js-k, a glutathione/glutathione s-transferase-activated nitric oxide donor of the diazeniumdiolate...
TRANSCRIPT
-
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
1/10
2003;2:409-417.Mol Cancer TherPaul J. Shami, Joseph E. Saavedra, Lai Y. Wang, et al.
1Antineoplastic ActivityPotentNitric Oxide Donor of the Diazeniumdiolate Class with
Transferase-activatedS-JS-K, a Glutathione/Glutathione
Updated version
http://mct.aacrjournals.org/content/2/4/409Access the most recent version of this article at:
Cited Articles
http://mct.aacrjournals.org/content/2/4/409.full.html#ref-list-1This article cites by 26 articles, 2 of which you can access for free at:
Citing articles
http://mct.aacrjournals.org/content/2/4/409.full.html#related-urlsThis article has been cited by 9 HighWire-hosted articles. Access the articles at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
SubscriptionsReprints and
[email protected] atTo order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
To request permission to re-use all or part of this article, contact the AACR Publications
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/content/2/4/409http://mct.aacrjournals.org/content/2/4/409http://mct.aacrjournals.org/content/2/4/409.full.html#ref-list-1http://mct.aacrjournals.org/content/2/4/409.full.html#ref-list-1http://mct.aacrjournals.org/content/2/4/409.full.html#related-urlshttp://mct.aacrjournals.org/content/2/4/409.full.html#related-urlshttp://mct.aacrjournals.org/cgi/alertshttp://mct.aacrjournals.org/cgi/alertsmailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/mailto:[email protected]:[email protected]://mct.aacrjournals.org/cgi/alertshttp://mct.aacrjournals.org/content/2/4/409.full.html#related-urlshttp://mct.aacrjournals.org/content/2/4/409.full.html#ref-list-1http://mct.aacrjournals.org/content/2/4/409 -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
2/10
JS-K, a Glutathione/Glutathione S-Transferase-activatedNitric Oxide Donor of the Diazeniumdiolate Class withPotent Antineoplastic Activity1
Paul J. Shami,2 Joseph E. Saavedra, Lai Y. Wang,Challice L. Bonifant, Bhalchandra A. Diwan,Shivendra V. Singh, Yijun Gu, Stephen D. Fox,Gregory S. Buzard, Michael L. Citro,David J. Waterhouse, Keith M. Davies, Xinhua Ji,and Larry K. Keefer
Division of Medical Oncology, Department of Internal Medicine,University of Utah and Salt Lake City Veterans Administration MedicalCenters, Salt Lake City, Utah 84148 [P. J. S., L. Y. W.]; Basic ResearchProgram [J. E. S., B. A. D., G. S. B., M. L. C.] and Analytical ChemistryLaboratory [S. D. F.], Science Applications International Corporation-Frederick, National Cancer Institute at Frederick, Frederick, Maryland21702; Department of Pharmacology, University of Pittsburgh CancerInstitute, Pittsburgh, Pennsylvania 15213 [S. V. S., Y. G.];Macromolecular Crystallography Laboratory, National Cancer Instituteat Frederick, Frederick, Maryland 21702 [Y. G., X. J.]; Department ofChemistry, George Mason University, Fairfax, Virginia 22030 [K. M. D.];and Laboratory of Comparative Carcinogenesis, National CancerInstitute at Frederick, Frederick, Maryland 21702 [C. L. B., D. J. W.,L. K. K.]
Abstract
We have previously shown that nitric oxide (NO)
inhibits growth and induces differentiation and
apoptosis in acute myeloid leukemia cells, with the HL-
60 human myeloid leukemia line being particularly
sensitive to NO-mediated cytolysis. With the goal
of identifying a prodrug that can target NO to theleukemia cells without inducing NO-mediated systemic
hypotension, we have screened a series of O2-aryl
diazeniumdiolates designed to be stable at
physiological pH but to release NO upon reaction
with glutathione. O2-(2,4-Dinitrophenyl) 1-[(4-
ethoxycarbonyl)piperazin-1-yl]diazen-1-ium-1,2-diolate
(JS-K) proved to be the most active antiproliferative
agent among those tested in HL-60 cells, with an IC50of 0.20.5 M. After 5 days of exposure to 0.5 M JS-K,
HL-60 cells had differentiated and acquired some of
the phenotypic features of normal monocytes. One- to
2-day treatment with JS-K at concentrations of 0.51
M resulted in apoptosis induction in a concentration-and caspase-dependent manner. JS-K also inhibited
the growth of solid tumor cell lines but to a lesser
extent than HL-60 cells. JS-K was administered i.v. to
nonobese diabetic-severe combined immune deficient
mice at doses of up to 4 mol/kg without inducing
significant hypotension. The growth of s.c. implanted
HL-60 cells was reduced by50% when the mice
received i.v. injections three times/week with 4 mol/
kg boluses of JS-K. Histological examination of tumor
explants from JS-K-treated animals revealed extensive
necrosis. Similar results were seen with s.c. human
prostate cancer (PPC-1) xenografts. Our data indicate
that JS-K is a promising lead compound for the
possible development of a novel class of antineoplastic
agents.
Introduction
AML3 is a life-threatening disease with an annual incidence
of 2.25 per 100,000 (1). Despite therapy with different classes
of chemotherapeutic agents, including anthracyclines, 1--
D-arabinofuranosylcytosine, etoposide, all-trans-retinoic
acid, as well as high-dose therapy with stem cell rescue, the
overall 5-year disease-free survival remains 30%. This il-
lustrates the need for agents with novel mechanisms of
action.This study is based on the premise that the unusual sen-
sitivity of AML cells to the cytotoxic effects of NO can be
exploited for improved therapy of this malignancy. We pre-
viously showed that NO-releasing drugs of the diazenium-
diolate class induce apoptosis in the HL-60 human myeloid
leukemia cell line at relatively low concentrations (for exam-
ple, an IC50 of 0.006 mM for a cell-permeant prodrug that is
activated for NO release by intracellular esterases; Ref. 2).
This is in marked contrast to, for example, normal vascular
smooth muscle cells, which experience cytostasis without
toxicity in the continuous presence of 0.20.5 mM (Z)-1-[2-
(2-Aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2-
diolate (half-life for NO release of 20 h) for 6 days (3), and to
primary hepatocytes, which can actually be protected from
apoptosis by exposure to a diazeniumdiolate that is acti-
vated for NO release by cytochrome P450 (4).
Received 11/25/02; revised 1/30/03; accepted 2/12/03.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indi-cate this fact.1 This work is supported by a Translational Research Award from theLeukemia and Lymphoma Society (to P. J. S.). This project was alsosupported, in part, through National Cancer Institute Contract No. NO1-CO-12400 (J. E. S., B. A. D., S. D. F., G. S. B., M. L. C.), and NationalInstitute of Environmental Health Sciences Grant ES09140 (to S. V. S.).2 To whom requests for reprints should be addressed, at SLC VA MedicalCenter, Box 151M, 500 Foothill Boulevard, Salt Lake City, UT 84148.Phone: (801) 582-1565; Fax: (801) 583-9624; E-mail: [email protected].
utah.edu.
3 The abbreviations used are: AML, acute myeloid leukemia; BSO, buthi-onine sulfoximine; CDNB, 1-chloro-2,4-dinitrobenzene; DNP-SG, S-(2,4-dinitrophenyl)glutathione; GSH, glutathione; GST, glutathione S-transfer-ase; GSTCD, 1-( S-glutathionyl)-2,4,6-trinitrocyclohexadienate anion;HPLC, high performance liquid chromatography; JS-K, O2-(2,4-dinitro-phenyl) 1-[(4-ethoxycarbonyl)piperazin-1-yl]diazen-1-ium-1,2-diolate; MS,mass spectrometry; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetra-zolium bromide; NAC, N-acetyl-L-cysteine; NO, nitric oxide; NOD-SCID,nonobese diabetic-severe combined immune deficient; NSE, nonspecificesterase; xGSTY1-1, class Y (where Y is: A, ; M, ; P, ) GST of subunit
type 1 from x (where x is: h, human; r, rat).
409Vol. 2, 409417, April 2003 Molecular Cancer Therapeutics
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
3/10
One might, therefore, expect NO-based therapies to ex-
hibit some selectivity for the NO-sensitive leukemia cells
relative to normal tissues. With this in mind, we have
screened a library of arylated diazeniumdiolates that were
designed to be activated for NO release by reaction with
cellular thiols such as GSH (5) with or without catalysis by
GST, the major isoforms (, , ) of which are expressed in
75, 55, and 95% of AML cases, respectively (6). We report
here on the in vitro and in vivo antineoplastic activity of the
most potent growth inhibitor of this family tested to date,JS-K (structure in Fig. 1).
Materials and Methods
Chemicals. JS-K (7) and 4-carbethoxy-PIPERAZI/NO (Ref.
8; structures shown in Fig. 1) were synthesized as described
previously. S-(2,4-Dinitrophenyl)glutathione was prepared by
the method of Mancini et al. (9). The pan-caspase inhibitor
Z-VAD-FMK was from Biomol (Plymouth Meeting, PA).
Daunorubicin and etoposide were from Dr. Grayden Harker
(University of Utah). HPLC grade solvents were purchased
from VWR Scientific Co. (South Plainfield, NJ). All other
chemicals were from Sigma (St. Louis, MO) unless otherwise
noted.
Chromatography and MS. HPLC separations were car-
ried out with a Phenomenex Luna 5 C18(2) column (Tor-
rance, CA) using a gradient of acetonitrile:water (each con-
taining 0.1% formic acid) at a flow rate of 1 ml/min
(percentage of acetonitrile: 25% for 0 5 min followed by a
linear program to 70% at 15 min).HPLC-MS studies were performed on an Agilent Capillary
Series 1100 LC/MSD Ion Trap mass spectrometer with elec-
trospray ionization in the positive ion mode. Separations
were effected as above, except that the flow rate was 15
l/min, and the gradient was 10% acetonitrile for 0 5 min
followed by a linear program to 70% at 15 min.
Molecular Modeling. CDNB is a model substrate used
extensively for monitoring activity of GST (10). It has been
shown that GSTs with a higher specific activity toward CDNB
stabilize the Meisenheimer complex (Fig. 1) at the transition
state better than the isoforms with poor catalytic activity for
CDNB-GSH conjugation (5). Crystal structures of a transition
state analogue, GSTCD
, in complex with rGSTM1-1, a class rat GST isoform (11), and with hGSTP1-1 (isoleucine-
104, alanine-113 variant), a class human GST isoform (12),
have been reported and provided the foundation for model-
ing the transition state of GST-catalyzed GSH conjugation of
JS-K. The initial model of the Meisenheimer complex of GSH
and JS-K for hGSTM1-1 and that for hGSTP1-1 built based
on the crystal structures of GSTCD in rGSTM1-1 (11) and
that in hGSTP1-1 (12), respectively. The models were subject
to geometry optimization using the conjugate gradient
method of Powell (13) and docked into the active sites of
ligand-free hGSTM1-1 and hGSTP1-1 built based on the
rGSTM1-1 (11) and hGSTP1-1 (12) structures, respectively.
The geometry of the protein-Meisenheimer complexes wasthen optimized, and the energy was minimized (13). Both
complexes were built in dimeric form considering the fact
that the biologically active forms of GSTs are dimeric pro-
teins and that the glutathionyl moiety of GSH interacts with
the side chains from both subunits (14 17). The Engh and
Huber (18) geometric parameters were used as the basis of
the force field. No crystal structure of class GST-bound
GSTCD is available. The Meisenheimer complex for class
GST was therefore built by modifying those for rGSTM1-1
and hGSTP1-1 based on the structures of hGSTA1-1-bound
S-benzyl-GSH (19) and the GSH conjugate of ethacrynic acid
(20). After energy minimization, the Meisenheimer complex
for hGSTA1-1 was docked in the active center of the enzyme,
and the protein-Meisenheimer complex was subject to en-
ergy minimization as described above. The molecular mod-
eling studies were carried out with program suites O (21) and
X-PLOR (22).
Determination of Specific Activity of Human GSTs to-
ward JS-K. Purified preparations of recombinant hG-
STA1-1, hGSTM1-1, and hGSTP1-1 were obtained from
Panvera (Madison, WI). The activity of human GST toward
CDNB was determined, as described by Habig et al. (10),
before activity measurements with JS-K to ensure that the
enzyme preparations were catalytically active. For activity
measurement toward JS-K, the reaction mixture in a final
volume of 1 ml contained 100 mM potassium phosphate
buffer (pH 6.5), 1 mM GSH, 0.045 mM JS-K, and an appro-
Fig. 1. Structures of JS-K, our control arylating agent CDNB, and theircommon GSH conjugate (DNP-SG). The structures of the Meisenheimercomplexespresumed to be intermediates in the two conjugation reactionsare also shown. 4-Carbethoxy-PIPERAZI/NO (the coproduct of the GSH/
JS-K reaction) generates NO spontaneously at physiological pH asshown; its half-life for NO release at 37C in 0.1 M phosphate buffer (pH7.4) was found to be 6.0 min (k 1.9 103 s1).
410 Antitumor Effects of JS-K, a Nitric Oxide Prodrug
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
4/10
priate amount of human GST isoenzyme protein. The reac-
tion was started by the addition of JS-K, and the rate of
reaction was monitored by measuring decrease in absorb-
ance of JS-K at 298 nm because of its reaction with GSH.
The specific activity toward JS-K was calculated using an
extinction coefficient of 18 mM1
cm1
at 298 nm.Measurement of NO Release. Chemiluminescence de-
tection and quantification of NO evolving from the reactions
of JS-K were conducted using an NO-specific Thermal En-
ergy Analyzer (Model 502A; Thermedics, Analytical Instru-
ment Division, Waltham, MA) essentially as described previ-
ously (23). Briefly, pH 7.4 phosphate buffer containing 1 mM
GSH was sparged with inert gas until a steady detector
response was established. Where indicated, GSTs were
added to a final concentration of 1.67 g of enzyme/ml. The
NO release profile was followed at 37C for 45 min after
injecting JS-K at a final concentration of 133 nM to start the
reaction. The resulting curve was integrated to quantify the
amount of NO released/mol of compound.Cell Lines and Culture Conditions. HL-60, DLD1, and
U937 cells were from American Type Culture Collection (Ma-
nassas, VA). Meth A cells were from Dr. Wolfram Samlowski
(University of Utah). The PPC-1 cell line was provided by Dr.
Graeme Bolger (University of Alabama). For the cell growth
and apoptosis experiments, cells were cultured at a density
of 150,000 cells/ml in RPMI 1640 with 10% fetal bovine
serum at 37C in a 5% CO2-humidified atmosphere. Agents
were added at the indicated concentrations 24 h after culture
initiation. At the indicated time intervals, cells were harvested
and washed twice in PBS before processing for analysis of
growth, differentiation, and apoptosis.
Cell Growth, Differentiation, and Apoptosis Assays.
The number of viable cells was determined using the MTT
assay according to the manufacturers protocol (Promega,
Madison, WI) or using a Coulter counter. Differentiation was
evaluated using Wright and NSE staining of cells collected on
microscope slides by cytospin as described previously (24).
Apoptosis was assayed by flow cytometry and by determin-
ing DNA fragmentation using agarose gel electrophoresis as
described previously (25). For the flow cytometry assay, we
used the propidium iodide staining method of Nicoletti et al.
(26).
In Vivo Studies of JS-K. NOD/SCID mice were bred and
maintained at the Huntsman Cancer Institute at the Univer-
sity of Utah. Experiments were performed on male or female
mice 6 8 weeks of age at the Animal Care Facility of the Salt
Lake City Veterans Administration Medical Center after ap-
proval by the Institutional Animal Care and Use Committees.
We measured systolic blood pressure on unanesthetized
NOD/SCID mice using an occluding tail cuff and a pulse
transducer connected to a blood pressure transducer/mon-
itor from World Precision Instruments (Sarasota, FL). Signals
from the blood pressure monitor and pulse transducer were
transmitted to a MacLab2 data acquisition device (pur-
chased from Stoelting, Wood Dale, IL) that feeds directly into
a Macintosh computer. The recorded data were analyzed
using the Chart data analysis software purchased from
Stoelting. Measurements were done in triplicate at each time
point.
To study the in vivo antineoplastic potency of JS-K, NOD/
SCID mice received injections in the flanks s.c. with HL-60 or
PPC-1 cells (2.5 106 cells/flank). When s.c. tumors were
palpable, treatment with JS-K or an equal volume of vehicle
(20% DMSO in PBS) was started using the indicated doses
and route. Tumor size was measured daily or every other dayusing a Vernier caliper. Tumor volume was calculated using
the formula: width length [(width length)/2] 0.5236.
Fifteen to 20 days after tumor cell implantation, animals were
sacrificed by CO2
inhalation, and tumors were collected for
histochemical analysis.
Histological Analysis of Tumors. At the completion of the
experiments, animals were sacrificed, and s.c. tumors were
dissected out, fixed in 10% formaldehyde, and imbedded in
paraffin. Four-m sections were cut and stained with H&E.
Calculations and Statistical Analysis. Results are ex-
pressed as averages of multiple experiments with SE. SE
was calculated as the SD of different measurements divided
by the square root of the number of measurements. Differ-ences were considered statistically significant if the P was
0.05 as calculated using the t test.
Results
Reactivity of JS-K with GSH. The diazeniumdiolate ion has
been judged to resemble chloride as a leaving group in SNAr
reactions (7). Because CDNB (structure shown in Fig. 1) is
known to react with GSH, we anticipated that JS-K would be
similarly converted to DNP-SG (structure also shown in Fig.
1). This was confirmed by HPLC-MS; as shown in Fig. 2A, an
85% conversion of JS-K to DNP-SG occurred within a 30-
min incubation period at 37C in pH 7.4 phosphate buffer.
Similar results were seen in RPMI 1640 cell culture. Pseudo-first-order kinetic plots for the reaction of GSH with JS-K in
0.1 M phosphate buffer (pH 7.4) were obtained with GSH (15
mM) in large excess of the substrate. Excellent first-order
behavior was observed over several half-lives and measured
first-order rate constants showed a linear dependence on
(GSH; Fig. 2C). The slope and y intercept of the line yielded
values for the second-order [k2 (1.02 0.04) M1 s1] and
first-order [k1 (4 12) 10
5 s1] rate constants for the
reactions of JS-K with GSH and water, respectively, at 37C
in 0.1 M phosphate buffer, pH 7.4. The UV spectral changes
accompanying the reaction in a second cell culture medium
(DMEM) are shown in Fig. 2B.
Hydrolysis of JS-K. JS-K proved resistant to simple hy-
drolysis under these conditions, as reflected in the near-zero
y intercept of Fig. 2C. The value of k1
obtained from the
intercept is in statistical agreement with the rate constant for
JS-K hydrolysis (1 106 s1) measured separately in the
absence of GSH. The small amount of the hydrolysis product
(2,4-dinitrophenol) seen in the chromatogram of Fig. 2A was
apparently formed in the DMSO stock solution, which in this
case had been stored in the refrigerator with intermittent use
during several weeks. Hydrolysis was much more facile at pH
12, which was expected for a compound type designed to be
activated for NO release by nucleophilic attack.
Solubility Limits. JS-K showed an interesting but poten-
tially nettlesome tendency to remain supersaturated in aque-
ous solutions, only to separate from the aqueous phase at a
411Molecular Cancer Therapeutics
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
5/10
time course that was difficult to predict or control. For ex-
ample, our first attempts to follow the hydrolysis of 50 M
JS-K spectrophotometrically sometimes showed little
change, as expected from Fig. 2C, but at other times re-
vealed a variable and sometimes rapid rate of absorbance
loss without production of new maxima. The problem couldbe overcome by cleaning the cuvette with nitric acid before
filling it with buffer then adding the DMSO/JS-K stock solu-
tion, in which case absorbance changed little with time.
However, rinsing this solution out with distilled water fol-
lowed by refilling the cuvette with buffer and adding the
stock solution often led to shrinking absorbance with con-
comitant appearance of cloudiness in the cuvette; dissolu-
tion of the insoluble material, after isolating it by centrifuga-
tion, showed that the resulting solid (dissolved in acetonitrile)
was essentially pure JS-K in an amount equivalent to the
quantity lost from the supernatant. Quantification of the JS-K
in the buffer by HPLC and UV spectrophotometry after all
spectral changes had stopped indicated that JS-Ks solubil-ity limit in 0.1 M phosphate buffer at 37C and pH 7.4 con-
taining 1% DMSO was 10 M. The relative insolubility of
JS-K in aqueous media should be taken into account during
any experimental work involving dilution of organic JS-K
solutions with aqueous media because the observed tend-
ency toward supersaturation can lead to JS-K concentra-
tions that are much lower or higher than expected and thus
to erroneous results.
Catalysis of the GSH/JS-K Reaction and NO Release
by GST. The reaction of GSH with CDNB (Fig. 1) is catalyzed
by several classes of human GSTs, and, thus, this electro-
philic substrate is often used for quantifying their activity.
Given the similarity of diazeniumdiolate ions to chloride as a
leaving group in SNAr reactions (7), we expected JS-K also to
undergo GST-catalyzed conjugation with GSH. To gain in-
sights into the effect the obvious steric differences between
chloride and the diazeniumdiolate ion shown in Fig. 1 may
imply, we modeled the accommodation of the Meisenheimer
complex of JS-K in the active sites of the three major classes
of human GSTs, i.e., hGSTA1-1, hGSTM1-1, and hGSTP1-1.
As illustrated in Fig. 3, AC, both hGSTA1-1 and hGSTM1-1
classes of GSTs accommodate the Meisenheimer complex
very well, but hGSTP1-1 appears to have serious steric con-
flicts with the diazeniumdiolate moiety of the transition state
complex. On the basis of molecular modeling, we predicted
that hGSTA1-1 and hGSTM1-1 should be more effective
than hGSTP1-1 for catalyzing the GSH conjugation of JS-K.These predictions were confirmed by determining the activ-
ities of recombinant hGSTA1-1, hGSTM1-1, and hGSTP1-1
preparations toward JS-K. The data are summarized in Table 1.
Specific activities of GSTs toward CDNB were determined be-
fore activity measurement with JS-K to ensure that the enzyme
preparations were catalytically active. The specific activities of
the GSTs toward CDNB were comparable with the values pub-
lished in the literature (Ref. 27; Table 1). In agreement with our
max 354 nm, rather than DNP-SG, max 340 nm. C, increases in JS-Kconsumption rate at an initial concentration of 50 M in 0.1 M phosphate
(pH 7.4) at 37C as a function of increasing GSH concentration.
Fig. 2. Reactions of JS-K with thiols. A, HPLC traces of a reactionmixture containing 50 M JS-K and 500 M GSH at 37C in 0.1 M phos-phate (pH 7.4) at an observing wavelength of 300 nm immediately aftermixing (top) and 30 min later (bottom). The structures of JS-K and its GSHconjugate DNP-SG (extinction coefficients at 300 nm of 16 and 3.7 m M1
cm1, respectively) are shown in Fig. 1; 2,4-DNP is the JS-K hydrolysisproduct, 2,4-dinitrophenol, and DMSO is the cosolvent. B, UV spectralchanges during 30 min at 37C of a 50 M JS-K solution in serum-freeDMEMin the absence (left) and presence (right) of 1 mM GSH. Note that in this
cystine-containing medium, the product is S-(2,4-dinitrophenyl)cysteine,
412 Antitumor Effects of JS-K, a Nitric Oxide Prodrug
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
6/10
prediction, hGSTP1-1 was much less active than hGSTA1-1 or
hGSTM1-1 for GSH conjugation of JS-K.
That these reactions led to NO generation as predicted
was demonstrated by purging gases from the solution as
they formed into an NO-specific chemiluminescence detec-
tor. The results are shown in Fig. 3D. Consistent with JS-Ks
resistance to hydrolysis but its reactivity toward GSH as
noted above, no NO was detected until 1 mM GSH was
added to the pH 7.4 phosphate/0.67 M JS-K solution. In the
absence of enzyme, NO release began immediately upon
adding GSH, increasing in rate until plateauing at 8 min, and
integrating to a total of 1.1 mol of NO/mol of JS-K within 43
min of mixing. The hGSTP1-1 isoform catalyzed this reac-
tion, but only weakly; that this was not attributable to a
deleterious effect of JS-K exposure on the enzymes activity
was demonstrated in its unfettered ability, in the presence of
up to 80 M JS-K, to catalyze CDNBs conjugation with GSH.
The hGSTA1-1 and hGSTM1-1 isoforms proved much supe-rior to hGSTP1-1 as catalysts for JS-K conjugation. The data
of Fig. 3D on NO release are thus consistent with the
conclusion from the JS-K consumption studies of Table 1
that JS-K is metabolized much better by hGSTA1-1 and
hGSTM1-1 than by hGSTP1-1.
Growth Inhibitory Properties of JS-K. Fig. 4A shows a
comparison between the growth inhibitory ability of JS-K and
those of the chemotherapeutic agents daunorubicin and eto-
poside in our HL-60 assay system. The IC50
s of JS-K, dauno-
rubicin, and etoposide were 0.5, 0.01, and 0.3 M, respec-
tively. CDNB, a compound with the same aryl ring as JS-K
that does not release NO (Fig. 1), inhibited the in vitro growth
of HL-60 cells but at much higher concentrations, with anIC
50estimated at 6.7 M (data not shown). JS-K also inhib-
ited the growth of U937 (monocytic leukemia) cells with an
IC50 of 0.3 M (data not shown). Solid tumor cell growth was
also inhibited by JS-K, although to a lesser extent than
leukemia cells (Fig. 5); the IC50
s for the three lines we tested,
PPC-1, DLD-1, and Meth A, were an order of magnitude
greater than those for the two leukemia lines.
To determine whether modulation of the GST pathway
affects JS-Ks antineoplastic properties, we performed ex-
periments using NAC or BSO. NAC increases intracellular
GSH levels, whereas BSO inhibits its synthesis (28). Treat-
ment of HL-60 cells with NAC (0.3 0.5 mM) or BSO (0.2 0.3
mM) did not significantly affect cell growth. Pretreatment of
Fig. 3. Catalysis of JS-Ks NO release by GST. AC illustrate the accommodationof the Meisenheimer complex of JS-K in the active center of (A) hGSTA1-1, (B)hGSTM1-1, and (C) hGSTP1-1 as predicted by the molecular modeling studies. Theactive centers of the GSTs are shown as surface representations and the Meisen-heimer complexes as ball-and-stick models with atomic color scheme (carbon, gray;nitrogen, blue; oxygen, red; and sulfur, yellow). The illustration was prepared usingGrasp (29) and Raster3D (30). D, chemiluminescence traces showing the time courseof NO release from 0.67 M JS-K at 37C in 0.1 M phosphate (pH 7.4) containing 1.0mM GSH both alone and with 1.67 g/ml hGSTA1-1, hGSTM11, or hGSTP11.
Table 1 Specific activities (mol min1 mg1) of GST-catalyzed
glutathionylation of CDNB and JS-K in 0.10 M phosphate buffer at pH6.5 and 25Ca
Enzyme CDNB JS-K
hGSTA1-1 34 3.7
hGSTM1-1 200 14.9
hGSTP1-1 (I104, A113) 95 0.15
a The concentrations of CDNB and JS-K were 1.0 and 0.045 mM, respec-tively. The GSH concentration was 1.0 mM for both substrates. The proteinconcentrations were 0.025, 0.01, and 0.215 mg/ml for hGSTA1-1,hGSTM1-1, and hGSTP1-1, respectively.
413Molecular Cancer Therapeutics
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
7/10
the cells for 2 6 h with NAC prevented the JS-K-induced
growth inhibition. Pretreatment of HL-60 cells with BSO for
2 6 h did not prevent the JS-K-induced growth inhibition,
whereas pretreatment of the cells with BSO for 24 h en-
hanced it (data not shown).
Induction of Leukemia Cell Apoptosis by JS-K. Be-
cause we had previously shown that spontaneous and
esterase-activated NO generators induce apoptosis in leu-
kemia cells (2, 25), we sought to determine whether JS-K had
a similar effect. Three days after addition of JS-K at concen-
trations of 0.5 and 1 M, the percentage of apoptotic HL-60
cells increased from 7 to 27 and 43%, respectively (Fig. 4B).
The flow cytometry experiments were confirmed with DNA
laddering assays (Fig. 4C). The pan-caspase inhibitor C-
VAD-FMK prevented the JS-K-induced growth inhibitory and
apoptotic effects (Fig. 4D).
Effect of JS-K on Leukemia Cell Differentiation. We
have previously shown that NO induces HL-60 cells to dif-
ferentiate along the monocytic phenotype (24). We therefore
determined whether JS-K induced differentiation as well.
HL-60 cells were treated with JS-K at a concentration of 0.5
M f or 35 days. Wright stain revealed morphological
changes consistent with a monocytic phenotype, namely
development of folded nuclei, large cytoplasms, and cyto-
plasmic vacuoles (Fig. 6). NSE (an enzyme specific to the
monocytic lineage) staining showed that JS-K increased the
percentage of HL-60 cells expressing NSE from 1 to 40%
(Fig. 6).
Fig. 4. Antiproliferative and proapoptotic effects of JS-K on HL-60 leukemia cells in vitro. A, comparison of the antiproliferative effects of JS-K with thoseof other chemotherapeutic agents. Cells were cultured with JS-K, daunorubicin, or etoposide for 3 days at the indicated concentrations. Cell viability wasdetermined using the MTT assay (averages and SE of three separate experiments). B, apoptosis induction by JS-K. Cells were cultured with JS-K at theindicated concentrations. At 72 h, the percentage of apoptotic cells was determined by flow cytometry (see text; averages and SE of three separateexperiments). C, DNA laddering in HL-60 cells resulting from 3-day exposure to 1 M JS-K (gel representative of three different experiments; MWM,molecular weight marker). D, reversal by the caspase inhibitor C-VAD-FMK of JS-Ks cytostatic and proapoptotic effects on HL-60 cells. Cells were culturedwith JS-K (0.75 M), C-VAD-FMK (50 M), or the combination for 3 days. Cell viability and apoptosis were determined using the MTT assay and flowcytometry, respectively (averages and SE of three separate experiments).
414 Antitumor Effects of JS-K, a Nitric Oxide Prodrug
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
8/10
In Vivo Effects of JS-K. Before determining in vivo an-
tineoplastic activity of JS-K, we next sought to identify a
dose that would not induce significant hypotension in NOD/
SCID mice. Using i.v. administration, we were able to esca-
late the dose of JS-K up to 4 mol/kg without observing
significant hypotension, whereas doses of 5 mol/kg or
higher had a notable hypotensive effect (data not shown).
Using estimates of the mouse blood volume, 4 mol/kg of
JS-K would be expected to yield peak blood levels of 17
M, which is far above its in vitro IC50
.
NOD/SCID mice were then implanted s.c. in the flanks with2.5 106 HL-60 cells/flank. When the tumors were palpable,
treatment was started with JS-K administered i.v. via the tail
vein three times/week at a dose of 4 mol/kg. Control mice
received an equal volume of vehicle through the same route
and according to the same schedule. JS-K treatment in-
duced a significant inhibition of in vivo leukemia cell growth
(Fig. 7A). Sixteen days after starting therapy, the average
tumor volumes in control and JS-K-treated mice were 8.34
0.72 and 3.13 1.14 cm3 (P 0.039), respectively, reflecting
a 50% reduction in tumor volume in treated mice. Histo-
logical analysis of HL-60 cell tumors obtained from vehicle-
treated mice revealed a uniform population of densely
packed myeloblasts. The cells were highly invasive, pene-
trating the surrounding tissues, and showing high mitotic
activity. Necrosis was minimal. On the other hand, histolog-
ical analysis revealed extensive (50%) cell necrosis in
HL-60 cell tumors obtained from the JS-K-treated mice as
compared with 10% in controls (Fig. 7B).
To determine whether JS-K inhibits thein vivo growth of solid
tumor cells, NOD/SCID mice were implanted with 2.5 106
PPC-1 (prostate carcinoma) cells and treated with 4 mol/kg
JS-K or an equal volume of vehicle i.v. three times/week. Similar
to the observation with HL-60 cells, JS-K treatment inhibited
the growth of PPC-1 cells in vivo. Nineteen days after start of
therapy, s.c. tumor implant volumes were 0.368 0.082 and
0.107 0.053 cm3 (P 0.0073) in vehicle and JS-K-treated
animals, respectively (Fig. 7C). Similar to HL-60 cells, PPC-1
cells were highly aggressive and invaded the surrounding tis-
sues. Histological analysis revealed extensive tumor necrosis in
implants obtained from JS-K-treated animals (Fig. 7D).
Discussion
The results suggest that JS-K might serve as a promising
lead compound in the search for new classes of antineoplas-
tic agents with novel mechanisms of action. In a test of its
antiproliferative effect on the HL-60 human myeloid leukemia
cell line, its consistently submicromolar IC50 (three separatedeterminations of 0.2, 0.5, and 0.5 M) compared favorably
with those of the clinical antileukemics daunorubicin (0.01
M) and etoposide (0.3 M). Chemical characterization of
JS-K showed that it resists hydrolysis in the absence of
strong nucleophiles but generates copious NO upon reaction
with GSH. The experiments additionally confirmed predic-
tions based on molecular models of transition state geometry
that two of three major human isoforms of GST could
strongly catalyze NO release from JS-K.
Consistent with our hypothesis that such an NO prodrug
might direct its toxic action selectively to the NO-sensitive
malignant cells while sparing less sensitive normal tissues,
JS-K injected i.v. three times/week inhibited the growth of
s.c. tumors formed from HL-60 xenografts in NOD-SCID
mice by approximately one-half. This suggests that JS-K
may serve as a useful lead for developing therapies against
leukemia.
As far as mechanism of action is concerned, the in vitro
experiments showed all of the characteristics of previous
studies with structurally different NO donors (including in-
duction of both apoptosis and differentiation in HL-60 cells)
and were thus consistent with NO-induced cytolysis. How-
ever, it is important to recognize that other pathways may
also be operative. For example, arylation of thiol groups in
critical protein residues could be contributing to the toxicity;
however, this may be a minor effect, as CDNB (which trans-
fers the same 2,4-dinitrophenyl group to cellular nucleophiles
Fig. 6. Effect of JS-K on differentiation of leukemia cells. HL-60 cellswere cultured with (right) and without (left) 0.5 M JS-K for a period of 5days. Differentiation was assessed morphologically (Wright stain, top) and
by determining NSE (bottom) expression (photomicrographs of one ex-periment representative of three).
Fig. 5. Inhibitory effect of JS-K on three different solid tumor cell lines.PPC-1 (prostate), DLD-1 (colon), and Meth A (mammary) cells were cul-tured with JS-K at the indicated concentrations for 3 days. Cell viabilitywas determined using the MTT assay (averages and SE of three separate
experiments).
415Molecular Cancer Therapeutics
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
9/10
as JS-K but without generating any NO) was an order of
magnitude less potent in the HL-60 screen. Whether there is
synergy between the growth inhibitory effects of NO and
arylation remains to be determined. Transcarbamoylation is
also a possible route to toxicity. The finding that N-acetyl-L-
cysteine prevented the growth inhibitory effects of JS-K
while BSO enhanced them suggests that one or more reac-
tive nitrogen/oxygen species (NO and/or the product of NO s
reaction with other reactive oxygen species) play a key role
in effecting the antineoplastic activity of JS-K.
As to the speci fic molecular targets that may be in-
volved, exposure to JS-K can in principle render any crit-
ical thiol group nucleophilically unreactive by S-nitrosation
or S-(2,4-dinitrophenyl)ation. It is likely that the cysteine-
containing caspases are directly or indirectly among the
mechanistically important targets of JS-Ks action be-
cause the nonspecific caspase inhibitor C-VAD-FMK re-
duced JS-Ks cytostatic and apoptotic effects on HL-60
cells. Other pathways are possible, though, and a fuller
explanation of the drugs antineoplastic activity must
await the outcome of future hypothesis testing.
Whatever the mechanism(s) that may be involved, the
present results support the choice of JS-K as a worthy
lead compound for additional drug discovery efforts. The
significant in vivo activity demonstrated in Fig. 7 was seen
in first generation experiments that could hardly be viewed
as optimized; the dose regimen was chosen based on
preliminary studies of bolus sizes that could be adminis-
tered without inducing systemic hypotension or other
manifestations of toxicity. It is entirely possible that a
continuous administration schedule would be much more
effective. Structural modification aimed at improving the
problematic solubility of this compound could beneficially
affect its absorption, distribution, and transport proper-
ties. With substantial efforts underway in other laborato-
ries4 as well as our own aimed at further characterizing
and refining the chemotherapeutic potential of this inter-
esting lead, we are hopeful that our ultimate goal of intro-
4 L. Jia; K. Tew and V. Findlay; J. Liu and M. Waalkes, personal commu-
nications.
Fig. 7. Effects of i.v. JS-K on growth of both human leukemia and solid tumor xenografts in NOD-SCID mice. A, growth rate of tumors produced when2.5 106 HL-60 cells were implanted s.c. in NOD/SCID mice [3 animals each for JS-K and vehicle treatments]. When tumors became palpable, JS-K wasadministered i.v. at a dose of 4 mol/kg three times/week. Tumor volumes were measured at regular intervals. At the time of experiment termination, tumor
implants were dissected out and analyzed histologically. Differences between vehicle and JS-K values from days 9 16 were significant at the P 0.05 level.B, JS-K induced extensive necrosis in these tumors (photomicrograph from 1 animal representative of 3; H&E stain, 100 magnification). C, growth rateof tumors produced when 2.5 106 human prostate carcinoma (PPC-1) cells were implanted s.c. in NOD/SCID mice (5 animals each for JS-K and vehicletreatments) and followed as in A above. Differences between vehicle and JS-K values were significant at the P 0.05 level for all time points after day 1.D, JS-K induced extensive necrosis in these tumors (photomicrograph from 1 animal representative of 5; H&E stain,100 magnification).
416 Antitumor Effects of JS-K, a Nitric Oxide Prodrug
Research.on March 22, 2013. 2003 American Association for Cancermct.aacrjournals.orgDownloaded from
http://mct.aacrjournals.org/http://mct.aacrjournals.org/http://mct.aacrjournals.org/ -
7/28/2019 JS-K, a Glutathione/Glutathione S-Transferase-activated Nitric Oxide Donor of the Diazeniumdiolate Class with Pote
10/10
ducing new and improved treatments for AML and other
malignant diseases will be significantly advanced.
References
1. Greer, J. P., Baer, and Kinney, M. C. In: G. R. Lee, J. Foerster, J.
Lukens, F. Paraskevas, J. P. Greer, and G. M. Rodgers (eds.), Wintrobes
Clinical Hematology, Ed. 10, pp. 22722319. Baltimore: Williams and
Wilkins, 1999.
2. Saavedra, J. E., Shami, P. J., Wang, L. Y., Davies, K. M., Booth, M. N.,
Citro, M. L., and Keefer, L. K. Esterase-sensitive nitric oxide donors of the
diazeniumdiolate family. In vitro antileukemic activity. J. Med. Chem., 43:
261269, 2000.
3. Mooradian, D. L., Hutsell, T. C., and Keefer, L. K. Nitric oxide (NO)
donor molecules: effect of NO release rate on vascular smoothmusclecell
proliferation in vitro. J. Cardiovasc. Pharmacol., 25: 674 678, 1995.
4. Saavedra, J. E., Billiar, T. R., Williams, D. L., Kim, Y-M., Watkins, S. C.,
and Keefer, L. K. Targeting nitric oxide (NO) delivery in vivo. Design of a
liver-selective NO donor prodrug that blocks tumor necrosis factor--
induced apoptosis and toxicity in the liver. J. Med. Chem., 40: 19471954,
1997.
5. Bico, P., Chen, C. Y., Jones, M., Erhardt, J., and Dirr, H. Class piglutathione S-transferase: Meisenheimer complex formation. Biochem.
Mol. Biol. Int., 33: 887 892, 1994.
6. Sargent,J. M., Williamson, C., Hall, A.G., Elgie, A.W., and Taylor, C. G.
In: G. J. L. Kaspers, R. Pieters, and A. J. P. Veerman (eds.), Drug Resist-
ance in Leukemia and Lymphoma, III, pp. 205 209. New York: Kluwer
Academic/Plenum Publishers, 1999.
7. Saavedra, J. E., Srinivasan, A., Bonifant, C. L., Chu, J., Shanklin, A. P.,
Flippen-Anderson, J. L., Rice, W. G., Turpin, J. A., Davies, K. M., and
Keefer, L. K. The secondary amine/nitric oxide complex ion R2N[N(O)NO]-
as nucleophile and leaving group in SNAr reactions. J. Org. Chem., 66:
3090 3098, 2001.
8. Saavedra, J. E., Booth, M. N., Hrabie, J. A., Davies, K. M., and Keefer,
L. K. Piperazine as a linker for incorporating the nitric-oxide releasing
diazeniumdiolate group into other biomedically relevant functional mole-
cules. J. Org. Chem., 64: 5124 5131, 1999.
9. Mancini, I., Guella, G., Chiasera, G., and Pietra, F. Glutathione S-
transferase catalysed dehalogenation of haloaromatic compounds which
lack nitro groups near the reaction centre. Tetrahedron Lett., 39: 1611
1614, 1998.
10. Habig, W. H., Pabst, M. J., and Jakoby, W. B. Glutathione S-trans-
ferases. The first enzymatic step in mercapturic acid formation. J. Biol.
Chem., 249: 7130 7139, 1974.
11. Ji, X., Armstrong, R. N., and Gilliland, G. L. Snapshots along the
reaction coordinate of an SNAr reaction catalyzed by glutathione trans-
ferase. Biochemistry, 32: 12949 12954, 1993.
12. Prade, L., Huber, R., Manoharan, T. H., Fahl, W. E., and Reuter, W.
Structures of class pi glutathione S-transferase from human placenta in
complex with substrate, transition-state analogue and inhibitor. Structure
(Lond.), 5: 12871295, 1997.
13. Powell, M. J. D. Restart procedures for the conjugate gradient
method. Math. Prog., 12: 241254, 1977.
14. Armstrong, R. N. Glutathione S-transferases: structure and mecha-
nism of an archetypical detoxication enzyme. Adv. Enzymol. Relat. Areas
Mol. Biol., 69: 1 44, 1994.
15. Wilce, M. C. J., and Parker, M. W. Structure and function of glutathi-
one S-transferases. Biochim. Biophys. Acta, 1205: 118, 1994.
16. Dirr, H., Reinemer, P., and Huber, R. X-ray crystal structures of
cytosolic glutathione S-transferases. Implications for protein architecture,
substrate recognition and catalytic function. Eur. J. Biochem., 220: 645
661, 1994.
17. Armstrong, R. N. Structure, catalytic mechanism, and evolution of the
glutathione transferases. Chem. Res. Toxicol., 10: 218, 1997.
18. Engh, R. A., and Huber, R. Accurate bond and angle parameters for
X-ray protein structure refinement. Acta Crystallogr., 47: 392 400, 1991.
19. Sinning, I., Kleywegt, G. J., Cowan, S. W., Reinemer, P., Dirr, H. W.,
Huber, R., Gilliland, G. L., Armstrong, R. N., Ji, X., Board, P. G., Olin, B.,
Mannervik, B., and Jones, T. A. Structure determination and refinement of
human alpha class glutathione transferase A1-1, and a comparison with
the mu and pi class enzymes. J. Mol. Biol., 232: 192212, 1993.
20. Cameron, A. D., Sinning, I., LHermite, G., Olin, B., Board, P. G.,
Mannervik, B., and Jones, T. A. Structural analysis of human alpha-class
glutathione transferase A1-1 in the apo-form and in complexes with
ethacrynic acid and its glutathione conjugate. Structure (Lond.), 3: 717727, 1995.
21. Jones, T. A., and Kjeldgaard, M. Electron-density map interpretation.
Methods Enzymol., 277: 173208, 1997.
22. Brunger, A. T., and Rice, L. M. Crystallographic refinement by simu-
lated annealing: methods and applications. Methods Enzymol., 277: 243
269, 1997.
23. Keefer, L. K., Nims, R. W., Davies, K. M., and Wink, D. A. NONOates
(1-substituted diazen-1-ium-1,2-diolates) as nitric oxide donors: conven-
ient nitric oxide dosage forms. Methods Enzymol., 268: 281293, 1996.
24. Magrinat, G., Mason, S. N., Shami, P. J., and Weinberg, J. B. Nitric
oxide modulation of human leukemia cell differentiation and gene expres-
sion. Blood, 80: 1880 1884, 1992.
25. Shami, P. J., Sauls, D. L., and Weinberg, J. B. Schedule and concen-
tration-dependent induction of apoptosis in leukemia cells by nitric oxide.
Leukemia (Baltimore), 12: 14611466, 1998.
26. Nicoletti, I., Migliorati, G., Pagliacci, M. C., Grignani, F., and Riccardi,
C. A rapid and simple method for measuring thymocyte apoptosis by
propidium iodide staining and flow cytometry. J. Immunol. Methods, 139:
271279, 1991.
27. Hayes, J. D., and Pulford, D. J. The glutathione S-transferase super-
gene family: regulation of GST and the contribution of the isoenzymes to
cancer chemoprotection and drug resistance. Crit. Rev. Biochem. Mol.
Biol., 30: 445 600, 1995.
28. Anderson, M. E. Glutathione: an overview of biosynthesis and mod-
ulation. Chem. Biol. Interact., 111112: 114, 1998.
29. Nicholls, A., Sharp, K. A., and Honig, B. Protein folding and associ-
ation: insights from the interfacial and thermodynamic properties of hy-
drocarbons. Proteins Struct. Funct. Genet., 11: 281296, 1991.
30. Merritt, E. A., and Bacon, D. J. Raster3D: photorealistic molecular
graphics. Methods Enzymol., 277: 505524, 1997.
417Molecular Cancer Therapeutics