1,3-disubstituted-4-aminopyrazolo [3, 4-d] pyrimidines, a new class of potent inhibitors for phospholipase D

Article abstract of DOI:10.1111/cbdd.12319

Phospholipase D enzymes cleave lipid substrates to produce phosphatidic acid, an important precursor for many essential cellular molecules. Phospholipase D is a target to modulate cancer-cell invasiveness. This study reports synthesis of a new class of phospholipase D inhibitors based on 1,3-disubstituted-4-amino-pyrazolopyrimidine core structure. These molecules were synthesized and used to perform initial screening for the inhibition of purified bacterial phospholipase D, which is highly homologous to the human PLD₁. Initially tested with the bacterial phospholipase D enzyme, then confirmed with the recombinant human PLD₁ and PLD₂ enzymes, the molecules presented here exhibited inhibition of phospholipase D activity (IC₅₀) in the low-nanomolar to low-micromolar range with both monomeric substrate diC₄PC and phospholipid vesicles and micelles. The data strongly indicate that these inhibitory molecules directly block enzyme/vesicle substrate binding. Preliminary activity studies using recombinant human phospholipase Ds in in vivo cell assays measuring both transphosphatidylation and head-group cleavage indicate inhibition in the mid- to low-nanomolar range for these potent inhibitory novel molecules in a physiological environment. This study reports synthesis of a new class of PLD inhibitors based on 1,3-disubstituted-4-amino-pyrazolopyrimidine core structure. These molecules exhibited inhibition of human recombinant PLD activity (IC ₅₀) in the low-nanomolar to low-micromolar range with monomeric substrate diC4PC and phospholipid vesicles and micelles. Preliminary activity studies using recombinant human PLDs in in vivo cell assays measuring both transphosphatidylation and head-group cleavage indicates inhibition in the mid- to low-nanomolar range for these potent inhibitory novel molecules in a physiological environment.

Full text of DOI:10.1111/cbdd.12319

Chem Biol Drug Des 2014; 84: 270–281
Research Article
1,3-Disubstituted-4-Aminopyrazolo [3, 4-d]
Pyrimidines, a New Class of Potent Inhibitors for
Phospholipase D
Aditya Kulkarni1, Phong Quang2, Victoriana
Curry2, Renee Keyes2, Weihong Zhou1, Hyejin
oleoyl-phosphatidic acid; POPC, 1-palmitoyl-2-oleoyl-PC; Ptd-
But, phosphatidyl butanol; SUV, small unilamellar vesicles.
1
1
1,
ꢀ
€ €
Cho , Jonathan Baffoe , Bela Torok * and
Received 2 August 2013, revised 19 December 2013 and
accepted for publication 4 March 2014
Kimberly Stieglitz2,*
1Department of Chemistry, University of Massachusetts
Boston, 100 Morrissey Blvd, Boston, MA 02125, USA
2STEM Biotechnology Division, Roxbury Community
College, Roxbury, MA 02120, USA
Phospholipase D (PLD) is an enzymatically active protein
that catalyzes the cleavage of phosphatidylcholine at the
ester linkage–releasing choline and phosphatidic acid (PA)
(1). Phosphatidic acid is a second lipid messenger located
at the intersection of several essential signaling and meta-
bolic pathways (2). Phosphatidic acid is involved in the
regulation of essential cellular functions and also mediates
enhancement of cell migration. Both increased PLD activity
and PLD expression have been observed in a variety of
human cancer tissues (3–5). While PLD promotes the abil-
ity of cells to initiate defence mechanisms, the inhibition of
PLD results in a diminished ability of cells to adhere.
Therefore, alleviating PLD activity has recently emerged as
a new drug target for human cancer-cell invasiveness (6).
Recent studies have also associated human PLD1 with
amyloid precursor protein in Alzheimer’s disease (7,8).
Hence, a therapeutic strategy, which involves specific inhi-
bition of PLD by small molecules, is highly desirable for
the treatment of cancer and Alzheimer’s disease.
ꢀ
€ €
*Corresponding authors: Bela Torok,
bela.torok@umb.edu; and Kimberly Stieglitz,
kastieglitz@rcc.mass.edu
Phospholipase D enzymes cleave lipid substrates to
produce phosphatidic acid, an important precursor for
many essential cellular molecules. Phospholipase D
is
This study reports synthesis of a new class of phos-
pholipase inhibitors based on 1,3-disubstituted-
a target to modulate cancer-cell invasiveness.
D
4-amino-pyrazolopyrimidine core structure. These
molecules were synthesized and used to perform initial
screening for the inhibition of purified bacterial phos-
pholipase D, which is highly homologous to the human
PLD1. Initially tested with the bacterial phospholipase
D
enzyme, then confirmed with the recombinant
human PLD1 and PLD2 enzymes, the molecules pre-
sented here exhibited inhibition of phospholipase D
activity (IC50) in the low-nanomolar to low-micromolar
range with both monomeric substrate diC4PC and
phospholipid vesicles and micelles. The data strongly
indicate that these inhibitory molecules directly
block enzyme/vesicle substrate binding. Preliminary
activity studies using recombinant human phospholi-
pase Ds in in vivo cell assays measuring both trans-
phosphatidylation and head-group cleavage indicate
inhibition in the mid- to low-nanomolar range for these
potent inhibitory novel molecules in a physiological
environment.
Mammalian PLD exists in the following two isoforms: PLD1
and PLD2. Both of these isoforms have 53% sequence
identity and show different modes of activation and func-
tional roles. Recently, extensive efforts have been focused
on selective inhibition of one isoform over the other.
Figure 1 shows the structures of successful dual- and
isoform-selective PLD inhibitors in previous in vitro and in
vivo quantitative cell-based assays. Isoform-selective PLD1
inhibitor 2 and PLD2 inhibitor 3 from Figure 1 are structural
modifications of dual PLD1/2 inhibitor halopemide, 1 (9,10).
In order to be PLD1-selective, the inhibitor requires a
(S)-configuration on the methyl-bearing carbon adjacent to
the amide linkage (9), whereas a PLD2-selective inhibitor
requires a spiro ring fused with the lactam (9,10). These
inhibitors were used with the bacterial PLD used in this
study for screening and preliminary kinetics, and the two
human PLD isoforms overexpressed to test novel inhibitors
reported here. The bacterial PLD used in preliminary
Key words: chemical biology, drug design, protease and
ligands (substrate/inhibitor)
Abbreviations: ARF, ADP-ribosylation factor; diC4PA, dibuty-
roylphosphatidic acid; diC4PC, dibutyroylphosphatidylcholine;
LUV, large unilamellar vesicles; MES, 2-(N-morpholo)ethane-
sulfonic acid; PAGE, polyacrylamide gel electrophoresis; PA,
phosphatidic acid; PC, phosphatidylcholine; PLD, phospholipase
D; PMA, phorbol 12-myristate 13-acetate; POPA, 1-palmitoyl-2-
270
ª 2014 John Wiley & Sons A/S. doi: 10.1111/cbdd.12319

Potent Inhibitors for Phospholipase D
Figure 1: Structures of successful dual and isoform selective PLD inhibitors.
testing of inhibitors is available to purchase from EnzoLife-
Sciences and has similar reactivity with the human PLD1.
nomic and commercially available starting materials in four
steps as shown in Scheme 1. A considerable diversity in
the basic core structure could be achieved by using vari-
ous hydrazines and acyl/aroyl chlorides.
This bacterial PLD, from Streptomyces species strain pmf
(PLD_sp) used in this study for preliminary screening of
newly synthesized molecules, exhibits a high specificity for
phosphatidylcholine (PC) cleavage to generate free choline
and PA. The advantage of using this bacterial PLD is that
the PLD_sp crystal structure has been solved, and the
active site contains the highly conserved HKD repeats
common within the PLD superfamily, including human and
other mammalian PLD’s (11). Therefore, when molecules
were found to be inhibitory in the in vivo and in vitro
assays with the purified enzyme, molecular docking and
viewing the 3D structure may aid in the ‘second genera-
tion’ of drugs design.
Materials and Methods
Synthesis of inhibitor candidates
All chemicals and solvents used for the synthesis of the
inhibitor candidates were purchased from Sigma-Aldrich
Corp., St. Louis, MO USA. All compounds prepared were
synthesized on the basis of a method described in litera-
ture (6,13,14) with minor alterations wherever necessary. A
variety of substituted derivatives was synthesized based
on the 4-aminopyrazolopyrimidine core structure. The mol-
ecules were synthesized in five steps starting from acyl/
aroyl chloride and malononitrile. In the first step, the acyl/
aroyl chloride was coupled with malononitrile with a simple
base-catalyzed addition/elimination reaction. In the second
step, enol ether was generated by the addition of dimethyl
sulfide. In step three, a pyrazole was synthesized by the
addition of a hydrazine. The first three steps were carried
out in a one pot manner to get pyrazoles in moderate
yields (30–60%). The intermediate was refluxed in formam-
ide to get the final product in high yields (step 4). The
identification and purity determination were carried out by
mass spectrometry and NMR spectroscopy (1H and 13C)
(step 5). Spectra of known compounds were in agreement
with the literature data (see Appendix S1).
Herein, we report highly potent inhibitors of PLD_sp and
human PLD1 and PLD2, which are based on an 4-amino-
pyrazolopyrimidine core structure (Scheme 1). These
enzymes were tested for the inhibition of PLD activity in
several in vitro biochemical assays using various forms of
the phospholipid substrate phosphatidylcholine, from
monomer substrate, micelles, mixed micelles, and small
unilamellar vesicles (SUV). Preliminary in vivo studies were
also carried out with various cell lines measuring in vivo
expression of human PLDs to test for the inhibition of
transphosphatidylation activity and PC head-group cleav-
age (PA production) from whole-cell lipid extraction. 4-
Aminopyrazolopyrimidines have earlier been used as tyro-
sine kinase inhibitors (12) and dual inhibitors of tyrosine
and phosphoinositide kinases (13) among other applica-
tions. The work presented here expands the application of
these molecules as potential potent inhibitors of enzymes
in PLD superfamily, which are emerging as targets for can-
cer research and other human diseases.
Preparation of commercially available PLD
inhibitors to test bacterial and human PLD
enzymes for specificity of inhibition
FIPI (4-Fluoro-N-(2-(4-(5-fluoro-1H-indol-1-yl)piperidin-1-yl)
ethyl)benzamide) was purchased from Sigma-Aldrich cata-
log number F5807. The following PLD inhibitors were pur-
chased from Avant Polar lipids: #857370 named
The inhibitors designed in this study lack a chiral center or
a spiro ring structure, which is attractive from the synthetic
point of view. The synthesis was accomplished from eco-
VU0155056
N-(2-{4-[2-oxo-2,3-dihydro-1H-benzo(d)imi-
Chem Biol Drug Des 2014; 84: 270–281
271

Kulkarni et al.
Scheme 1: Synthesis of PLD inhibitors.
dazol-1-yl]piperidin-1-yl}ethyl)-2-naphthamide,
#857371
In vitro kinetic assay using soluble substrate
DiC4PC for PLD inhibition studies
named
VU0359595(1R,2R)-N-([S]-1-{4-[5-bromo-2-oxo-
2,3-dihydro-1H-benzo(d)imidazol-1-yl]piperidin-1-yl}propan-
2-yl)-2-phenylcyclopropanecarboxamide, and #857372
named VU0285655-1 N-{2-[4-oxo-1-phenyl-1,3,8-triaza-
spiro(4.5)decan-8-yl]ethyl}quinoline-3-carboxamide. Lyoph-
olized power was reconstituted in DMSO as a 50X stock.
Negative controls were run with all enzymes in the kinetic
assays using 2% DMSO with enzymes in the absence of
inhibitor to subtract off any inhibition due to DMSO.
For initial screening of inhibitors for the inhibition of PLD
activity, the pH-stat assay +/À inhibitors was used with
the bacterial PLD_sp enzyme. About 50 mM NaOH stock
solution was standardized with KHP and then diluted in
Millipore water (10-fold serial dilutions) and then used to
hold the pH constant (at 8.0) in fixed concentrations of
5 mM for diC4PC, 150 mM NaCl. The diC4PA product
decreased the pH and was in a 1:1 ratio with moles NaOH
needed to stabilize the pH at 8.0. This pH is appropriate
as the pKa2 of the diC4PA side chain is ~ 6.8 so that all of
this product will be titrated at 8.0 (15).
Preparation of PLD enzymes for kinetic assays and
POPC vesicles
PLD_sp from Streptomyces species strain pmf was pur-
chased from EnzoLifeSciences catalog number BML-
SE301-0025 at 95% pure and dialyzed in 25% glycerol
(cryo-protectant) and frozen in À80 °C freezer in small aliqu-
ots and was used without further purification for initial
screening of small molecules. Human PLD1 and PLD2
sequences were purchased Thermo Scientific Inc., (Wal-
tham, MA, USA). The linear sequences were modified with
PCR to have EcoR1- and BamH1-flanking sequences that
were then digest along with the vector pMET and then
ligated into the vector’s multiple cloning site (MCS) from the
Pichia (P. methanolica) Kit catalog number K1780-01
(http://www.lifetechnologies.com/us/en/home.html
accessed 12/16/13). After selecting, scale up, harvesting
the vector and insert, the construct was sent out for
sequencing (Genewiz, Cambridge, MA, USA) and once con-
firmed the PLD sequence was intact, it was cut and trans-
fected with electroporation in the yeast cells to select cells,
induce expression and harvest PLD proteins. Details of the
cloning, purification, and initial characterization of the
human PLD1 and human PLD2 clones will be published
in subsequent publications. Preparation of 1-palmitoyl-2-
oleoyl-PC (POPC) vesicles was performed with slight
modifications (microscale) as previously described using a
sonicator with vesicle samples on ice at 50% duty cycle until
the sample was optically clear as possible (OD580 < 0.100)
(15–17). Dynamic light scattering was used to confirm that
vesicles were unilamellar and estimate the size of the SUV
and LUV vesicles as previously described (17).
In vitro kinetic assay using [choline-methyl-3H]di-
palmitoylPC vesicles for PLD inhibition studies
Compounds that yielded and estimated IC50 of <100 lM
were further quantified and tested using a PLD activity
assay with human PLD1 and PLD2 recombinant proteins
measuring the release of [methyl-3H] choline from
[choline-methyl-3H]di-palmitoylphosphatidylcholine
([cho-
line-methyl-3H]PC) vesicles as previously described (1,18).
Modifications were made to optimize activity especially for
the human PLDs. For example, PLD1 required the addition
of cofactor ADP-ribosylation factor (ARF) for optimal activ-
ity (18). Briefly, PLD enzymes (at 50 nM) were reconstituted
with phospholipid vesicle substrates. Lipid solutions were
dried and resuspended, and SUV were prepared by soni-
cation as sited previously. All assays were performed at
37 °C with agitation for 30 min. Reactions were stopped
by the addition of trichloroacetiacid and bovine serum
albumin. Free [methyl-3H] choline was separated from pre-
cipitated lipids and proteins by centrifugation and analyzed
by liquid scintillation counting. Raw data are normalized
and are presented as relative activity. Experiments are per-
formed in triplicate (18). Once the baseline specific activity
was determined for the enzyme, known amounts of inhibi-
tor ranging from 0.1 nM–0.1 mM concentration were
added. Initial molecule screening was performed with small
molecules listed in Table 2 in the micromolar range (~ 10
data points in duplicate). Then, full duplicate curves were
272
Chem Biol Drug Des 2014; 84: 270–281

Potent Inhibitors for Phospholipase D
generated higher or lower than these initial concentrations
depending on the efficiency of each individual molecule.
The pH-stat assay was run and analyzed as previously
described (15–17). Briefly, serial dilutions were made from
a small initial volume of inhibitor 0.1 mM stock in 5%
DMSO. The IC50 (point of 50% inhibition) was determined
graphically by graphing the relative specific activity (Y axis)
and inhibitor concentration (X axis) using KALEIDAGRAPH
curve-fitting software (2457, Reading, PA, USA). Prepara-
tion of POPC vesicles and diC4Pa was performed with
slight modifications (microscale) as previously described
using a sonicator with vesicle samples on ice at 50% duty
cycle until the sample was optically clear as possible
(OD580 < 0.100) (15–17).
was generated. All rates are based on assays run at least
in duplicate. 31P NMR was also used to quantify PA pro-
duction in in vivo assays with PLD2 in HEX-293 cells after
phorbol 12-myristate 13-acetate (PMA) treatment (19).
Cell culture and transfection
Human embryonic kidney cells (HEK-293) were cultured in
DMEM with 10% fetal calf serum and used to study
endogenous expression of PLDs as well as transfection
with exogenous human PLDs. Cells were grown in 24-well
plates. A construct with modified AAV vector, GFP reporter
protein and human PLD1 and human PLD2 was made.
About 0.5 mg of DNA of either construct was used to per-
form transfection with LipofectAMINE Plus (Invitrogen,
Carlsbad, CA, USA). Twelve hours after transfection, the
medium was replaced with fresh growth medium, and the
cells were cultured for another 18 h before viewing on
fluorescent microscope to confirm expression of the con-
struct and proceeding with the in vivo assay described
briefly below and processing the cells for analysis of PLD
activity in the absence and presence of the most potent
inhibitors. Preliminary fluorescent imaging of cells and
cytology studies of cell morphology with inhibitors intro-
duced herein was performed in an inducible Chinese
Hamster Ovary (CHO) cell line cultured in Ham’s F12
media and 10% tetracycline free fetal bovine serum (FBS).
The CHO cell line has been confirmed to induce (weakly)
overexpression wild-type PLD1 and PLD2 by adding
2.5 lg/mL doxycycline (Dox) for 24 h in a protocol modi-
fied from work previously described (21–24). Preliminary in
vivo PLD inhibitory assays were carried out under culture
conditions as described previously with these cells (21).
1H NMR assay of PLD activity with phospholipid
vesicles
1H NMR (500 MHz) spectra, monitoring choline produc-
tion, were acquired at a temperature of 37 °C. The param-
eters used in acquiring spectra, acquisition time, relaxation
delay, and pulse width were modeled from those used in
previous studies (19). Chemical shifts were referenced to
the residual water resonance (before presaturation) as pre-
viously described (15–17,19,20). The total volume of each
assay sample was 0.250 mL. An initial spectrum was
acquired before adding the enzyme and calcium of the
prepared phospholipid samples. This served as the zero
time control. PLD_sp enzymes (2.5 lg/sample) +/À inhibi-
tors were preincubated together for 1.5 h prior to adding
to the vesicles. This preincubation of enzyme and inhibitor
before adding enzyme to the vesicles for a kinetic run had
a marked effect on the efficiency of the inhibitors. After the
addition of PLD, an arrayed experiment was carried out for
about 1 h. Initial rates were obtained from the progress
curve for ~5–30% PC hydrolysis as monitored by the
increase in choline N(CH3)3 intensity.
PLD in vivo transphosphatidylation and head-
group hydrolysis (PA sensor) activity assays
In vivo PLD activities were determined using transphos-
phatidylation to measure the accumulation of Ptd-But in
intact cells in a protocol with the new constructs of AAV-
GFP-PLD1 and AAV-GFP-PLD2 modified from previous
research of in vivo kinetic studies of PLD (21–24). HEX-
293 cells were transfected with PLD1 +ARF or PLD2 then
incubated with Inh_1 or Inh_2 (diluted from a 5009 stock
concentration in DMSO) or medium containing a matching
concentration of DMSO for 30 min before addition of 0.6%
1-butanol for 30 min. For the time–course study monitor-
ing phosphotidyl butanol (Ptd-But) production, cells were
preincubated with increasing amounts of Inh_1 for PLD1
and Inh_2 for PLD2 for 10, 20, 30, or 60 min followed by
a 30-min incubation with 0.6% 1-butanol.
31P NMR assays of PLD activity with phospholipid
vesicles
PLD-specific
activities
toward
short-chain
diacy-
lphosphatidylcholine (diCnPC), including diC4PC, diC6PC
and diC7PC, and detergent/POPC mixed micelles were
measured by 31P NMR spectroscopy in the absence of
Ca2+ as described previously (20). Samples typically
included 10 mM phospholipid in 25 mM buffer (variable
depending on pH) in a volume of 0.25 mL, of which
0.05 mL is D2O. The diCnPA product was easily
detectable as a pH-dependent resonance downfield from
the PC resonance (À0.5 ppm). The transphosphatidylation
reaction was examined with diC4PC and CH3OH as sub-
strates. Most rates were determined from fixed time-point
assays (6 to ~25 min) where the PLD_Sp reaction was
stopped by heating at 100 °C in a hot water bath. Amount
of enzymes used was 2.5 mg for each time-point with
fresh enzyme added each time, data collected to prevent
depletion of enzyme due to sequestering when PA product
Both endogenous and exogenous (PLD2 in AAV vector)
human PLD2 hydrolytic activity cleaving PC to PA product
in vivo was analyzed by using methods modified from previ-
ously described in vivo kinetic studies (18,21). Briefly, a fluo-
rescent PA sensor consisting of enhanced GFP fused to a
40-amino acid PA-binding domain from the yeast Spo20
Chem Biol Drug Des 2014; 84: 270–281
273

Kulkarni et al.
protein was utilized GFP-Spo20-PABD or the mutant GFP-
Spo20-PABD-L67R that does not bind PA (21,24–27). Fluo-
rescent intensity generated from the PA sensor and mutant
sensor was tracked by lysing the cells with a sonicator at
50% duty cycle and sucrose gradient ultracentrifugation to
separate the nuclei of cells from the plasma membrane
(and endosomes). A total lipid extraction procedure on the
harvested ultracentrifuged fractions (Nuclear and cell mem-
brane portions separated) of the whole cells was performed
(nuclear and cell membrane portions separated before lipids
were extracted from each) (24,25). Quantitative analysis of
fluorescent intensity on a Perkin-Elmer fluorometer (Model
LS-45) was performed (24–27). Fluorescent intensity was
reported after normalizing for endogenous PLD activity for
HEK-293 cells that were transfected when appropriate.
Endogenous PLD activity was also measured and PA pro-
duction quantified with 31P NMR. HEK-293 cells were cul-
tured in small flasks and serum-starved overnight before
incubation with 500 nM Inh_2 for 60 min before the addition
of 150 ng/mL PMA. Cells were incubated for a further
60 min, then washed in ice-cold PBS, and scraped into
ice-cold methanol. Lipids were extracted using acidified
organic solvents. Dried lipid extracts were resuspended in
methanol, and an aliquot was removed for determination of
total lipid phosphorous after wet digestion in perchloric acid
(23–25). Phosphatidic acid production was then quantified
by 31P NMR.
formed the docking of the ligands to a set of grids describ-
ing the target protein. As a control to make sure the docking
parameters were accurate, the synthetic product diC4PA
that was in the active site of the PLD_sp structure (Figure
3A) (1V0Y), was docked into the active site and compared
with the actual position of that product in the crystal struc-
ture. AutoDock4 studies resulted in docking scores. Once
experimental IC50’s were determined; output files and dock-
ing scores were analyzed for comparison with actual kinetic
data obtained and log coordinates were converted to out-
put.pdb to produce figures of the most potent inhibitors or if
not potent, to illustrate how the molecule may be modified
to increase potency for that category of molecules (Table 2
and Figure 3C).
Results and Discussion
Synthesis of inhibitors and testing of PLD_Sc and
PLD_SpPLD kinetic assay with monomeric
substrate DiC4PC and [choline-methyl-3H]di-
palmitoylPC vesicles to screen potential PLD
inhibitors
All inhibitors were synthesized in good yields as depicted
in Scheme 1 (6,14). Nineteen compounds were prepared
for the biochemical assays. The identification and purity
determination were carried out by mass spectrometry and
NMR spectroscopy (1H and 13C). Details of chemical char-
acterization can be found in Appendix S1.
Western blotting and imaging
Cells were lysed by sonication and run on a 12% SDS-
PAGE gel. The blots were blocked in 1% dry milk in Tris
HCl pH 8.0. Western blotting for primary (antiphospho-
ERK and total ERK and antiphospho-AKT and total AKT)
were purchased from Sigma-Aldrich and Thermo Scientific
(Pierce Protein Biology Products) and used according to
recommended protocols from these companies. DyLight
conjugated secondary antibodies (DyLight 488) were used
for fluorescent detection of proteins.
All small molecules synthesized were initially evaluated for
their ability to inhibit PLD activity with enzyme from S.
species strain pmf (PLD_Sp) in the pH-stat assay using a
synthetic substrate diC4PC as previously described
(1,11,19). PLD activity assay measuring the release of
[methyl-3H] choline from [choline-methyl-3H]di-palmi-
toylphosphatidylch oline ([choline-methyl-3H]PC) vesicles
was used as previously described (17,19,20). Once the
PLD_Sp was prepared, the enzyme was tested with the
commercially available PLD inhibitors to demonstrate that
the bacterial enzyme, which is readily available to pur-
chase from EnzoLifeSciences, has cross-reactivity with the
human PLDs. The results are summarized in Table 1.
PLD_Sp was inhibited by Inh_VU0155056 and
Inh_VU0359595 (a PLD1 specific inhibitor) with IC50’s of
125 and 50 nM, respectively. PLD_Sp also exhibited inhibi-
tion with FIPI, a potent commercially available modified ha-
lopemide PLD1/2 inhibitor. The IC50 for FIPI with PLD_Sp
was 100 nM (18). Like the human PLDs, PLD_Sp is a
member of the PLD superfamily and performs both hydro-
lase and transferase activities. This is consistent with a
mechanism that proceeds through a covalent phosphat-
idylhistidyl intermediate where the rate-limiting step is for-
mation of the covalent intermediate.
Fluorescent signals were generated and imaged UVP gel
docking and imaging system (www.uvp.com 12/16/13)
equipped with a transilluminator, light box, digital camera,
and densitometry software (26,27).
Computer docking of inhibitors into PLD-active
site
Computer Dockings of the Small Molecules synthesized in
this study was carried out into the PLD_sp-active site (PDB
Code 1V0Y). The Autodock4 program suite was used as
previously described (28). Inhibitor structures were drawn in
ChemDraw and transferred into PRODRG to convert inh.pdb
to inh.pdbq for AutoDock (28,29). Three-dimensional coor-
dinates of phospholipase D were downloaded from RCSB.
All the experimental ligands were removed from the pocket
of interest. The energy-minimized ligand with full charges
was loaded into the PRODRG program. The AutoDock4 per-
Human recombinant PLD1 exhibited an IC50 of 255 nM
with Inh_VU0155056 reported to be the dual PLD1 and
PLD2 inhibitor (Table 1). The FIPI inhibitor was 10-fold
274
Chem Biol Drug Des 2014; 84: 270–281

Potent Inhibitors for Phospholipase D
Table 1: Comparison of IC50’s of bacterial and human PLDs
with commercially available PLD inhibitors and [choline-methyl-3H]
labeled di-palmitoylPC vesicles as substrate (15–20)
Table 2: Structure–activity relationship for the inhibition of PLD
enzymes determined by using both monomer substrate DiC4PC
and [choline-methyl-3H] labeled di-palmitoylPC vesicles
Human
PLD1 (nM)
Human
PLD2 (nM)
R1
NH2
a
a
PLD_Spa (nM)
N
N
PLD enzyme
N
N
Inh_VU0155056
255
36
225
1200
450
15
125
50
R2
b
IC50
Inh_VU0359595
Group Inhibitor R1
R2
PLD1 PLD_sp PLD2
b
IC50
Inh_VU0285655-1
>1000
25
1225
100
I
1
2
1-naphthyl 4-MeO-
Ph
4-MeO-
Ph
2-naphthyl tert-butyl
1-naphthyl tert-butyl
5
10
125
15
b
IC50
FIPI_HClc
phenyl
250
275
aHuman PLD1 and PLD2 sequences were purchased.
bAll data reported here use the following inhibitors purchased from
Avanti Polar Lipids: Inh_VU0359595 is a PLD1-specific inhibitor
and Inh_VU0285655-1 is PLD2 specific.
cFIPI was purchased from Sigma-Aldrich and used as a control
for successful.
II
III
3
4
5
6
7
500
426
550
350
1200
375
385
650
790
390
150
350
290
1300
1500
phenyl
methyl
tert-butyl
tert-butyl
1-naphthyl 2-Et-
phenyl
2-Et-
phenyl
2-naphthyl 2-Et-
phenyl
1-naphthyl phenyl
phenyl
phenyl
1-naphthyl methyl
phenyl methyl
2-naphthyl methyl
8
9
phenyl
1500
1550
450
355
1250
1375
more potent than Inh_VU0155056 for PLD1 exhibiting an
IC50 of 25 nM. FIPI is even more potent than the PLD1-
specific inhibitor from Avanti Polar Lipids Inh_VU0359595
which exhibited an IC50 of 36 nM for PLD1 (Table 1). The
human recombinant PLD2 enzyme activity IC50 was
225 nM by dual PLD Avanti inhibitor Inh_VU0155056.
Interestingly, the PLD1-specific inhibitor Inh_VU0359595
inhibited PLD2 in the low micromolar range. The PLD2-
specific inhibitor exhibited an IC50 of 450 nM with PLD2.
IV
V
10
11
12
13
14
15
16
17
18
19
875
1250
900 1500
20
620
250
150 1200
–
545
700
675
1050
120
2300
150
1200
1500
–
phenyl
methyl
800
695
650
methyl
phenol
methyl
VI
H
H
H
1950
1750
2-naphthyl
methyl
–
–
500 1550
1250
As shown in Table 2, the structure–activity relationship
(SAR) studies indicated that all 1,3-disubstituted-4-amino-
pyrazolopyrimidines showed potency in the low-nanomolar
to low-micromolar range for the inhibition of PLD_Sp and
human PLDs. In general, the potency of inhibition was
dependent on the bulkiness of the substituent on 1- and
3- positions R2 and R1, respectively, and the polarity of
the R2 position (Table 2). For PLD_Sp, larger groups (2-
naphthyl or 1–naphthyl) as R1 resulted in a trend toward
higher potency. More specifically, when comparing Inh_1
and Inh_2 IC50s for PLD_sp, the R2 group is playing a lar-
ger role for potency. The ability to hydrogen bond in the
R2 position dominates over bulkiness for the R1 position.
The observation that substituents at R2 that are polar and
can form hydrogen bonds (O, N or F atoms), such as entry
1 in Table 2, may be the most potent inhibitors of the PLD
enzymes for this new class of inhibitors gives us clues as
to what the second generation of inhibitors should look
like. The next generation of inhibitors to be developed may
be with fluorocarbons at this position as in the literature
fluorocarbons have been known to be potent inhibitors.
Inh_2 with an IC50 of 15 nM. Smaller trends are also useful
when designing second-generation inhibitors. Chemically
1-napthyl as an R1 group and 4-MeO-Ph as an R2 group
has the most potent inhibition for PLD1 enzyme (Inh_1).
The second most potent inhibitor is 1-napthyl as R1 group
with methyl as R2 group. The third most potent is methyl
groups in both R1 and R2.
The most effective inhibitor for PLD2 has a phenyl group at
R1 and 4-MeO-Ph group at R2. As with PLD1, a hydro-
gen-bonding atom in R2 makes a large difference in inhibi-
tor potency. Second best is methyl as R1 and phenyl as
R2 (Inh_12). The third best is IC50 of 125 nM (Inh_1) with
R1 napthyl and 4-MeO-Ph as R2. Second-generation
inhibitors might include a methyl as an R1 group and 4-
MeO-Ph as R2 might make a really potent inhibitor.
PLD NMR activity assay on phospholipids in
monomer, micelles, SUV vesicles and mixed
micelles at 37 °C with +/À PLD inhibitors
PLD1 shows highest potency for Inh_1 like PLD_sp with
an IC50 of 5 nM 4- to 20-fold lower than the next best two
inhibitors. Inh_1 was used for the in vivo studies with this
enzyme. In contrast, PLD2 exhibited highest potency with
Once small molecules that were potent inhibitors of PLD
activity were identified by synthetic short-chain phospholipids
Chem Biol Drug Des 2014; 84: 270–281
275

Kulkarni et al.
1
0.9
0.8
and 3.18 ppm were used to monitor inner and outer leaflet
of PC in SUV and the emergence of free choline at
~3.09 ppm (16,19). Although overall we had slightly higher
(20-30%) specific activities than previously reported for
these NMR assays for PLD_Sp [See Table 3 and references
(20)], many of the trends previously published for monomer
and vesicle kinetics were reproduced for these inhibitor
studies. As shown in Figure 2B, the hyperbolic kinetics of
PLD activity toward POPC previously described was repro-
duced in the absence (and presence) of PLD inhibitor 1
from Table 2. The inhibitory effects of Inh_1 reduced activity
of the PLD enzymes to 1–5% or original activity for mono-
meric substrates and <10% for vesicle substrates (Fig-
ure 2B and Table 3).
A
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Inh_1 chosen for NMR kinetic and vesicle binding studies
showed high potency in the low-nanomolar range with
both the monomeric pH-stat ande vesicle kinetic assay
(Table 2) had to be used in concentrations in the med-
ian-nanomolar range ~300 nM (Table 3) to get significant
inhibitory effects (~3–4 times more concentrated). This
may be partly due to the fact that phospholipases have a
unique obstacle that complicates kinetics relative to other
enzymes in that to cleave this type of substrate. For
every catalytic cycle, enzymes in the phospholipid family
must have a mechanism to insert into the leaflets of the
bilayer, extract monomer substrate from the lipid bilayer,
cleave the monomer to products, and then release the
products into the lipid bilayer to turnover efficiently. The
complexity of the kinetics is reflected in the disparity
between the amount of substrate necessary to be pre-
sented to the enzyme in vesicle form to reach the appar-
ent Km and Vmax. Due to low yields of some inhibitors,
the full vesicle kinetic studies were limited to Inh_1 for
PLD_sp.
0
5
10
15
20
25
[Inhibitor 1] nM
B
0.3
0.25
0.2
0.15
0.1
0.05
0
0
10
20
30
40
50
60
Time (min)
Figure 2: Effect of Inhibitor 1 on PLD_sp (A) with monomer
diC4PC pH state assay showing a sample of data from a pH stat
assay run in duplicate in the presence of 5 mM diC4PC and
increasing amount of Inhibitor 1 in the nanomolar range. All data
points were run in duplicate. Error for duplicate data points was
between 15% and 20% (B) Choline production monitoring
hydrolysis of 1-palmitoyl-2-oleoyl-PC (POPC) unilamellar vesicles
as substrate with 1H Spectra. The formation of free choline was
monitored to calculate specific activity as previously described
(17,18,21) using 10 mM POPC vesicles in 25 mM imidazole pH 7.0
here in the absence (circles) and presence (Triangles) of Inhibitor
1. All data were collected in duplicate experiments, and
experimental error was between 15% and 20%.
Overall, we had slightly higher (20–30%) specific activities
than previously reported for these NMR assays (See
Table 3 and references (18,20,21)). Table 3 showed
remarkable elimination of transferase activity for the
PLD_sp enzyme in the presence of inhibitor 1 (negligible
transferase activity detected at as low as 50 nM inhibitor
concentration). Inhibition of PLD activity toward vesicle
substrate was observed to be 2–3-fold less potent overall
than monomer substrate. This is probably due in part to
the complex mode of binding and catalysis of enzymes
that work at the lipid bilayer interface to extract monomer
substrate discussed previously. Interestingly, for both
NMR and in vivo studies, the inhibition of vesicle hydroly-
sis was much more effective if the enzyme was preincu-
bated for 60–90 min at 37 °C with inhibitor. Directly
adding inhibitor and enzyme to the vesicles for these
studies resulting in addition of up to 300 nM inhibitor to
get the same effect observed with 30–120 nM (with prein-
cubation with PLD enzyme) reported in Table 3. This
implies a conformational change may be required in the
PLD enzyme to limit binding to the POPC vesicle. It is
hoped that a future crystal structure of the PLD/inhibitor
(Table 2 and Figure 2a), further kinetic studies were carried
out on vesicles using 31P and H NMR. NMR studies have
1
been established as the preferred method to measure PLD
activity with vesicles instead of pH-stat assays with vesicles
due to the high pKa2 of PA in vesicles (pKa2 of 7.6 in pre-
dominantly PC bilayers increasing as PA increases) which
makes pH-stat assays of PLD-mediated cleavage of PC to
PA problematic because at an end-point of 8.0 only part of
the PA is titratable. 1H NMR spectroscopy was used to
monitor PLD activity by measuring the intensity of the
water-soluble choline N-methyl resonance as previously
described in detail (19,20). Briefly, chemical shifts of 3.15
276
Chem Biol Drug Des 2014; 84: 270–281

Potent Inhibitors for Phospholipase D
Trp470
Table 3: PLD_sp activity toward phospholipids in monomer,
micelles, vesicles, and mixed micelles at 37 °C +/À Inh_1 (inhibi-
tor) using NMR
A
Asn465
Leu471
Lys450
Conc. Ca2+ Physical
Substrate (mM) (mM) state
Specific
hydrolasea
Activity
transferasea
Leu351
Tyr467
Lys172
His170
His448
diC4PC
diC6PC
diC7PC
POPC
POPC
POPC/
chol
5
5
5
10
10
5/5
0
0
0
5
5
5
Monomer 0.15 (15.5)
Monomer 0.22 (11.5)
0.00 (10.3)
0.00 (3.26)
0.25 (2.96)
Phe166
Micelle
SUV
0.58 (8.5)
diC4PA
Trp168
0.49 (5.4)b
0.23 (2.10)b
0.44 (5.90)b,c
Asn187
LUV
SUV
Trp189
POPC +
TX-100
5/20
0
Mixed
5.25 (125)d
micelle
Trp470
B
Asn465
PA, phosphatidic acid; POPC, 1-palmitoyl-2-oleoyl-PC; SUV, small
unilamellar vesicles.
Leu471
Lys450
aHydrolase (generation of PA) and transferase (generation of PG)
activities of PLD_sp (2.5 ug) toward 5 mM diCnPC with 5 mM glyc-
erol present in 50 mM MES-NaOH buffer, pH 6.5; 31P NMR was
used to monitor PA and PG production.
Leu351
Lys172
His170
Tyr467
His448
bThe specific activity of the enzyme was measured by 1H NMR
spectroscopy monitoring for choline production. This is known as
a combination of hydrolase and transphosphatidylation reaction
rates. Duplicate assay samples were used to obtain rates; errors
in the rates were between 15% and 20%.
Phe166
Trp168
1
Asn187
cThe specific activity is based on the rate over the first 25 min;
unlike with pure POPC SUVs, the rate decreased over longer
times when cholesterol was present in the vesicles.
Trp189
Trp470
C
dWith this amount of TX-100 present, the major (>90%) product
for PLD_sp is the transphosphatidylation product with the Triton
molecule (20).
Asn465
Leu471
Lys450
Lys172
Leu351
His170
Tyr467
complex will supply information needed to fully interpret
these results.
His448
Trp168
Asn187
Phe166
Analysis of docking of small molecules into
PLD_sp-active site (PDB_ID 1V0Y)
9
Trp189
Dibutyrylphosphatidylcholine (diC4PC) is the soluble syn-
thetic substrate for PLD (also used in pH-stat assay in this
study) that was reacted with the crystallized PLD_sp
enzyme crystal. The crystallized enzyme actually cleaved
the substrate to products choline and 2-butyryloxy-1-
tetrahydroxyphosphoranyloxy-methyl-ehtylbutyrate (diC4PA)
before data were collected for the structure/ligand com-
plex (PDB code 1V0Y) (11) Figure 3A shows the PLD_sp
enzyme (1V0Y) cocrystallized with diC4PA. The active site
consists of hydrophobic areas to the left and upper right.
The short alkyl chains of the diC4PA product lie toward the
hydrophobic region of the interior of the pocket. There are
hydrophilic areas of the enzyme below center of gravity of
the active site cavity (Asn465, His170, Lys450, Lys172,
His448, and Asn187) surrounding the phosphate moiety
where the polar portion of this molecule is bonded (Fig-
ure 3A). The most potent inhibitor for the PLD_sp enzyme
is docked as described in experimental methods (inhibitor
1 from Table 2) with IC50 of 10 nM in Figure 3B to look for
Figure 3: The PDB use for Docking Studies is 1V0Y (S_sp. PLD)
(A) PLD bound to synthetic product diC4PA (after being soaked in
synthetic substrate diC4PC) in the crystal structure from RCSB
(PDB code 1V0Y) (B) 1 docked in PLD (C) 9 docked in PLD.
possible trends on how these types of inhibitors may bind
to the enzyme. The tight binding observed from the actual
IC50 may be due to the hydrogen-binding capacity of the
oxygen of the R2 (4-methoxyphenyl) substituent in the
phosphate-binding pocket of the enzyme suggested by
the predicted binding in the molecular docking (Figure 3B).
In contrast, PLD_sp binding to docked 9 is shown in Fig-
ure 3C; without hydrogen-binding capacity at R2, the IC50
increases to 350 nM. This trend for PLD_sp inhibition
shows that the R2 group of Inh_1 has enhanced inhibitory
capacity for the PLD_sp enzyme when able to hydrogen
Chem Biol Drug Des 2014; 84: 270–281
277

Kulkarni et al.
1
A
Figure 4: Inh_1 and Inh_2 are potent in vivo inhibitors. (A) HEX-
293 cell lines were transfected with PLD1 and PLD2 constructs
overnight, and then once expression confirmed (GFP reporter
protein), 0.6% 1-butanol was added, and Ptd-But production was
PLD2 + Inh_2
PLD1 + Inh_1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
measured
to
quantify
transphosphatidylation
reaction.
Measurements were taken in triplicate and normalized for any
endogenous PLD activity (+/À PMA). Increasing concentrations of
Inh_1, Inh_2 or DMSO control were added. (B) Inh_2 blocks
translocation of a phosphatidic acid (PA) sensor to the plasma
membrane. PLD2 over-expressing HEX-293 cells (transfected
overnight with AAV-GFP-PLD2 construct) were transfected with
the PA sensor GFP-Spo20-PABD or the mutant GFP-Spo20-
PABD-L67R that does not bind PA. After the transfection with the
PA sensors, cells were viewed under the fluorescent microscope
to confirm negative and positive controls were working. Then,
after separating nuclei from membrane with ultracentrifugation and
total lipid extraction, PA sensor for each subcellular location was
quantified on a Perkin-Elmer LS-45 flourometer. Assays were
performed in duplicate. (C) Analysis of PA production in
endogenously PLD expressing HEX-293 cells. HEK-293 cells À/+
PMA treatment as a positive control. Phosphatidic acid production
was evaluated in 500 nM Inh_2. Phosphatidic acid was quantified
after total lipid extraction followed by 31P NMR.
0
2
4
6
8
10
12
14
[Inhibitor] nM
B
Endosomes
plasma membrane
Nuclear
fraction
4000
3500
3000
2500
2000
1500
1000
500
0
Inh_2 exhibited potent inhibition of exogenous PLD
transphosphatidylation activity. Inh_1 IC50 is ~3.75 nM and
Inh_2 IC50 is ~9.5 nM. Figure 4B shows that Inh_2 is a
potent in vivo inhibitor blocking phospholipid head-group
cleavage of PC. When cells are inactive, the majority of
the sensors are localizes to the nucleus (Figure 4B control
group 1). When cells are overexpressing PLD2, the PA
sensor fluorescent intensity was found to increase by
about 100-fold in fluorescent intensity within the subcellu-
lar plasma membrane fraction, strongly suggesting that
more PA was moving into the plasma membrane fraction
and to intracellular membrane vesicles (endosomes) colo-
calized outside the nucleus (Figure 4B compare groups 1
and 2 plasma membrane/endosome subcellular fraction
fluorescent intensity). This is known to happen when PLDs
are actively generating PA from PC. After addition of
100 nM of Inh_2 to the PLD2-overexpressing cells, the
fluorescent intensity of the nuclear fraction and plasma
membrane/endosomes were reversed back to the profile
of group 1 with low or no PLD activity in resting cells, indi-
cating a lack of generation of PA by PLD2. Thus, Inh_2 is
a potent inhibitor of head-group hydrolysis of PC to PA.
Repeating the assay with the mutant PA sensor indicates
1
2
3
4
5
6
Control
PLD2
PLD2
100 nM Inh_2
Control PLD2
PLD2
100 nM Inh_2
Mutant PA sensor
45
40
35
30
25
20
15
10
5
C
Basal
+PMA
0
1
2
Control
500 nM Inh_1
bond to polar side chains in the phosphate-binding
a more random distribution of fluorescent intensity
pocket. (Table 2: compare 1 and 9 and Figure 3B,C).
between the plasma membrane and nuclear fractions ana-
lyzed (negative controls Figure 4B Groups 4, 5, and 6).
Endogenous activity of PLD2 is increased by stimulating
HEK-293 cells with PMA. Figure 4C group 1 shows
endogenous PLD activity (relative PA levels) nearly 30%
higher in the presence of PMA. As shown in Figure 4C
group 2, treatment of HEK293 cells with 500 nM Inh_2
blocks both endogenous and PMA stimulated PLD activity
(PA production by dropping relative levels of PA by 20%
without PMA stimulation and over 50% in PMA-stimulated
cells comparing groups 1 and 2 in Figure 4C.
In vivo transphosphatidylation and head-group
hydrolysis (PA production) kinetic assays
Figure 4A measures Inh_1 and Inh_2 capacity to inhibit
PLD trasnphosphatidylation by measuring the production
of Ptd-But. HEX-293 cells overexpressing PLD1 or PLD2
were incubated with increasing amounts of Inh_1 and
Inh_2, respectively, or DMSO control before addition of
0.6% 1-butanol. As seen in Figure 4A, both Inh_1 and
278
Chem Biol Drug Des 2014; 84: 270–281

Potent Inhibitors for Phospholipase D
the release of [methyl-3H] choline from [choline-methyl-3H]
di-palmitoylphosphatidylcholine
A
B
([choline-methyl-3H]PC)
vesicles as previously described (18,21). For PLD_sp, the
hydrophobicity of the active site requires some ‘bulk’ at
position R2 comparing inhibitors 18 (no bulk at R2) with 2,
9, and 15 (Table 2). The most potent inhibitors for PLD_sp
would be those that have some hydrophobicity at R2 and
are not too bulky, yet can engage in hydrogen bonding as
shown with 1 (Table 2 and Figure 3B). NMR vesicle kinetic
assays were also performed. For the recombinant human
PLD1 and PLD2 enzymes, Inh_1 and Inh_2 molecules were
used in the in vivo cell-based kinetic assays. In the trans-
phosphatidylation kinetic assay (increase of Ptd-But prod-
uct) in vivo, Inh_1 and Inh_2 exhibited IC50’s of
approximately 3.75 and 8.5 nM with PLD1 and PLD2,
respectively (Figure 4A). For in vivo head-group cleavage
Figure 5: PLD Inhibitors Inh_1 and Inh_2 do not alter upstream
signaling events. Inh_1 and Inh_2. Do not effect AKT
phosphorylation upon serum stimulation. MDA-MB-231 cells were
starved overnight before being pretreated with 500 nM Inh_1 or
Inh_2 or DMSO (control) for 12 h and then stimulated with serum
for 45 min. Cell lysates were analyzed by SDS-PAGE and
Western blotting with antibodies against: (A) phosphorylated AKT,
total AKT, and (B) phosphorylated ERK, or total ERK.
assay,
a PA product-tracking molecule was utilized.
Results of subcellular fractionation separating nuclei from
plasma membrane and analysis of total lipids of each
fraction showed that Inh_2 was very potent (100 nM) at
preventing the translocation of the PA tracker from the
nucleus to the cytoplasm and intracellular vesicles, which
happens when PA product builds up in the absence of the
inhibitor (18). There was remarkable consistency between
the in vitro and in vivo assays, which have physiological
relevance. The docking studies using AutoDock4 sug-
gested that the naphthyl group as R1 lies in the hydropho-
bic region of the pocket but there is room for expansion or
elongation to go deeper into the pocket (more like the alkyl
chains of phospholipids (Figure 3B,C). Overall, the exoge-
nous inhibition and in vivo inhibition observed with some of
the novel molecules in this study are at least as potent as
previously reported specific and highly effective inhibitors
to mammalian PLD tested in both cell-based and in vitro
assays.
Inh_1 and Inh_2 do not affect PLD subcellular
localization, actin stress fiber network in resting
HEK-293 cells and do not affect signaling events
upstream from PLD activation
There are molecules known to dislocate PLDs from the
plasma membrane where they are associated when PLD
is active. This would be indirect inhibition. Patterns in
actin fibers viewed through the fluorescent microscope
did not show visual differences when comparing cells
used in these In vivo kinetic assays over-expressing
human PLD1 and PLD2 in the presence and absence of
Inh_1 and Inh_2 (data not shown). As PLD activation has
been reported in the literature to occur downstream of
cell signaling events such as AKT and/or ERK activation
(18), examination of phosphorylated and total AKT and
ERK was performed with Western blot analysis in the
presence and absence of Inh_1 and Inh_2. As shown in
Figure 5A,B, serum-starved cells did not express high lev-
els of phosphorylated AKT and ERK. The presence of
serum stimulates phosphorylation of AKT and ERK. Com-
paring lanes 2, 4, and 6 for phosphorylated AKT and
ERK (p-AKT and p-ERK), the presence of 500 nM of
either Inh_1 or Inh_2 for 12 h does not significantly alter
the expression of the phosphorylated form of these sig-
naling molecules. This is strong evidence that these PLD
inhibitors act specifically on the PLD enzyme in these in
vivo assays.
Acknowledgments
Financial help from University of Massachusetts Boston
and technical support from Roxbury Community College
Biotechnology Program are gratefully acknowledged. KS
would like to thank Feng Wang formerly of Boston Univer-
sity for helpful discussions about molecular modeling of
the binding of enzyme/lipid bilayers for follow-up and
future studies and the technical staff at Boston University’s
Chemical Instrumentation Center (CIC) for their assistance
with NMR to complete some of these kinetic studies. A
special thanks to Vicky Curry formerly of RCC, now at
Dana Farber Cancer Institute for all her extra work on this
project. Also, thanks to the laboratory staff in Pediatric
Oncology to prepare the PA fluorescent probes (wild type
and mutant) and the CHO PLD–inducible cell line and to
maintain the MDA-MB-231 cells for the Western blotting
experiments for this study. Thanks to Dr. Polina Anikeeva
and Ritchie Chen of MIT Center of Material Sciences and
Engineering for their time training me in molecular biology
techniques in Dr. Anikeeva’s RET program (summer 2013)
Conclusions
In conclusion, a diverse set of substituted pyrazolopyrimi-
dines have been prepared in a five-step synthesis. These
compounds were tested for the inhibition of PLD_Sp and
human PLD1 and PLD2 enzymes initially using the pH-stat
assay, followed by a standard PLD kinetic assay measuring
Chem Biol Drug Des 2014; 84: 270–281
279

Kulkarni et al.
and for donating the original AAV-gCAMP vector, HEX-293
cells and enzymes used to prepare the PLD1 and PLD2
constructs with Green Fluorescent Protein (GFP). Thanks
to Mark R. Payne and Julie V. Charbeneau for their excel-
lent editorial skills and summary of PLD1 and PLD2 trends
for in vitro inhibition (Table 2).
11. Leiros I., Mcsweeney S., Hough E. (2004) The reaction
mechanism of phospholipase D from Streptomyces sp
strain PMF. Snapshots along the reaction pathway
reveal a pentacoordinate reaction intermediate and an
unexpected final product. J Mol Biol;339:805–820.
12. Bishop A.C., Kung C., Shah K., Witucki L., Shokat
K.M., Liu Y. (1999) Generation of onospecific nanomo-
lar tyrosine kinase inhibitors via a chemical genetic
approach. J Am Chem Soc;121:627–631.
References
13. Apsel B., Blair J.A., Gonzalez B., Nazif T.M., Feldman
M.E., Aizenstein B., Hoffman R., Williams R.L., Shokat
K.M., Knight Z.A. (2008) Targeted polypharmacology:
discovery of dual inhibitors of tyrosine and phospho-
inositide kinases. Nat Chem Biol;4:691–699.
1. Brown H.A., Henage L.G., Preininger A.M., Xiang Y.,
Exton J.H. (2007) Biochemical analysis of phospholi-
pase D. Methods Enzymol;434:49–87.
2. Foster D.A. (2009) Phosphatidic acid signalling to
mTOR: signals for the survival of human cancer cells.
Biochim Biophys Acta;1791:949–955.
14. Hanefeld U., Rees C.W., White J.P., Williams D.J.
(1996) One-pot synthesis of tetrasubstituted pyrazoles-
3. Zhao Y., Ehara H., Akao Y., Shamoto M., Nakagawa
Y., Banno Y., Deguchi T., Ohishi N., Yagi K., Nozawa
Y. (2000) Increased activity and intranuclear expression
of phospholipase D2 in human renal cancer. Biochem
Biophys Res Commun;278:140–143.
4. Uchida N., Okamura S., Kuwano H. (1999) Phospholi-
pase D activity in human gastric carcinoma. Anticancer
Res;19:671–676.
5. Yamada Y., Hamajima N., Kato T., Iwata H., Yamam-
ura Y., Shinoda M., Suyama M., Mitsudomi T., Tajima
K., Kusakabe S., Yoshida H., Banno Y., Akao Y.,
Tanaka M., Nozawa Y. (2003) Association of a poly-
morphism of the phospholipase D(2) gene with the
proof of regiochemistry.
Trans;1:1545–1552.
J
Chem Soc Perkin
15. Geng D., Baker D.P., Foley S.F., Zhou C., Stieglitz K.,
Roberts M.F. (1999) A 20-kDa domain is required for
phosphatidic acid-induced allosteric activation of phos-
pholipase D from Streptomyces chromofuscus. Bio-
chim Biophys Acta;1430:234–244.
16. Stieglitz K., Seaton B., Roberts M.F. (1999) The role of
interfacial binding in the activation of Streptomyces
chromofuscus phospholipase D by phosphatidic acid.
J Biol Chem;274:35367–35374.
17. Stieglitz K.A., Seaton B., Roberts M.F. (2001) Binding
of proteolytically processed phospholipase
D from
prevalence of colorectal cancer.
81:126–131.
J
Mol Med;
Streptomyces chromofuscus to phosphatidylcholine
membranes facilitates vesicle aggregation and fusion.
Biochemistry;40:13954–13963.
6. Scott S.A., Selvy P.E., Buck J.R., Cho H.P., Criswell
T.L., Thomas A.L., Armstrong M.D., Arteaga C.L.,
Lindsley C.W., Brown H.A. (2009) Design of isoform-
selective phospholipase D inhibitors that modulate can-
cer cell invasiveness. Nat Chem Biol;5:108–117.
7. Jin J.K., Ahn B.H., Na Y.J., Kim J.I., Kim Y.S., Choi
E.K., Ko Y.G., Chung K.C., Kozlowski P.B., Mindo S.
(2007) Phospholipase D1 is associated with amyloid
precursor protein in Alzheimer’s disease. Neurobiol
Aging;28:1015–1027.
8. Oliveira T.G., Di Paolo G. (2010) Phospholipase D in
brain function and Alzheimer’s disease. Biochim Bio-
phys Acta;1801:799–805.
9. Lewis J.A., Scott S.A., Lavieri R., Buck J.R., Selvy
P.E., Stoops S.L., Armstrong M.D., Brown H.A.,
Lindsley C.W. (2009) Design and synthesis of iso-
form-selective phospholipase D (PLD) inhibitors. Part
I: impact of alternative halogenated privileged struc-
tures for PLD1 specificity. Bioorg Med Chem
Lett;19:1916–1920.
18. Su W., Yeku O., Olepu S., Genna A., Park J.S., Ren
H., Du G., Gelb M.H., Morris A.J., Frohman M.A.
(2009) 5-Fluoro-2-indolyl des-chlorohalopemide (FIPI),
a phospholipase D pharmacological inhibitor that alters
cell spreading and inhibits chemotaxis. Mol Pharma-
col;75:437–446.
19. Geng D., Chura J., Roberts M.F. (1998) Activation of
PLD by phosphatidic acid: enhanced vesicle binding,
phosphatidic acid-Ca2+ interaction, or an allosteric
effect? J Biol Chem;273:12195–12202.
20. Yang H., Roberts M.F. (2003) Phosphohydrolase and
transphosphatidylation reactions of two Streptomyces
phospholipase D enzymes: covalent versus non-cova-
lent catalysis. Protein Sci;12:2087–2098.
21. Hammond S.M., Altshuller Y.M., Sung T.C., Rudge
S.A., Rose K., Engebrecht J., Morris A.J., Frohman
M.A. (1995) Human ADP-ribosylation factor-activated
phosphati- dylcholine-specific phospholipase D defines
a new and highly conserved gene family. J Biol
Chem;270:29640–29643.
10. Lavieri R., Scott S.A., Lewis J.A., Selvy P.E., Arm-
strong M.D., Brown H.A., Lindsley C.W. (2009) Design
and synthesis of isoform-selective phospholipase D
(PLD) inhibitors. Part II: identification of the 1, 3, 8-tri-
azaspiro[4,5]decan-4-one privileged structure that
engenders PLD2 selectivity. Bioorg Med Chem
Lett;19:2240–2243.
22. Su W., Chardin P., Yamazaki M., Kanaho Y., Du
RhoA-mediated G. (2006) Phospholipase D1 signaling
is not required for the formation of stress fibers and
focal adhesions. Cell Signal;18:469–478.
23. Du G., Huang P., Liang B.T., Frohman M.A., MA,
(2004) Phospholipase D2 localizes to the plasma mem-
280
Chem Biol Drug Des 2014; 84: 270–281

Potent Inhibitors for Phospholipase D
brane and regulates angiotensin II receptor endocyto-
sis. Mol Biol Cell;15:1024–1030.
24. Morris A.J., Frohman M.A., Engebrecht J. (1997) Mea-
28. Morris G.M., Goodsell D.S., Halliday R.S., Huey R.,
Hart W.E., Belew R.K., Olsen A.J.J. (1998) Automated
docking using a Lamarckian genetic algorithm and
empirical binding free energy function. Comput
Chem;19:1639–1662.
surement of phospholipase
D activity. Anal Bio-
chem;252:1–9.
€
25. Zeniou-Meyer M., Zabari N., Ashery U., Chasserot-Go-
laz S., Haeberle A.M., Demais V., Bailly Y., Gottfried I.,
Nakanishi H., Neiman A.M., Du G., Frohman M.A.,
Bader M.F., Vitale N. (2007) Phospholipase D1 produc-
tion of phosphatidic acid at the plasma membrane pro-
motes exocytosis of large dense-core granules at a late
stage. J Biol Chem;282:21746–21757.
29. Schuttelkopf A.W., van Aalten D.M.F. (2004) A tool for
high-throughput crystallography of protein-ligand com-
plexes. Acta Crystallogr D Biol Crystallogr;D60:1355–
1363.
Supporting Information
26. Kang J.H., Chung J.K. (2008) Molecular-genetic imag-
Additional Supporting Information may be found in the
online version of this article:
ing based on reporter gene expression.
J Nucl
Med;49:164S–179S.
27. Jung A.C., Criqui M.C., Rutschmann S., Hoffmann
J.A., Ferrndon D. (2001) Microfluorometer assay to
measure the expression of beta-galactosidase and
green fluorescent reporter genes in single Drosophila
files. Biotechniques;30:594–598, 600-1.
Appendix S1. 1H and 13C NMR spectra of the molecules
synthesized for PLD Inhibitor Studies.
Chem Biol Drug Des 2014; 84: 270–281
281