In vitro studies of dna damage caused by tricyclic antidepressants: A role of peroxidase in the side effects of the drugs

Article abstract of DOI:10.1021/tx100221b

Studies show that tricyclic antidepressants prescribed for migraines, anxiety, and child enuresis have numerous adverse effects in living cells. One of the undesired outcomes observed under treatment with these drugs is DNA damage. However, the mechanisms underlying damage have yet to be elucidated. We performed in vitro studies of the DNA damage caused by four tricyclic antidepressants: imipramine, amitriptyline, opipramol, and protriptyline. We focused particularly on the DNA damage aided by peroxidases. As a model of a peroxidase, we used horseradish peroxidase (HRP). At pH 7, reactions of HRP with excess hydrogen peroxide and imipramine yielded an intense purple color and a broad absorption spectrum with the maximum intensity at 522 nm. Reactions performed between DNA and imipramine in the presence of H₂O ₂ and HRP resulted in the disappearance of the DNA band. In the case of the other three drugs, this effect was not observed. Extraction of the DNA from the reaction mixture indicated that DNA is degraded in the reaction between imipramine and H₂O₂ catalyzed by HRP. The final product of imipramine oxidation was identified as iminodibenzyl. We hypothesize that the damage to DNA was caused by an imipramine reactive intermediate.

Full text of DOI:10.1021/tx100221b

Chem. Res. Toxicol. 2010, 23, 1497–1503
1497
In Vitro Studies of DNA Damage Caused by Tricyclic
Antidepressants: A Role of Peroxidase in the Side Effects of the
Drugs
Ekaterina A. Korobkova,* William Ng, Abhishek Venkatratnam, Alicia K. Williams,
Madina Nizamova, and Nikolay Azar
Science Department, John Jay College of Criminlal Justice, 445 W. 59th AVenue, New York, New York 10019
ReceiVed June 30, 2010
Studies show that tricyclic antidepressants prescribed for migraines, anxiety, and child enuresis have
numerous adverse effects in living cells. One of the undesired outcomes observed under treatment with
these drugs is DNA damage. However, the mechanisms underlying damage have yet to be elucidated.
We performed in Vitro studies of the DNA damage caused by four tricyclic antidepressants: imipramine,
amitriptyline, opipramol, and protriptyline. We focused particularly on the DNA damage aided by
peroxidases. As a model of a peroxidase, we used horseradish peroxidase (HRP). At pH 7, reactions of
HRP with excess hydrogen peroxide and imipramine yielded an intense purple color and a broad absorption
spectrum with the maximum intensity at 522 nm. Reactions performed between DNA and imipramine in
the presence of H2O2 and HRP resulted in the disappearance of the DNA band. In the case of the other
three drugs, this effect was not observed. Extraction of the DNA from the reaction mixture indicated that
DNA is degraded in the reaction between imipramine and H2O2 catalyzed by HRP. The final product of
imipramine oxidation was identified as iminodibenzyl. We hypothesize that the damage to DNA was
caused by an imipramine reactive intermediate.
radicals formed during peroxidase catalysis are potentially
harmful to the living cells, as they react with DNA, proteins,
fatty acids, and other molecules (16, 17). These reactive
intermediates may abstract a hydrogen atom from a ribose ring
causing DNA cleavage or inter- and intramolecular cross-links.
Molecules with an unpaired electron may also bind to DNA,
producing stable covalent complexes (18). The resulting prod-
ucts prevent certain DNA processes from functioning normally.
We performed in Vitro DNA damage studies of two high
suicide-risk antidepressants, imipramine and amitriptyline (19, 20),
and structurally related opipramol and protriptyline (Figure 1).
As a model of peroxidase, we employed horseradish peroxidase
(HRP). We attempted to correlate the DNA damage potency of
the antidepressants with their structural characteristics. We
performed comparative analysis of their DNA damage potency
aided by hydrogen peroxide and HRP.
Introduction
Tricyclic antidepressants have been extensively used over the
last several decades for the treatment of insomnia, headaches,
depression, and childhood enuresis. These drugs can be very
toxic not only if taken in overdosage but also under normally
prescribed amounts. Antidepressants and their metabolites are
potentially very reactive toward DNA in living cells. DNA
presents many reactive rich sites; its bases and backbones are
often targets for various chemicals present in the cell. The
damage, if not repaired, can lead to very serious health issues,
particularly inducing cancer (1, 2). Tricyclic antidepressants with
their planar structures can insert between DNA bases, forming
stacking complexes. For example, imipramine has been found
to be a DNA intercalator (3, 4). The metabolism of antidepres-
sants may lead to even more severe DNA damage, such as strand
breaks (5).
The DNA damage caused by tricyclic antidepressants has
been studied in cells, animals, and humans. Strand breaks,
chromosome aberrations, and sister chromatid exchanges have
been detected in cells exposed to imipramine and amitriptyline
(5, 6). Mice studies revealed the decrease of marrow-mitotic
division caused by imipramine and desipramine (7). DNA
damage was also detected in a group of children treated with
imipramine for nocturnal enuresis (8). The mechanisms underly-
ing DNA damage caused by tricyclic antidepressants and their
metabolites are not completely understood.
Experimental Procedures
Materials. Imipramine hydrochloride, purity g99% (Allentown,
PA), amitriptyline hydrochloride, purity g98% (Allentown, PA),
opipramol dihydrochloride, purity g98% (Saint Louis, MO),
protriptyline hydrochloride, purity g99% (Saint Louis, MO), DNA
type I from calf thymus (Allentown, PA), peroxidase from
horseradish type VI, 250-330 units/mg (Allentown, PA), and
hydrogen peroxide (Allentown, PA) were purchased from Sigma
Aldrich. Wide range DNA marker, 16 fragments, 50-10,000 bp
(Saint Louis, MO) was also purchased from Sigma Aldrich. Plasmid
pBR322 (0.25 µg/µL, 4363 bp) purified from E.coli DH10B
(Carlsbad, CA) was purchased from Invitrogen. PCI mixture, ratio
25:24:1, biotech grade, 1 M Tris buffer (pH 6.7), was purchased
from Fisher Scientific (Fairlawn, NJ).
Recently, much interest has been drawn toward radical
mediated DNA damage (9-12). There is also sufficient evidence
of free radical formations due to the oxidation of drugs,
carcinogens, and xenobiotics by peroxidases (13-15). The
Reactions. The standard reaction mixture with calf thymus DNA
contained (unless otherwise specified) 0.2 µM HRP, 13 µM per bp
calf thymus DNA (∼10 ng/µL), 500 µM antidepressant, and 500
µM H2O2 in Sorenson buffer (pH 7.0) containing 67 mM dibasic
* To whom correspondence should be addressed. Tel: 1-212-237-8064.
E-mail: ekorobkova@jjay.cuny.edu.
10.1021/tx100221b  2010 American Chemical Society
Published on Web 08/30/2010

1498 Chem. Res. Toxicol., Vol. 23, No. 9, 2010
KorobkoVa et al.
Figure 1. Structures of tricyclic antidepressants.
sodium phosphate and 67 mM monobasic potassium phosphate.
The standard reaction mixture with pBR322 plasmid contained 0.2
µM HRP, 3 ng/µL (∼5 µM per bp) plasmid, 500 µM antidepressant,
and 500 µM H2O2 in 67 mM Sorenson buffer (pH 7.0). The volume
of the reaction mixtures prepared for the 15-well gels was 10 µL;
the volume of the reactions prepared for 8-well gels was 20 µL.
The components were added in the order listed. All reaction
mixtures were incubated at 37 °C for 1 h in a water bath.
Gel Electrophoresis. One percent agarose gels were prepared
in TAE buffer. Electrophoresis was performed at 80 V for 120 min.
In the case of 15-well gels, the loading samples contained 10 µL
of the reaction mixture, and 1.1 µL of the 10× loading buffer. In
the case of 8-well gels, the loading samples contained 20 µL of
the reaction mixture, and 2.2 µL of the 10× loading buffer. The
10× loading buffer consisted of 50% glycerol, 0.25% bromophenol
blue, and 0.25% xylene cyanole FF in TAE buffer. The gel was
stained in ethidium bromide (0.5 µg/mL) for 20 min and then
destained in distilled water for 20 min. The gel images were
captured with a PerkinElmer Geliance 200 Imaging system.
PCI DNA Extraction Protocol. Forty microliters of DNA
sample was taken in a 1.5 mL Eppendorf tube. Equal volume of
phenol/chloroform/isoamyl alcohol (PCI, 25:24:1) was added to the
tube. The two liquid layers (aqueous and organic) were mixed by
flicking the tubes. The tubes were then centrifuged at 12,000 rpm
for 5 min. The aqueous phase (upper layer) was carefully transferred
to a clean centrifuge. To this tube, equal volume of chloroform
was added. The two liquid layers (aqueous and organic) were mixed
by flicking the tubes. The tubes were then centrifuged at 12,000
rpm for 5 min. The aqueous phase (upper layer) was carefully
transferred to a clean centrifuge. The chloroform extraction
procedure was repeated 7 times. The extracted samples were
quantified with NanoDrop 2000C Spectrophotometer (Thermo
Scientific).
Part II, MS. (a) Transfer line: 300 °C isocratic. (b) MSD
acquisition parameters: acquisition mode, scan; solvent delay, 2
min; EM offset, 0, EM voltage 965 was used. Scan parameters:
low mass, 50.0; high mass, 800; threshold, 150. Sampling rate (2n)
was 2. MS Quad temperature, 150 °C; MS Source temperature,
230 °C.
Results
Reactions between DNA and Antidepressants. Reactions
of antidepressants with DNA were performed in the presence
and absence of HRP/H2O2. Antidepressants were imipramine,
amitriptyline, opipramol, or protriptyline (Figure 1). Reaction
mixtures were electrophoressed, and the gel images are shown
in Figure 2. Reactions were performed with calf thymus DNA
(Figure 2A) and digested plasmid pRB322 (Figure 2B). Reac-
tions between DNA and the drugs showed a maintaining of the
DNA band (Figure 2A and B, lanes 4-7). However, reactions
between DNA and imipramine conducted in the presence of
Absorption Spectra of HRP. All UV-vis absorption spectra
were recorded with Shimazu UV-visible Spectrophotometer, UV-
1700. The original solution contained 10 µM HRP in 500 µL of
67 mM Sorenson buffer (pH 7.0). H2O2 stock solution of 0.5 mM
was added to a final concentration of 10 µM, after which the
spectrum of HRP-I was recorded. Antidepressants were added to a
final concentration of 5-20 µM from the 0.5 mM stock solutions,
after which the spectra of HRP-II and regenerated HRP were
recorded. The temperature in the cuvette chamber was maintained
at 20 °C. The reference cuvette contained 67 mM Sorenson buffer
(pH 7.0). The concentration of HRP was calculated using the
extinction coefficients at 404 nm of 102,000 M-1 cm-1 (21).
Gas Chromatography-Mass Spectrometry (GC-MS). Mass
spectrometry data was acquired on an Agilent Technologies 7890A
GC/MS instrument with 5975C Mass Selective Detector (MSD).
NIST MS Search 2.0 MS Spectral library was supplied with the
Agilent GC/MS instrument. The instrument was setup as follows.
Part I, GC. (a) Autosampler setup: 2 µL injections at fast speed
setting with 7 s viscosity delay. (b) Injection port setup: splitless
injection, front inlet, 250 °C; purge flow, 0.5 mL/min; purge time,
2 min; pressure, 10.523 psi; septum purge flow, 3 mL/min; gas
saver on at 20 mL/min after 2 min. (c) Column: an HP-5MS
column, 30 m × 25 mm × 0.25 µm film thickness; column He 5.0
gas flow, 1.0 mL/min at 7.37 psi. (d) Oven: oven initial temperature,
100 °C held for 0 min, at 8 °C/min ramp to 300 °C, held for 2
min; total run time, 27 min.
Figure 2. Agarose gel electrophoresis of reaction mixtures containing
DNA and antidepressants. All reactions mixtures were incubated for
1 h at 37 °C. Concentrations of all drugs, HRP, and H2O2 were 0.5
mM, 0.2 µM, and 0.5 mM, respectively, in all reactions. The
concentration of calf thymus DNA was 13 µM per bp in all reaction
mixtures. The concentration of the digested plasmid was 3 ng/µL. (A)
Reactions with calf thymus DNA. Lane 1, wide range DNA marker,
50-10,000 bp, 5 µL; lane 2, DNA only; lane 3, DNA and HRP/
H2O2 (control reaction); lane 4, DNA and imipramine; lane 5, DNA
and amitriptyline; lane 6, DNA and opipramol; lane 7, DNA and
protriptyline; lanes 8 and 9, DNA, imipramine, and HRP/H2O2; lanes
10 and 11, DNA, amitriptyline, and HRP/H2O2; lanes 12 and 13,
DNA, opipramol, and HRP/H2O2; lanes 14 and 15, DNA, protrip-
tyline, and HRP/H2O2. (B) Reactions with digested plasmid pBR322.
Lane 1, DNA marker, 5 µL; lane 2, plasmid only; lane 3, plasmid
and HRP/H2O2; lane 4, plasmid and imipramine; lane 5, plasmid
and amitriptyline; lane 6, plasmid and opipramol; lane 7, plasmid
and protriptyline; lanes 8 and 9, plasmid, imipramine, and HRP/
H2O2; lanes 10 and 11, plasmid, amitriptyline, and HRP/H2O2; lanes
12 and 13, plasmid, opipramol, and HRP/H2O2; lanes 14 and 15,
plasmid, protriptyline, and HRP/H2O2.

Interactions of Tricyclic Antidepressants with DNA
Chem. Res. Toxicol., Vol. 23, No. 9, 2010 1499
Scheme 1. Catalytic Cycle of HRP
UV spectra of native HRP and its oxidized forms are shown
in Figure 4A-D. Line a represents the spectrum of native
HRP (∼10 µM). One equivalent of H2O2 was added to a
brown solution of HRP with an appearance of a light green
color. Line b represents the spectrum of HRP oxidized by
H2O2, which is consistent with the spectrum recorded for
HRP-I (23). The spectrum shifted toward the longer wave-
lengths, and the intensity of the absorbance decreased
significantly. These observations are in agreement with the
other studies (24, 25).
The addition of one equivalent of imipramine or half
equivalent of opipramol to the solution of HRP-I led to the
UV-vis spectrum characteristic for HRP-II (23) with a
significant increase in signal compared to that of HRP-I, and
the maximum peak shifted to a longer wavelength with
respect to the spectrum of native HRP (Figure 4A and C,
line c). The addition of another equivalent of imipramine or
another half equivalent of opipramol resulted in the regenera-
tion of the spectrum characteristic for native HRP (Figure
4A and C, line d). The addition of 1 and 10 equivalents of
amitriptyline or protroptyline to the solution of HRP-I did
not lead to a spectrum change (Figure 4B and D, lines c and
d). This suggests that amitriptyline and protriptyline do not
react with the oxidized prosthetic group of HRP. The slight
increase in the intensity is associated with the instability of
the HRP-I, which is reduced spontaneously by solution
impurities (26). The addition of either antidepressant to a
solution of native HRP not containing H2O2 demonstrated
no effect on the spectrum.
Figure 3. Quantitation of DNA extracted from the reaction mixture
by PCI procedure (phenol/chloroform/isoamyl alcohol). Solid line, UV
spectrum of sample extracted from a 40 µL solution containing 100
µM calf thymus DNA in water. Dashed line, UV spectrum of sample
extracted from a 40 µL solution containing a reaction mixture: 100
µM calf thymus DNA, 2 µM HRP, 0.5 mM imipramine, and 5 mM
H2O2 in water. (Inset) Agarose gel electrophoresis of DNA (lane 2)
and DNA reaction mixture (lane 3). Gel electrophoresis was performed
before the PCI extraction. Twenty microliters was loaded in each lane.
Lane 1 is the wide range DNA marker, 50-10,000 bp.
HRP/H2O2 led to complete fading of the bands. (Figure 2A and
B, lanes 8 and 9). No similar effect was observed with
amitriptyline, opipramol, and protriptyline (Figure 2A and B,
lanes 10-15). The control reaction, DNA + HRP + H2O2,
resulted in the retention of the bands (Figure 2A and B, lane 3)
suggesting an involvement of the drug in the DNA damaging
process. The following control reactions were also performed
and resulted with no loss of DNA: (1) DNA + antidepressant
+ H2O2 and (2) DNA + antidepressant + HRP (data not
shown).
PCI DNA Extraction. DNA extraction was performed from
samples containing (a) 100 µM DNA and (b) 100 µM DNA
and imipramine incubated for 1 h at 37 °C in the presence of
HRP/H2O2 in water. The extraction was performed using a
Phenol/Chloroform/Isoamyl alcohol system. The detailed de-
scription of the procedure is shown in Experimental Procedures.
Briefly, DNA was collected in the aqueous layer, and then the
aqueous phase was purified with chloroform extractions. DNA
was finally quantified with the NanoDrop spectrophotometer,
and the results are shown in Figure 3. The control experiment
(Figure 3, solid line) shows the UV spectrum of DNA extracted
from the solution using the procedure described above. The
absorbance of 1.3 at λmax) 260 nm corresponds to a DNA
concentration of 100 µM with an extinction coefficient for calf
thymus DNA of 12,824 M(bp)-1 cm-1 (22). The dashed line in
Figure 3 shows the UV spectrum of the sample collected from
the reaction mixture initially containing 100 µM DNA. DNA
that was originally present in the solution disappeared from the
sample in the course of the reaction as indicated by the absence
of the peak at 260 nm. A peak at 270 nm is due to the remaining
phenol. Gel electrophoresis was also performed, and the results
are shown in the inset of Figure 3. This experiment combined
with the gel electrophoresis studies shown in Figure 2 indicates
that DNA degrades in the presence of imipramine, HRP, and
H2O2.
Both imipramine and opipramol proved to be substrates for
HRP. However, DNA degradation was observed only in the
presence of imipramine (Figure 2A and B, lanes 8 and 9; Figure
3). We performed more detailed studies of the spectra appearing
in the reaction of imipramine with HRP/H2O2.
The addition of excess imipramine to the mixture of HRP
and H2O2 yielded a spectrum with the maximum absorption at
522 nm. Figure 5A shows the UV-vis absorption spectrum
recorded from the mixture containing 0.2 µM HRP, 500 µM
H2O2, and 500 µM imipramine in 2 mM phosphate buffer (pH
7). The intensity of the spectrum increased for approximately
1 h and 20 min, after which it decreased showing a new
absorption maximum at 340 nm. The color of the solution
proceeded from pink to light purple and to brown as the
absorbance at 522 nm increased. A reaction between imipramine
and HRP/H2O2 performed in water yielded a similar spectrum
with a peak at 522 nm (data not shown). Mixing imipramine
with H2O2 or HRP alone did not yield a purple color or the
corresponding absorption spectrum.
Absorption Spectrophotometry Studies. The UV-vis ab-
sorption spectra were recorded for native and modified HRP.
The catalytic cycle of HRP is shown in Scheme 1. In this
scheme, HRP (native enzyme) is converted to the oxidized form
HRP-I by hydrogen peroxide. The reactions HRP-I f HRP-II
and HRP-II f HRP proceed through the abstraction of electrons
from the drug molecules resulting in the formation of free
radicals. In general, the oxidizing agent can be any organic
peroxidase.
The same experiments were repeated with different pH
values, and the results are shown in Figure 5B, C, and D.
The maximum of the spectrum shifted to longer wavelengths
as the pH decreased. At pH 5.5, the color of the solution
evolved from dark blue-brown (λmax ) 560 nm) to brown

1500 Chem. Res. Toxicol., Vol. 23, No. 9, 2010
KorobkoVa et al.
Figure 4. UV-vis absorption spectra of HRP treated with H2O2 and antidepressants. (A) Line a, native HRP 10 µM; line b, HRP with one equivalent
of H2O2; line c, spectrum recorded after the addition of 1 equivalent of imipramine; line d, after the addition of 2 equivalents of imipramine. (B)
Line a, 10 µM native HRP; line b, HRP with one equivalent of H2O2; line c, after the addition of 1 equivalent of amitriptyline; line d, after the
addition of 10 equivalents of amitriptyline. (C) Line a, 10 µM native HRP; line b, HRP with one equivalent of H2O2; line c, after the addition of
1 half equivalent of opipramol; line d, after the addition of 1 equivalent of opipramol. (D) Line a, 10 µM native HRP; line b, HRP with one
equivalent of H2O2; line c, after the addition of 1 equivalent of protriptyline; line d, after the addition of 10 equivalents of protriptyline.
Figure 5. UV-vis absorption spectra of reaction mixtures containing HRP and excess of H2O2 and imipramine. (A) Phosphate buffer (2 mM, pH
7). Concentrations of HRP, H2O2, and imipramine are 0.2 µM, 0.5 mM, and 0.5 mM, respectively. Spectrum a, 3 min after mixing the reagents;
spectrum b, 6 min after mixing; spectrum c, 10 min; spectrum d, 17 min; spectrum e, 29 min; spectrum f, 1 h 17 min; spectrum g, 1 h and 23 min;
spectrum h, 1 h and 45 min. (B) Phosphate buffer (2 mM, pH 5.5). Concentrations of HRP, H2O2, and imipramine are 0.2 µM, 0.5 mM, and 0.5
mM, respectively. Spectrum a, 2 min after mixing the reagents; spectrum b, 5 min after mixing; spectrum c, 7 min; spectrum d, 17 min; spectrum
e, 24 min; spectrum f, 1 h and 24 min; spectrum g, 3 h and 33 min. (C) Phosphate buffer (2 mM, pH 4.7). Concentrations of HRP, H2O2, and
imipramine are 0.2 µM, 0.5 mM, and 0.5 mM, respectively. Spectrum a, recorded after mixing the reagents; spectrum b, 2 min after mixing;
spectrum c, 4 min; spectrum d, 6 min; spectrum e, 14 min; spectrum f, 30 min. (D) Phosphate buffer (2 mM, pH 2). Concentrations of HRP, H2O2,
and imipramine are 2 µM, 1 mM, and 1 mM, respectively. Spectrum a, recorded after mixing the reagents; spectrum b, 2 min after mixing;
spectrum c, 4 min; spectrum d, 6 min; spectrum e, 8 min.
and to light orange (Figure 5B). At pH 4.7, the blue solution
(λmax ) 600 nm) turned red-brown in the course of the
reaction (Figure 5C). At pH 2, the blue solution (λmax) 650
nm) turned green with the maximum intensity decreasing,
yielding a new peak at ∼400 nm (Figure 5D). Borg (27)
recorded similar spectra in imipramine reaction mixtures with
ferric salts, ceric ion, and HRP/H2O2 in acid. Sakurai et al.
(28) observed the blue spectrum in the reaction of imipramine
and vanadate ion. In both studies, the blue color converted
to green spontaneously. EPR (electron paramagnetic reso-

Interactions of Tricyclic Antidepressants with DNA
Chem. Res. Toxicol., Vol. 23, No. 9, 2010 1501
Scheme 2. Resonance Structures of Imipraminea
Figure 6. Drug concentration dependence study. Gel electrophoresis
of the reaction mixtures containing plasmid DNA, imipramine, and
HRP/H2O2. All reaction mixtures contained 3 ng/µL (∼5 µM) of
digested pBR322 plasmid, 0.2 µM HRP, and 0.5 mM H2O2. Concentra-
tion of imipramine varied as follows: lane 1, none; lane 2, 2 µM; lane
3, 5 µM; lane 4, 8 µM; lane 5, 11 µM; lane 6, 14 µM; lane 7, 17 µM;
lane 8, 20 µM. All reaction mixtures were incubated for 1 h in a 37 °C
water bath.
a Numbers, 2, 4, and 6 indicate the location of the negative charge on
the carbon atoms of the benzene rings.
Discussion
Analysis of the reaction mixtures containing calf thymus
genomic DNA or digested plasmid pBR322, imipramine, and
HRP/H2O2 resulted in the disappearance of the DNA band from
the gel images (Figure 2A and B, lanes 8 and 9; Figure 3, inset).
The PCI extraction of calf thymus DNA from the reaction
mixtures containing imipramine and the HRP/H2O2 oxidation
system showed that DNA degrades in the reaction (Figure 3,
dashed line). The DNA damage is probably caused by the
reactive intermediate formed during electron transfer from
imipramine to the oxidized forms of HRP, HRP-I, and HRP-II.
Oxidation-reduction cycle of HRP proceeds through the
formation of HRP-I and HRP-II (Scheme 1). The structure of
the active sites of native and oxidized forms has been character-
ized for some peroxidases. The prosthetic group of the native
peroxidase contains iron in its ferric state. In HRP-I, the iron is
in the ferryl state in the iron-oxo complex (23). Moreover, HRP-I
is characterized by the presence of the radical cation, which
rests on the porphyrine ring of the heme (29) or in an amino
acid residue (30). The HRP-II complex of peroxidases contains
a fully covalent ferryl-oxo complex (23).
Spectroscopy studies of HRP were performed to study the
processes that affect DNA when antidepressants are present with
HRP/H2O2. The UV-vis absorption spectra of HRP and its
oxidized forms have been well characterized (23). The addition
of imipramine, amitriptyline, opipramol, and protriptyline to the
HRP-I solution led to different effects (Figure 4A-D). In the
case of amitriptyline or protriptyline, the addition of excess
amounts of the drug to the solution of HRP-I resulted in almost
no spectrum change (Figure 4B and D). However, in the case
of imipramine or opipramol, the consecutive addition of small
aliquots (within the 2 equivalents of HRP) of the drug led to a
spectrum characteristic of the HRP-II compound (Figure 4A
and C, line c) and a spectrum, which coincided with that of the
native HRP (Figure 4A and C, line d).
We envision that the difference in reactivity of the drugs with
HRP-I is associated with their nucleophilic properties. The
overlapping of the free electron pair on the N-atom of imi-
pramine may lead to an increase in nucleophilicity as a result
of negative charge delocalization on the benzene rings. This
can be rationalized through the resonance structures as shown
in Scheme 2. Both amitripyline and protriptyline do not have a
nitrogen atom bridging the aromatic rings and consequently do
not display this increased nucleophilicity (Figure 1). The
localization of the negative charge on the benzene rings of the
drug can also facilitate the abstraction of the electron, therefore
increasing its reduction potential and aiding the reactions,
HRP-I f HRP-II and HRP-II f HRP. As the enzyme carries
on through a large number of cycles, radicals may generate
(Scheme 1).
Figure 7. Mass spectral analysis of the imipramine oxidation product.
(A) Mass spectrum of the major product formed in the reaction of
imipramine with HRP/H2O2. The GC retention time of the compound
was 16 min. (B) Matching spectrum from the library and structure of
10,11-dihydro-5H-dibenz[b,f]azepine (iminodibenzyl).
nance) spectra were recorded from the reaction mixtures and
associated with a transitory free radical from imipramine
(27, 28).
Concentration Dependence Studies. Reactions between
plasmid DNA and imipramine were performed in the presence
of HRP/H2O2 with the drug concentration ranging from 2
µM to 20 µM (10 to 100 mol equiv of HRP). The reaction
mixtures were analyzed with gel electrophoresis, and the
image is shown in Figure 6. It can be seen that the band
intensity faded as the drug concentration increased. At
imipramine concentrations greater than 30 µM, the DNA band
disappeared completely.
Mass Spectral Analysis. Overnight incubation of the reaction
mixture containing imipramine, H2O2, and HRP at room
temperature or 1 h of incubation at 37 °C resulted in the
formation of a brown precipitate, which dissolved freely in
acetonitrile. The sample was analyzed with GC-MS. The
retention time of a major peak was 16 min. The mass spectrum
is shown in Figure 7A. The spectrum matched that of iminod-
ibenzyl (Figure 7B). The systematic name of the compound is
10,11-dihydro-5H-dibenz[b,f]azepine.
The addition of the excess of imipramine and H2O2 to the
HRP yielded a purple color which intensified with time. At pH

1502 Chem. Res. Toxicol., Vol. 23, No. 9, 2010
KorobkoVa et al.
7, a spectrum with a strong absorption maximum at 522 nm
appeared (Figure 5A). The same reaction conducted at pH 2
resulted in the appearance of a spectrum with λmax ) 650 nm
(Figure 5D). The spectrum was very unstable at pH 2 and
disappeared in 4 min. These observations are consistent with
the studies of imipramine mixtures with ceric ion and vanadyl
ion. EPR of transitory imipramine radical was detected in these
reactions (27, 28). In the present work, the formation of the
imipramine radical is supported by stoichiometry. HRP-I f
HRP-II transitions occur through the abstraction of one electron
from the reducing agent (26). In the case of imipramine, the
HRP-II was produced after the addition of one HRP equivalent
of the drug to the HRP-I solution (Figure 4A). This indicates
that each imipramine molecule loses one electron in this step,
suggesting the formation of an unpaired-electron species.
Criminal Justice. We cordially thank Professor Nathan Lents
for technical help and valuable discussions. We are grateful to
Professor Marcel Roberts for the help with the manuscript, Dr.
Steffen Jockusch for his constructive comments, and Dr. Arkadiy
Bolshov for a number of useful conversations. We also thank
anonymous referees for their many constructive remarks on the
original version of the manuscript.
References
(1) Yarosh, D. B. (1985) The role of O6-methylguanine-DNA methyl-
transferase in cell survival, mutagenesis, and carcinogenesis. Mutat.
Res. 145, 1–16.
(2) Friedberg, E. C. (2001) How nucleotide excision repair protects against
cancer. Nature ReV. Cancer 1, 22–33.
(3) Snyder, R. D., and Arnone, M. R. (2002) Putative identification of
functional interactions between DNA intercalating agents and topoi-
somerase II using the V79 in Vitro micronucleus assay. Mutat. Res.
503, 21–35.
(4) Snyder, R D., Ewing, D. E., and Hendry, L. B. (2004) Evaluation of
DNA intercalation potential of pharmaceutical and other chemicals
by cell-based and three-dimensional computational approaches. En-
Viron. Mol. Mutagen. 44, 163–173.
(5) Slamon, N. D., Ward, T. H., Butler, J., and Pentreath, V. W. (2001)
Assessment of DNA damage in C6 glicoma cells after antidepressant
treatment using an alkaline comet assay. Arch. Toxicol. 75, 243–250.
(6) Saxena, R., and Ahuja, Y. R. (1988) Genotoxicity evaluation of the
tricyclic antidepressants amitriptyline and imipramine using human
lymphocyte cultures. EnViron. Mol. Mutagen. 12, 421–430.
(7) Madrigal-Bujaidar, E., Madrigal-Santilla´n, E. O., Alvarez-Gonzalez,
I., Baez, R., and Pilar Marquez, P. (2008) Micronuclei induced by
imipramine and desipramine in mice: a subchronic study. Basic Clin.
Pharmacol. Toxicol. 103, 569–573.
Reactions between aromatic amines and hydroperoxides have
been studied in the presence of peroxidases; the appearance of
purple color was associated with the formation of the radical
cation. Griffin and co-workers (31) detected free radicals of
aminopyrine formed by microsomal cytochrome P450 in the
presence of cumene hydroperoxide. Free radicals of aminopyrine
have also been observed in the lipoxygenase/H2O2 system (32).
Various phenothiozines have been shown to produce radical
cations in the presence of HRP and H2O2, the maximum
wavelength varying between 500 and 630 nm (33, 34). Crystal
violet has been shown to produce free radicals in the reaction
with H2O2 catalyzed by HRP (35).
The final product in the reaction between imipramine and
HRP/H2O2 was identified as a dealkylated imipramine, imin-
odibenzyl (Figure 7C). The N-dealkylation mechanism was also
proposed for aromatic amines, aminopyrine (32) and crystal
violet (35). Hufford and co-workers identified iminodibenzyl
as an imipramine metabolite using fungal organisms (36).
(8) Dungaros, R., Turkbay, T., Surer, I., Gok, F., Denli, M., and Baltaci,
V. (2002) DNA damage in children treated with imipramine for
primary nocturnal enuresis. Pediatr. Int. 44, 617–621.
(9) Yamamoto, K., and Kawanishi, S. (1992) Site-specific DNA damage
by phenylhydrazine and phenelzine in the presence of Cu(II) ion or
Fe(III) complexes: roles of active oxygen species and carbon radicals.
Chem. Res. Toxicol. 5, 440–446.
(10) Balasubramanian, B., Pogozelski, W. K., and Tullius, T. D. (1998)
DNA strand breaking by the hydroxylradical is governed by the
accessible surface areas of the hydrogen atoms on the DNA back bone.
Proc. Natl. Acad. Sci.U.S.A. 95, 9738–9743.
(11) Henle, E. S., Han, Z., Tang, N., Rai, P., Luo, Y., and Linni, S. (1999)
Sequence-specific DNA cleavage by Fe2+-mediated Fenton reactions
has possible biological implications. J. Biol. Chem. 274, 962–971.
(12) Murata, M., Ohnishi, S., and Kawanishi, S. (2001) Acrylonitrile
enhances H2O2-mediated DNA damage via nitrogen-centered radical
formation. Chem. Res. Toxicol. 14, 1421–1427.
(13) Eghbal, M. A., Tafazoli, S., Pennefather, P., and O’Brien, P. J. (2004)
Peroxidase catalyzed formation of cytotoxic prooxidant phenothiazine
free radicals at physiological pH. Chem.-Biol. Interact. 151, 43–51.
(14) Ritter, C. L., Malejka-Giganti, D., and Polnaszek, C. F. (1983)
Cytochrome c/H2O2-mediated one electron oxidation of carcinogenic
N-fluorenylacetohydroxamic acids to nitroxyl free radicals. Chem.-
Biol. Interact. 46, 317–334.
Our experiments showed that oxidation of imipramine
catalyzed with HRP leads to the degradation of DNA (Figure
2A and B, lanes 8 and 9; Figure 3). Radical induced DNA
cleavage has been studied extensively with respect to OH
radicals formed in Fenton reactions (10, 11). DNA cleavage by
OH radicals has biotechnological applications (11). The anti-
depressant, 2-phenylethylhydrazine, was shown to induce DNA
strand breaks through the metabolically generated carbon-
centered radicals (37). The studies related to the interactions
between DNA and nitrogen-centered radicals, however, are very
limited. Murata and co-workers showed that the N-centered
radical formed from acrylonitrile enhances DNA damage
produced by H2O2 and Cu(II) (12).
(15) O’Brien, P. J. (2000) Peroxidases. Chem.-Biol. Interact. 129, 113–
139.
Genotoxicity of imipramine has been assessed in various
systems, and it has been found to cause significant DNA damage
in mice (7), human lymphocytes (6, 8), and C6 rat glioma cells
(5). Saxena and Ahuja performed comparative studies of
genotoxicity of imipramine and amitriptyline in human lym-
phocyte cultures (6). They found that amitriptyline caused
chromosome damage at concentrations significantly greater than
those attained in patients under normal therapy. However,
imipramine exhibited a substantial increase in chromosome
aberrations at the upper level of therapeutic doses, 500 ng/mL
(∼2 µM). Our studies indicate that imipramine may cause
significant DNA degradation at concentrations of 2-5 µM
(Figure 6). Amitriptyline shows no effect on DNA at the same
conditions. The damage caused by imipramine is associated with
the formation of a very reactive imipramine intermediate.
(16) O’Brien, P. J. (1988) Radical formation during the peroxydase
catalyzed metabolism of carcinogens and xenobiotics: The reactivity
of these radicals with GSH, DNA, and unsaturated lipid. Free Radical
Biol. Med. 4, 169–183.
(17) Miura, T., Muraoka, S., and Fujimoto, Y. (2001) Inactivation of
creatine kinase during the interaction of indomethacin with horseradish
peroxidase and hydrogen peroxide: Involvement of indomethacin
radicals. Chem.-Biol. Interact. 134, 13–25.
(18) O’Brien, P. J. (1985) Free-radical-mediated DNA binding. EnViron.
Health. Perspect. 64, 219–232.
(19) Mittmann, N. M., Lui, B. A., Bparm, M. I., Bradley, C. A., Pless, R.,
Shear, N. H., and Einarson, T. R. (1997) Drug-related mortality in
Canada (1984-1994). Pharmacoepidemiol. Drug Saf. 6, 157–168.
(20) Caplan, Y. H., Ottinger, W. E., Jongsei, P., and Smith, T. D. (1985)
Drug and chemical related deaths: Incidence in the state of Maryland.
J. Forensic Sci. 30, 1012–1021.
(21) Song, H.-Y., Yao, J.-H., Liu, J.-Z., Zhou, S.-J., Xiong, Y.-H., and Ji,
L.-N. (2005) Effects of phthalic anhydride modification on horseradish
peroxidase stability and structure. Enzyme Microb. Technol. 36, 605–
611.
Acknowledgment. This work was supported by a PSC-
CUNY grant and the PRISM program at John Jay College of
(22) Suh, D., and Chaires, J. B. (1995) Criteria for the mode of binding of
DNA binding agents. Bioorg. Med. Chem. 3, 723–728.

Interactions of Tricyclic Antidepressants with DNA
Chem. Res. Toxicol., Vol. 23, No. 9, 2010 1503
(23) Dunford, H. B., and Stillman, J. S. (1976) On the function and
mechanism of action of peroxidases. Coord. Chem. ReV. 19, 187–
251.
(24) Aitkem, S. M., Ouellet, M., Percival, M. D., and English, A. M. (2003)
Mechanism of Horseradish peroxidase inactivation by benzhydrazide:
a critical evaluation of arylhydrazides as peroxidase inhibitors.
Biochem. J. 375, 613–621.
(25) Ator, M., and Ortiz de Montellano, P. R. (1987) Protein control of
prosthetic heme reactivity. J. Biol. Chem. 262, 1542–1551.
(26) Ortiz de Montellano, P. R. (1986) In Cytochrome P-450: Structure,
Mechanism and Biochemistry (Ortiz de Montellano, R. P. Ed.) Plenum
Press, New York.
(27) Borg, D. C. (1965) Free radicals from imipramine. Biochem. Phar-
macol. 14, 115–120.
(28) Saakurai, H., and Goda, T. (1983) Formation of imipramine free radical
and vanadyle ion in the reaction of imipramine and pentavalent
vanadium ion. Inorg. Chim. Acra 78, L33–L34.
(29) Edwards, S. L., Raag, R., Wariishi, H., Gold, M. H., and Poulos, T. L.
(1993) Crystal structure of lignin peroxidase. Proc. Natl. Acad.
Sci.U.S.A. 90, 750–754.
(30) Sivaraja, M., Goodin, D. B., Smith, M., and Hoffman, B. M. (1989)
Identification by ENDOR of Trp191 as the free-radical site in
cytochrome c peroxidase compound ES. Science 245, 738–740.
(31) Griffin, B. W., Marth, C., Yasukochi, Y., and Masters, B. S. S. (1980)
Radical mechanism of aminopyrine oxidation by cumene hydroper-
oxide catalyzed by purified liver microsomal cytochrome P-450. Arch.
Biochem. Biophys. 205, 543–553.
(32) Perez-Gilabert, M., Sanchez-Ferrer, A., and Garcia-Carmona, F. (1997)
Oxidation of aminopyrine by the hydroperoxidase activity of lipoxy-
genase: a new proposed mechanism of N-demethylation. Free Radical
Biol. Med. 23, 548–555.
(33) Vazquez, A., Tudela, J., Varon, R., and Garcia-Canovas, F. (1991)
Determination of the molar absorptivities of phenothiazine cation
radicals generated by oxidation with hydrogen peroxide/peroxidase.
Anal. Biochem. 202, 245–248.
(34) Sackett, P. H., Mayausky, J. S., Smith, T., Kalus, S., and McCreery,
R. L. (1981) Side-chain effects on phenothiazine cation radical
reactions. J. Med. Chem. 24, 1342–1347.
(35) Gadelha, F. R., Hanna, P. M., Mason, R. P., and Docampo, R. (1992)
Evidence for free radical formation during horseradish peroxidase-
catalyzed N-demethylation of crystal violet. Chem.-Biol. Interact. 85,
35–48.
(36) Hufford, C. D., Capiton, G. A., Clark, A. M., and Baker, J. K. (1981)
Metabolism of imipramine by microorganisms. J. Pharm. Sci. 70, 151–
155.
(37) Saez, G., Thornalley, P. J., Hill, H. A. O., Hems, R., and Bannister,
J. V. (1982) The production of free radicals during the authoxidation
of cysteine and their effect on isolated rat hepatocytes. Biochim.
Biophys. Acta 719, 24–31.
TX100221B