Inhibition of human glutathione transferases by dinitronaphthalene derivatives

Article abstract of DOI:10.1016/j.abb.2014.06.002

Glutathione transferase (GST) enzymes catalyze the conjugation of glutathione with reactive functional groups of endogenous compounds and xenobiotics, including halonitroaromatics. 1-Chloro-2,4-dinitrobenzene (CDNB) is one of the most commonly used substr

Full text of DOI:10.1016/j.abb.2014.06.002

Archives of Biochemistry and Biophysics 555–556 (2014) 71–76
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Inhibition of human glutathione transferases by dinitronaphthalene
derivatives
Hilary Groom a, Moses Lee b,1, Pravin Patil b,2, P. David Josephy a,
⇑
a Department of Molecular and Cellular Biology, University of Guelph, Guelph, Ontario N1G 2W1, Canada
b Department of Chemistry, Hope College, Holland, MI 49423, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 April 2014
and in revised form 29 May 2014
Available online 12 June 2014
Glutathione transferase (GST) enzymes catalyze the conjugation of glutathione with reactive functional
groups of endogenous compounds and xenobiotics, including halonitroaromatics. 1-Chloro-2,
4-dinitrobenzene (CDNB) is one of the most commonly used substrates for GST activity assays. We have
studied the interactions of dinitronaphthalene analogues of CDNB with recombinant human GST enzymes
(Alpha, Mu, and Pi classes) expressed in Escherichia coli. Dinitronaphthalene derivatives were found to be
GST inhibitors. The highest potency of inhibition was observed towards Mu-class GSTs, M1-1 and M2-2;
IC50 values for 1-methoxy- and 1-ethoxy-2,4-dinitronaphthalene were in the high nanomolar to low
micromolar range. Inhibition accompanies the formation, at the enzyme active site, of very stable
Meisenheimer complex intermediates.
Keywords:
Glutathione transferase
1-Chloro-2,4-dinitronaphthalene
Meisenheimer complex
Ó 2014 Elsevier Inc. All rights reserved.
Introduction
chemotherapeutic agents that are detoxified by glutathione (GSH)
conjugation [7]. GST P1-1, an enzyme that is frequently overexpres-
sed in tumors, has received particular attention as a target [8,9]. Bet-
ter understanding of the interactions of small molecules with the
GST active site will facilitate drug discovery, as has been discussed
by Wu and Dong [2].
Glutathione transferase (GST)3 enzymes catalyze reactions of the
endogenous nucleophile glutathione with a wide range of electro-
philes, including arene oxides, epoxides, alkyl halides, quinones,
and a,b-unsaturated carbonyls [1–3]. The diversity of GST substrates
is partly accounted for by the existence of many different GST
enzymes: at least 17 cytosolic GSTs are found in humans, in classes
Alpha, Mu, Omega, Pi, Sigma, Theta, and Zeta [2]. A second aspect of
GST substrate diversity is that each GST enzyme can accept a range
of substrates – so-called ‘‘catalytic promiscuity’’ [4–6]. Since the cor-
respondence between GST substrates and enzymes is many-to-many
rather than one-to-one, it would be desirable to identify both ‘‘uni-
versal’’ (active with any form) and ‘‘specific’’ (active with only one
form) GST substrates and GST inhibitors. Specific inhibitors might
be used, for example, to counteract the resistance of tumor cells to
1-Chloro-2,4-dinitrobenzene (CDNB) is
a
substrate for
most forms of GST; its reaction with glutathione (GSH) yields
2,4-dinitrophenyl glutathione and chloride ion. The use of CDNB as
a GST substrate was introduced by Clark and colleagues, who wrote
that ‘‘it is tempting to regard this compound as a general substrate
which may function more or less effectively with any GSH transfer-
ase’’ [10]. CDNB did not prove to be a truly universal substrate
(having undetectable activity with human GST T1-1, for example
[11]), but it is commonly used for assaying GST activity [12,13].
The mechanism of the reaction of CDNB with GSH is nucleo-
philic aromatic substitution (SNAr) via an anionic Meisenheimer
complex intermediate [14–16]. When 1,3,5-trinitrobenzene
(TNB), a CDNB analogue that lacks a leaving group, is added to
rat liver GST 3-3 (M1-1) or GST 4-4 (M2-2) [17] in the presence
of GSH, a stable red complex of the enzyme with 1-(S-glutathio-
nyl)-2,4,6-trinitro-cyclohexadienate (a ‘‘dead-end’’ Meisenheimer
intermediate) is formed. The complex can be isolated [14] and its
crystal structure has been obtained [18] as has that of the corre-
sponding complex with recombinant human GST P1-1 [19]. The
TNB-GSH Meisenheimer complex is an inhibitor of GST activity at
⇑
Corresponding author. Fax: +1 (519) 837 1802.
E-mail address: djosephy@uoguelph.ca (P.D. Josephy).
Present address: Department of Chemistry, Georgia State University, Atlanta, GA
1
30303, United States.
2
Present address: Department of Chemistry, University of Louisville, Louisville, KY
40292, United States.
3
Abbreviations used: CDNB, 1-chloro-2,4-dinitrobenzene; CDNN, 1-chloro-2,4-
dinitronaphthalene; ESI, electrospray ionization; GSH, glutathione; GST, glutathione
transferase; HPLC–MS, high performance liquid chromatography–mass spectrometry;
IPTG, isopropyl b-D-1-thiogalactopyranoside; LB, lysogeny broth; PMSF, phen-
ylmethanesulfonylfluoride; TNB, 1,3,5-trinitrobenzene.
l
M concentrations, as tested with a moth GST [20] and with
http://dx.doi.org/10.1016/j.abb.2014.06.002
0003-9861/Ó 2014 Elsevier Inc. All rights reserved.

72
H. Groom et al. / Archives of Biochemistry and Biophysics 555–556 (2014) 71–76
Table 1
X
X
Y
Dinitronaphthalene compounds used in this study.
-
NO2
NO2
Dinitronaphthalene derivatives
X
Abbreviation
CAS
logP
H–
Cl–
HO–
DNN
CDNN
DNNOH
606-37-1
2401-85-6
605-69-6
55154-12-6*
13772-69-5
15352-94-0
–
2.77
3.23
2.90
–
2.83
3.34
–
NO2
NO2
Fig. 1. Left: General structure of the dinitronaphthalene compounds used in this
study. Right: general structure of dinitronaphthalene Meisenheimer complexes. In
the complexes studied here, Y = glutathione.
CH3O–
CH3CH2O–
Glutathione–
DNNOMe
DNNOEt
DNNSG
The general structure is shown in Fig. 1. The substituents, abbreviations, and
chemical abstracts registry numbers are shown. The last column gives the esti-
mated logP (logarithm of octanol–water partition coefficient) values, taken from
the American Chemical Society (SciFinder) database (calculated using Advanced
Chemistry Development™ Software v. 11.02).
affinity-purified human liver and placenta GST preparations [21].
Recombinant human GST P1-1 also forms a Meisenheimer complex
with TNB [22].
Naphthalene derivatives form Meisenheimer complexes (Fig. 1)
much more readily than do benzene derivatives, due to the greater
delocalization of negative charge in the larger ring system [23,24].
For example, the methoxy complex of 1,3-dinitronaphthalene is
about 104-fold more stable than that of 1,3-dinitrobenzene, as mea-
sured by the equilibrium constants for their formation [25] (p. 128).
The stabilizing effect of the second aromatic ring in naphthalene is
almost as large as that of an additional nitro group (e.g., going from
1,3-dinitrobenzene to TNB) [25]. The facile formation of Meisenhei-
mer intermediates from dinitronaphthalene derivatives suggested
to us that these compounds might be good substrates or inhibitors
of GST enzymes. 1-Chloro-2,4-dinitronaphthalene (CDNN) is the
naphthalene analogue of CDNB [26–28]. The goal of our study was
to examine the interactions of CDNN and other dinitronaphthalene
derivatives with human cytosolic GST enzymes.
*
Martius yellow anion.
neutralized with a few drops of HCl. The solid precipitate that
formed was collected by vacuum filtration and washed sequen-
tially with cold water, ethanol, and ethyl acetate. Air-dried solid
was recrystallized from 5 mL hot ethanol; yield, 22%. No residual
CDNN starting material remained, as determined by HPLC analysis.
ESI–LC–MS gave an (M + H)+ peak at m/z = 524.0, as expected, and
a smaller (M + Na)+ peak at m/z = 546.0.
Expression and purification of human cytosolic GSTs
Recombinant hGST proteins were expressed in E. coli and puri-
fied using a modification of the published protocol [30]. All cul-
tures were grown with vigorous shaking at 37 °C. An aliquot of
overnight culture grown in LB (lysogeny broth) with ampicillin,
0.1 mg per mL, was diluted 100-fold into 2TY (tryptone yeast
extract) medium (1L) with ampicillin, 0.1 mg per mL, in a 2 L flask.
When the culture reached OD600 = 0.5, IPTG (1 mM) was added,
and induction proceeded overnight. Cells were then harvested by
centrifugation and resuspended in lysis buffer (50 mM Tris buffer,
pH 7.4; 100 mM NaCl; 1 mM EDTA; 1 mM b-mercaptoethanol;
12 mL). Cells were lysed as previously described [30]; phen-
ylmethanesulfonylfluoride (PMSF) protease inhibitor, 0.1 mM,
was added after sonication. Lysate enzyme activity was measured
with CDNB (see below) and protein concentration was determined
with the Bradford assay, using BSA as reference standard [31]. GST
purification was performed on a column (2 mL bed volume) of
glutathione-agarose affinity resin pre-equilibrated with cold
phosphate-buffered saline (PBS). To remove non-specific binding
proteins, the column was washed with PBS until no protein was
present in the eluate, as measured by absorbance at 280 nm. GST
was then eluted with 50 mM Tris, pH 9, containing 50 mM GSH
(3 mL). The eluate was dialyzed overnight at 4 °C against 100 mM
potassium phosphate buffer, pH 6.5, containing 10% glycerol,
1 mM b-mercaptoethanol; 1 L. Protein samples were frozen on
dry ice and stored at À80 °C until use. GST protein homogeneity
was confirmed by SDS–PAGE (data not shown).
Materials and methods
Sources of chemicals were as follows: glutathione-agarose, IPTG
(isopropyl-b-D-1-thiogalactopyranoside), ampicillin sodium salt,
tetracycline HCl, b-mercaptoethanol, Martius Yellow, and bovine
serum albumin (BSA): Sigma–Aldrich (Oakville, ON); 1-chloro-
2,4-dinitrobenzene (CDNB; 98%): Alfa Aesar (Ward Hill, MA);
glutathione (99.6%): ChemImpex (Wood Dale, IL); 2,4-dinitronaph-
thalene (DNN; 100%): Accu-Standard, Inc. (New Haven, CT); Difco
agar, Tryptone, Yeast Extract: Becton, Dickson Co. (Sparks, MD);
Oxoid Nutrient Broth No. 2: Oxoid, Ltd. (Hampshire, England);
lysozyme: Boehringer-Mannheim (Germany).
Recombinant hGST-expressing Escherichia coli strains were the
kind gift of Dr. Bengt Mannervik (Uppsala University, Sweden).
UV–visible spectra were recorded on
spectrophotometer.
a
Cary BIO300
Synthesis and characterization of dinitronaphthalene derivatives
CDNN (See Fig. 1 and Table 1 for structures) was synthesized
from 1-hydroxy-2,4-dinitrobenzene (Martius Yellow, DNNOH)
according to published procedures [26,27]. The methoxy and eth-
oxy derivatives (DNNOMe and DNNOEt) were also synthesized in
good yields (56% and 78%, respectively) from the reactions of CDNN
with the corresponding sodium alkoxides, instead of the potassium
salts as previously reported [24]. Reaction conditions: DNNOMe:
Na(s), dry MeOH, CDNN, 10 °C to RT, 1 h (56%); DNNOEt: Na(s),
dry EtOH, CDNN, 10 °C to RT, 1 h (78%). The melting points and
1H NMR data of DNNOMe and DNNOEt were in agreement with lit-
erature values [24].
The GSH conjugate of CDNN was synthesized by a modification
of the method of Shiotsuki et al. [29]. GSH and CDNN, each 1 mM,
were dissolved in EtOH, 1.2 mL. NaOH (2 M, 1 mL) was added drop-
wise. The color of the solution immediately changed from yellow
to red–orange. The reaction was stirred for 30 min at RT and then
CDNB enzyme assay; inhibition by dinitronaphthalene and
derivatives; enzyme kinetics
GST activity was assayed with CDNB. Formation of the dinitro-
phenyl-GSH was monitored at 340 nm with a Cary 300 dual-beam
spectrophotometer (1.5 mL quartz cuvettes). Enzyme assays were
performed at RT in 100 mM potassium phosphate buffer, pH 6.5,
with 1 mM GSH and 0.5 mM CDNB [32].
hGST/inhibitor combinations chosen for detailed analysis were
identified by screening the effect of 25
lM inhibitor. For those
combinations which showed decreased activity, inhibitor IC50

H. Groom et al. / Archives of Biochemistry and Biophysics 555–556 (2014) 71–76
73
values were measured as follows. Recombinant hGST was pre-
incubated with inhibitor (dissolved in MeCN) and GSH, 1 mM, for
5 min at 37 °C; the final concentration of MeCN was 1%. The reac-
tion was initiated by addition of CDNB, 0.5 mM, in ethanol (1% final
concentration).
E. coli [30]. These enzymes are representative of three major
classes of xenobiotic-conjugating GST enzymes, Alpha, Mu, and Pi
[38,39], and each is a catalyst of CDNB conjugation [40,41]. Theta
class enzymes were not tested, since their CDNB conjugation activ-
ity is very low.
The amount of enzyme per incubation was 15
following cases, where less enzyme was used, to maintain the con-
dition [I] > [E]: M2-2, all inhibitors except DNNOH, 1 g; M1-1 and
DNN, 1 g; A1-1 and DNNOH, 5 g. All activity assays were con-
ducted in triplicate. The IC50 values were measured by fitting the
four-parameter logistic function [33] to each data set, using Sigma-
Plot™ software.
lg, except for the
CDNN glutathione adduct
l
l
l
CDNN reacted with GSH under the conditions used for syn-
thesis of the CDNB adduct: addition of NaOH to a solution of
GSH and the electrophile in aqueous ethanol solution [42]. The
product 2,4-dinitronaphthylglutathione was isolated and charac-
terized by HPLC–ESI–MS (M + H+, m/z = 524; M + Na+, m/z = 546).
To our surprise, however, no reaction was observed in enzyme
Mechanism of inhibition
incubations of GST (15
lg/mL) with GSH (1 mM) and CDNN
For selected combinations of enzyme and potent inhibitor, dou-
ble-reciprocal plots were constructed by varying the concentra-
tions of both the inhibitor and CDNB. Enzyme assays were
carried out as described above, with [GSH] = 1 mM, and [inhibitor]
and [CDNB] as indicated.
(e.g., 10 M) at pH values ranging from 6.5 to 7.4; no change
l
in the CDNN absorbance spectrum was seen; all six GSTs were
tested.
Inhibition of GST activity by dinitronaphthalene derivatives
Reversibility of inhibition/dialysis
Finding no evidence that CDNN is a substrate for GST-catalyzed
GSH conjugation, we then tested the compound as a possible inhib-
itor of the GST-catalyzed GSH conjugation of CDNB [32]. Indeed,
CDNN proved to be a potent inhibitor. Inhibition curves were mea-
sured for all six dinitronaphthalene derivatives and all six
enzymes, although in several cases (e.g., GST A2-2, A4-4, P1-1)
inhibition was weak or undetectable. A typical inhibition curve is
shown in Fig. 2 and the IC50 values are summarized in Table 2.
The strongest inhibitory effects were observed with GST M2-2
and the O-alkyl compounds 1-methoxy- and 1-ethoxy-2,
4-dinitronaphthalene, with IC50 values below 1 lM. There was no
clear relationship between predicted partition coefficient and
inhibitory potency (Fig. 3).
CDNN-treated GST A1-1 was subjected to dialysis (see Materials
and Methods). Following dialysis, enzyme activity was equal to that
of enzyme incubated without the inhibitor (data not shown), indi-
cating that the inhibition was fully reversible. Double-reciprocal
plot analysis was consistent with a competitive mode of inhibition
(Fig. 4).
To test reversibility of inhibition, the inhibitor was removed by
dialysis. Incubations (1 mL) containing GST A1-1 (25
lg), CDNN
(25 M), and GSH (1 mM) in potassium phosphate buffer, pH 6.5,
l
were dialyzed (molecular weight cut-off, 12–14 kDa) against 1 L
of buffer for 25 h at 4 °C. Control incubations lacked CDNN.
Meisenheimer complex formation: spectrophotometric experiments
Meisenheimer complex spectra were recorded as follows [14].
hGST preparations were dialyzed overnight against 100 mM potas-
sium phosphate buffer, pH 6.5, to remove b-mercaptoethanol.
Inhibitors (12.5
lM) were added to 100 mM potassium phosphate
buffer, pH 6.5, containing 50
lM hGST active sites (i.e., 25 lM
dimeric enzyme) and 1 mM GSH. The solvent (MeCN, except MeOH
in the case of DNNOH) concentration was 2%.
Mutagenicity assays
The mutagenicity assay (Ames test) [34] was performed with
Salmonella strain YG1024, a derivative of TA98 which expresses
high levels of arylamine N-acetyltransferase/N-hydroxyarylamine
O-acetyltransferase and is highly sensitive to the mutagenicity of
nitroaromatic compounds [35,36]. Bacteria were grown overnight
80
60
40
20
0
in Oxoid Nutrient Broth No. 2 containing ampicillin 25
and tetracycline 6.25 g/mL. Mutagen stocks were prepared in
MeCN, except for DNNOH, which was prepared in MeOH. Mutagen
stock (10 L) was added to an aliquot of culture (100 L) and 0.1 M
lg/mL
l
l
l
sodium phosphate buffer, pH 7.4, 0.5 mL, for 30 min at 37 °C, with
shaking, prior to plating. Assays were done in triplicate on at least
two different days. Colony counting was performed using NIST
Integrated Colony Enumerator (NICE) software [37]. The plate
was photographed with a digital camera; colonies were counted
in the inscribed square region of the circular 90 mm-diameter dish
and the counts were corrected by multiplication by the ratio of the
0
1
10
100
1000
[DNNOEt] (nM)
areas (p/2).
Fig. 2. Inhibition of GST M2-2 activity by DNNOEt. GST M2-2 (1lg) was incubated
with the inhibitor (dissolved in acetonitrile) and GSH, 1 mM, at 37 °C for 5 min, in
100 mM potassium phosphate buffer, pH 6.5. CDNB (dissolved in ethanol) was
added to initiate the reaction (final concentration, 0.5 mM), and product formation
was monitored at 340 nm. Final concentrations of acetonitrile and ethanol were 1%
each. Data points represent the mean s.e. of three independent experiments. IC50
values were determined by curve-fitting, as described in Materials and methods.
The corresponding experiment was carried out with the other inhibitors and
enzymes, as summarized in Table 2.
Results
GST expression
Six recombinant human cytosolic GST enzymes, A1-1, A2-2,
A4-4, M1-1, M2-2, and P1-1, were prepared by expression in

74
H. Groom et al. / Archives of Biochemistry and Biophysics 555–556 (2014) 71–76
Table 2
IC50 values (lM) of inhibitors towards recombinant human GST enzymes.
0.3
0.2
DNNOMe
DNNOEt
A1-1
A2-2
A4-4
M1-1
M2-2
P1-1
Sp. Act.
CDNN
DNNOH
DNNSG
DNNOMe
DNNOEt
DNN
27.4
3.6
0.9
5.1
6.1
41.8
35.7
4.1
7.6
2.9
47.5
0.7
97.6
3.4
0.12
0.18
1.3
39.2
91.2
*
*
*
*
*
*
*
76.2
89.7
*
*
9.8
*
*
*
*
*
*
*
0.1
ΔA
*
*
2.1
1.3
CDNN
0.0
*
Indicates no effect at inhibitor concentrations up to 25
l
M. The IC50 values were
determined as shown in Fig. 3.
DNN
-0.1
300
400
500
600
1.4
1.2
1.0
λ (nm)
Fig. 5. Absorption difference spectra of Meisenheimer complexes. Yankeelov
(tandem) cuvettes were used, to facilitate comparison of pre- and post-mixing
spectra. Final concentrations were: dinitronaphthalene (12.5
lM), GSH (1 mM), and
GST M2-2 (25 M; concentration of active sites; each enzyme molecule is
l
a
0.8
IC50
homodimer with two identical active sites) in 100 mM potassium phosphate buffer,
pH 6.5. Final concentration of acetonitrile (solvent for DNN) was 1%. One
compartment of the tandem sample cuvette was filled with enzyme and GSH in
buffer; the other compartment contained dinitronaphthalene solution in buffer. The
reference tandem cuvette was prepared identically, except that dinitronaphthalene
and enzyme were omitted. Spectra (sample – reference) were recorded before and
immediately after inverting and mixing both sample and reference cuvettes, at
room temperature. The plots show the difference spectra obtained by numerically
subtracting the pre-mixing spectrum from the corresponding post-mixing spec-
trum, for each dinitronaphthalene derivative. Heavy line: DNNOEt; thin line: DNN;
short-dashed line: CDNN; long-dashed line, DNNOMe.
(μ M)
0.6
0.4
0.2
0.0
2.6
2.8
3.0
3.2
3.4
logP
Fig. 3. Correlation analysis of IC50 values (GST M2-2, Table 2) vs. logP values
(Table 1) for the four potent dinitronaphthalene inhibitors for which calculated
logP values are available. Symbols: diamond, DNN; circle, DNNOMe; triangle,
CDNN; square, DNNOEt.
Table 3
UV–visible spectra of inhibitors and Meisenheimer complexes. The inhibitors (first
column) are listed with the kmax values of the free dinitronaphthalene compound
(second column) and of the corresponding Meisenheimer complex with GST M2-2
and GSH (third column) (by comparison, the kmax value for the dimethoxy complex
(Fig. 2, X = Y = -OMe) was given as 495 nm [24]).
0.03
0.02
0.01
0.00
-0.01
kmax (nm)
Inhibitor
Complex
CDNN
372
393, 436
512
295, 360
295, 360
288, 371
478
–
–
480
474
495
DNNOH
DNNSG
DNNOMe
DNNOEt
DNN
DNNOMe, and DNNOEt (Dissociation constants were not deter-
mined, because the inhibitors were generally so potent that, under
conditions where [E] ꢀ [I], the absorbance spectra were too weak
for accurate measurement.).
Once formed, the Meisenheimer complexes were very stable.
No significant loss of red color was seen when an incubation of
-5
0
5
10
15
20
25
1/[CDNB] (mM-1)
Fig. 4. Double-reciprocal plots of GST M2-2 inhibition by DNNOEt. DNNOEt
concentrations were 1.2 M (filled circles), 0.6 M (triangles), 0.3 M (squares);
MÀ1 mM. Assay condi-
l
l
l
GSTM2-2 (25 lM), GSH (1 mM), and CDNN (12.5 lM) was held at
no inhibitor, open circles. CDNB concentrations were 50
l
room temperature over a period of hours. Identical incubations
held at 37C gradually decolorized over the same length of time.
Subsequent addition of a second aliquot of GSH (1 mM) restored
the color, indicating that the decolorization is due to the gradual
autoxidation of GSH.
tions were as indicated for Fig. 2.
Formation of dinitronaphthalene Meisenheimer complexes at the GST
active site
Immediately upon addition of CDNN to an incubation contain-
Mutagenicity assays
ing a high concentration of GST (approx. 650 lg/mL) and GSH,
the characteristic red color of a Meisenheimer complex appeared
(Fig. 5 and Table 3). This complex formation had not been seen
at low (catalytic) concentrations of enzyme. Meisenheimer com-
plex formation was reversible: the red color was lost after over-
night dialysis against buffer as verified with DNN, CDNN,
CDNB is a direct-acting frameshift mutagen in the Ames test
[43]. Both DNN and DNNOH are direct-acting mutagens in Ames
test strains TA100 and TA98, indicating both base-substitution
and frameshift mutagenic activity [44]. We tested each of the dini-
tronaphthalene derivatives (except DNNSG) in strain YG1024. The

H. Groom et al. / Archives of Biochemistry and Biophysics 555–556 (2014) 71–76
75
3000
2500
2000
1500
1000
500
CDNN
DNN-OMe
DNN-OEt
0
1500
DNN
DNN-OH
1000
500
0
Fig. 7. Model of the structure of human GST M1A-1A complexed with glutathionyl
dinitronaphthalene. The figure shows the result of adding a second aromatic ring to
the structure of the bound glutathionyl-S-dinitrobenzene ligand in structure
1XWK.pdb [15], without further optimization.
0.01
0.1
1
10
100
1000
Kunze and Heps synthesized several GST inhibitors, GSH ana-
logues bearing dialkoxyphosphinyl (O@PA(OR)2) groups replacing
the side-chain of the GSH cysteine residue [52]. These compounds
were tested against several recombinants human and purified por-
cine GSTs. The IC50 values for recombinant human GST M1-1 were
in the low to high micromolar range, and class Mu enzymes were
more sensitive than classes Alpha or Pi. In contrast, a lipophilic
nmol per plate
Fig. 6. Ames test mutagenicity (His+ revertants per plate) for dinitronaphthalene
derivatives; strain YG1024. Data points represent mean s.e. of at least six plates
from two independent experiments. The declines in revertant yield at higher doses
were, in each case, due to toxicity, as evident from thinning of the background lawn
of auxotrophic colonies.
GSH conjugate,
c-glutamyl-(S-9-fluorenylmethyl)-L-cysteinyl-gly-
cine, showed sub-micromolar potency with human GST A1-1, but
weaker inhibition of human GSTs M2-2 and P1-1 [53]. Therefore,
the presence of a hydrophobic aryl substituent on the S atom does
not guarantee specificity for Mu class GSTs.
A crystal structure is available of GST M1-1 complexed with
glutathionyl-S-dinitrobenzene [15]. We simply replaced the ben-
zene ring in the structure with a naphthalene ring, without any
further modeling of possible favorable interactions between the
added aromatic ring and the protein (Fig. 7), and it can be seen that
the larger substrate is accommodated without steric clashes. This
is consistent with the observed formation of stable complexes with
the dinitronaphthalene derivatives.
results are shown in Fig. 6. All of the tested compounds were
direct-acting frameshift mutagens and all showed bacterial toxic-
ity at higher doses, as indicated by a decrease in yield of revertant
colonies as well as by the thinning of the background auxotrophic
lawn, characteristic of toxicity in the Ames test [45]. CDNN was
much more toxic than the other compounds, with toxicity evident
above 0.3 nmol per plate, although it was nevertheless possible to
observe a mutagenic response at lower doses.
Discussion
All of the dinitronaphthalene derivatives tested were direct-
acting frameshift mutagens. This result was expected, in view of
previous studies of these and related nitroaromatic compounds
[44,54]. The increased mutagenicity observed in strain YG1024 rel-
ative to strain TA98 is consistent with the view that the com-
pounds are activated by the well-established mechanism of
nitroreductase-mediated reduction to N-hydroxyarylamine deriva-
tives, followed by O-acetylation, ultimately generating a DNA-reac-
tive nitrenium ion [55]. Without doubt, the mutagenicity of
nitroaromatics is a major liability to their possible therapeutic
application. For this reason, it will be of interest to test whether
naphthalene derivatives bearing other electron-withdrawing func-
tional groups (e.g., cyano) are also active as GST inhibitors.
Among the three GST Alpha class enzymes examined, we found
that GST A1-1 is sensitive to inhibition by certain dinitronaphtha-
lenes, notably Martius yellow (DNNOH), whereas forms A2-2 and
A4-4 are much more resistant (Table 2). This result is consistent
with previous studies of these enzymes, which have characterized
GST A1-1 as ‘‘promiscuous’’ and GST A4-4 as ‘‘highly selective’’
[46]. GST A2-2 is best known for its GSH peroxidase activity [46]
and has much lower activity than GST A1-1 with respect to the
GSH conjugation of (for example) polycyclic aromatic hydrocarbon
dihydrodiol epoxides [47]. The Mu class enzyme GST M2-2 was the
most dinitronaphthalene-sensitive enzyme. As noted by Wu and
Dong, ‘‘mu-class GSTs have a larger and more open active site than
Alpha GSTs. This is consistent with the fact that Mu-class GST sub-
strates include many bulkier electrophilic agents such as aflatoxin
B1-epoxides and benzpyrene diols.’’ [2]. GST M2-2 is usually
referred to as a muscle-specific enzyme [48], but it is also
expressed in brain [49], kidney, heart, testis, and other organs
[50]. The BioGPS gene portal [51] database, surprisingly, shows
liver as the human cell type with much the highest level of GSTM2
mRNA expression (Possibly, DNA sequence homology with other
Mu-class GSTs led to spurious results in some of these expression
analyses.).
Acknowledgments
This work was supported by a grant from the Natural Sciences
and Engineering Research Council of Canada (Josephy) and by
National Science Foundation, USA, grant CHE 0809162 (Lee). We
wish to thank Dr. Matt Kimber for assistance with preparation of
Fig. 7.

76
H. Groom et al. / Archives of Biochemistry and Biophysics 555–556 (2014) 71–76
[29] T. Shiotsuki, A. Koiso, M. Eto, Pestic. Biochem. Physiol. 37 (1990) 121–129.
References
[30] S. Mukanganyama, M. Widersten, Y.S. Naik, B. Mannervik, J.A. Hasler, Int. J.
Cancer 97 (2002) 700–705.
[31] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[32] W.H. Habig, M.J. Pabst, W.B. Jakoby, J. Biol. Chem. 249 (1974) 7130–7139.
[33] M.J. Healy, Biochem. J. 130 (1972) 207–210.
[34] D.M. Maron, B.N. Ames, Mutat. Res. 113 (1983) 173–215.
[35] M. Watanabe, M.J. Ishidate, T. Nohmi, Mutat. Res. 234 (1990) 337–348.
[36] P.D. Josephy, P. Gruz, T. Nohmi, Mutat. Res. 386 (1997) 1–23.
[37] M.L. Clarke, R.L. Burton, A.N. Hill, M. Litorja, M.H. Nahm, J. Hwang, Cytometry
A. 77 (2010) 790–797.
[38] R.L. Schecter, M.A. Alaoui-Jamali, G. Batist, Biochem. Cell Biol. 70 (1992) 349–
353.
[39] D. Townsend, K. Tew, Am. J. Pharmacogenomics 3 (2003) 157–172.
[40] R. Hurst, Y. Bao, P. Jemth, B. Mannervik, G. Williamson, Biochem. J. 332 (1998)
97–100.
[41] B. Mannervik, M. Widersten, in: G.M. Pacifici, G.N. Fracchia (Eds.), Advances in
Drug Metabolism in Man, European Commission, Luxembourg, 1995, pp. 407–
459.
[42] V.L. Schramm, R. McCluskey, F.A. Emig, G. Litwack, J. Biol. Chem. 259 (1984)
714–722.
[1] B. Mannervik, Biochem. Soc. Trans. 24 (1996) 878–880.
[2] B. Wu, D. Dong, Trends Pharmacol. Sci. 33 (2012) 656–668.
[3] P.G. Board, D. Menon, Biochim. Biophys. Acta 1830 (2013) 3267–3288.
[4] K.E. Griswold, N.S. Aiyappan, B.L. Iverson, G. Georgiou, J. Mol. Biol. 364 (2006)
400–410.
[5] M.A. Norrgård, B. Mannervik, J. Mol. Biol. 412 (2011) 111–120.
[6] M.T. Honaker, M. Acchione, W. Zhang, B. Mannervik, W.M. Atkins, J. Biol. Chem.
288 (2013) 18599–18611.
[7] S. Mahajan, W.M. Atkins, Cell. Mol. Life Sci. 62 (2005) 1221–1233.
[8] Y. Musdal, U.M. Hegazy, Y. Aksoy, B. Mannervik, Chem. Biol. Interact. 205
(2013) 53–62.
[9] I. Sekine, J.D. Minna, K. Nishio, N. Saijo, T. Tamura, Jpn. J. Clin. Oncol. 37 (2007)
329–336.
[10] A.G. Clark, J.N. Smith, T.W. Speir, Biochem. J. 135 (1973) 385–392.
[11] P. Jemth, B. Mannervik, Arch. Biochem. Biophys. 348 (1997) 247–254.
[12] J.H. Keen, W.H. Habig, W.B. Jakoby, J. Biol. Chem. 251 (1976) 6183–6188.
[13] D.S. Pushparajah, M. Umachandran, K.E. Plant, N. Plant, C. Ioannides,
Mutagenesis 23 (2008) 299–308.
[14] G.F. Graminski, P.H. Zhang, M.A. Sesay, H.L. Ammon, R.N. Armstrong,
Biochemistry 28 (1989) 6252–6258.
[15] Y. Patskovsky, L. Patskovska, S.C. Almo, I. Listowsky, Biochemistry 45 (2006)
3852–3862.
[16] A.L. Bowman, L. Ridder, I.M. Rietjens, J. Vervoort, A.J. Mulholland, Biochemistry
46 (2007) 6353–6363.
[17] B. Mannervik, P.G. Board, J.D. Hayes, I. Listowsky, W.R. Pearson, Methods
Enzymol. 401 (2005) 1–8.
[18] X. Ji, R.N. Armstrong, G.L. Gilliland, Biochemistry 32 (1993) 12949–12954.
[19] L. Prade, R. Huber, T.H. Manoharan, W.E. Fahl, W. Reuter, Structure 5 (1997)
1287–1295.
[20] A.G. Clark, M. Sinclair, Biochem. Pharmacol. 37 (1988) 259–263.
[21] M. Sajan, G. Reddy, A.P. Kulkarni, Int. J. Toxicol. 19 (2000) 285–292.
[22] P. Bico, C.Y. Chen, M. Jones, J. Erhardt, H. Dirr, Biochem. Mol. Biol. Int. 33 (1994)
887–892.
[43] K.H. Summer, W. Goggelmann, H. Greim, Mutat. Res. 70 (1980) 269–278.
[44] E.C. McCoy, E.J. Rosenkranz, L.A. Petrullo, H.S. Rosenkranz, R. Mermelstein,
Environ. Mutagen. 3 (1981) 499–511.
[45] D. McGregor, R.D. Prentice, M. McConville, Y.J. Lee, W.J. Caspary, Environ.
Mutagen. 6 (1984) 545–557.
[46] C. Blikstad, A. Shokeer, S. Kurtovic, B. Mannervik, Biochim. Biophys. Acta 1780
(2008) 1458–1463.
[47] K. Dreij, K. Sundberg, A.S. Johansson, E. Nordling, A. Seidel, B. Persson, B.
Mannervik, B. Jernstrom, Chem. Res. Toxicol. 15 (2002) 825–831.
[48] L. Wei, Y.A. Abdellatif, D. Liu, T. Kimura, M. Coggan, E.M. Gallant, N.A. Beard,
P.G. Board, A.F. Dulhunty, Int. J. Biochem. Cell Biol. 40 (2008) 1616–1628.
[49] C.T. Beuckmann, K. Fujimori, Y. Urade, O. Hayaishi, Neurochem. Res. 25 (2000)
733–738.
[50] J.D. Rowe, E. Nieves, I. Listowsky, Biochem. J. 325 (1997) 481–486.
[51] C. Wu, C. Orozco, J. Boyer, M. Leglise, J. Goodale, S. Batalov, C.L. Hodge, J. Haase,
J. Janes, J.W. Huss III, A.I. Su, Genome Biol. 10 (2009) R130.
[52] T. Kunze, S. Heps, Biochem. Pharmacol. 59 (2000) 973–981.
[53] I. Cacciatore, A.M. Caccuri, A. Cocco, F. De Maria, A. Di Stefano, G. Luisi, F.
Pinnen, G. Ricci, P. Sozio, P. Turella, Amino Acids 29 (2005) 255–261.
[54] H.S. Rosenkranz, R. Mermelstein, Mutat. Res. 114 (1983) 217–267.
[55] P.D. Josephy, M. Novak, Front. Biosci. (Schol. Ed.) 5 (2013) 341–359.
[23] R. Foster, C.A. Fyfe, P.H. Emslie, M.I. Foreman, Tetrahedron 23 (1967) 227–237.
[24] J.H. Fendler, E.J. Fendler, W.E. Byrne, C.E. Griffin, J. Org. Chem. 33 (1968) 977–
985.
[25] F. Terrier, Modern Nucleophilic Aromatic Substitution, Wiley-VCH, New York,
2013.
[26] F. Ullman, W. Bruck, Ber. Dtsch. Chem. Ges. 41 (1908) 3932–3939.
[27] H.W. Talen, Rec. Trav. Chim. Pays-Bas 47 (1928) 329–345.
[28] A. Sato, A. Scott, T. Asao, M. Lee, J. Org. Chem. 71 (2006) 4692–4695.