization competition assay: the
range of resolvable inhibitor potency is limited by the affinity of the
fluorescent ligand. J. Biomol. Screening 2003, 8, 34–38.
(30) Cer, R. Z.; Mudunuri, U.; Stephens, R.; Lebeda, F. J. IC50-to-Ki:
a Web-based tool for converting IC50 to Ki values for inhibitors
of enzyme activity and ligand binding. Nucleic Acids Res. 2009, 37,
W441–W445.
(31) Wisniewska, M.;Karlberg, T.; Lehtio, L.; Johansson, I.; Kotenyova,
T.; Moche, M.; Schuler, H. Crystal structures of the ATPase domains of
four human Hsp70 isoforms: HSPA1L/Hsp70-hom, HSPA2/Hsp70-2,
HSPA6/Hsp70B0, and HSPA5/BiP/GRP78. PLoS One 2010, 5, No. e8625.
(32) Altschul, S. F.; Madden, T. L.; Schaffer, A. A.; Zhang, J.; Zhang,
Z.; Miller, W.; Lipman, D. J. Gapped BLAST and PSI-BLAST: a new
generation of protein database search programs. Nucleic Acids Res. 1997,
25, 3389–3402.
(33) Schmitt, L.; Tampe, R. ATP-lipids-protein anchor and energy
source in two dimensions. J. Am. Chem. Soc. 1996, 118, 5532–5543.
(34) Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data
collected in oscillation mode. Methods Enzymol. 1997, 276, 307–326.
(35) Emsley, P.; Cowtan, K. Coot: model-building tools for molec-
ular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004, 60,
2126–2132.
(36) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of
macromolecular structures by the maximum-likelihood method. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 1997, 53, 240–255.
(37) Potterton, E.; Briggs, P.; Turkenburg, M.; Dodson, E. A
graphical user interface to the CCP4 program suite. Acta Crystallogr.,
Sect. D: Biol. Crystallogr. 2003, 59, 1131–1137.
4041
dx.doi.org/10.1021/jm101625x |J. Med. Chem. 2011, 54, 4034–4041