The design of agents to control DNA methylation adducts. Enhanced major groove methylation of DNA by an N-Methyl-N-nitrosourea functionalized phenyl neutral red intercalator

Article abstract of DOI:10.1021/tx960007n

An N-methyl-N-nitrosourea (MNU) moiety [CH₃N(N=O)C(=O)NH-] linked to the C4'-position of the 5-substituted phenyl ring of phenyl neutral red (PNR), 2-methyl-3-amino-5-[p-[[2-[(N-nitroso-N- methylcarbamoyl)amino]ethyl]carbamoyl]phenyl]-7-(dimethylamino)phenazenium chloride (MNU-PNR), has been synthesized as an approach to design a molecule that will deliver alkylating agents with some preference to guanine (Gua) in the major groove of DNA. The PNR nucleus was chosen because previous studies suggested the following: (1) PNR binds with a slight preference for G/C rich sequences; and (2) PNR intercalates into DNA from the major groove with the 5-phenyl ring pointing out into the major groove (Muller, W., Bunemann, H., and Dattagupta, N. (1975) Eur. J. Biochem. 54, 279-291). It is demonstrated that MNU-PNR yields 2.6 and 6.0 times more N7-methylguanine (7-MeGua) than MNU at low salt (10 mM Tris buffer) and high salt (10 mM Tris buffer + 200 mM NaCl), respectively. It is also shown that the ratio of 7-MeGua (a major groove adduct) to N3-methyladenine (a minor groove adduct) is approximately 5 times higher for MNU-PNR than for MNU. The yield of the 7-MeGua adduct is decreased by the coaddition of a nonmethylating analogue of MNU-PNR or NaCl, but increased in the presence of the minor groove intercalator, ethidium bromide. Using a ³²P-end-labeled restriction fragment, the enhanced methylation by MNU-PNR at 7-Gua is confirmed, and it is demonstrated that the sequence-dependent formation of 7-MeGua from MNU-PNR is the same as that seen with MNU. UV, circular dichroism, and viscosity studies are consistent with MNU-PNR binding to DNA via an intercalation-based process.

Full text of DOI:10.1021/tx960007n

Chem. Res. Toxicol. 1996, 9, 939-948
939
Th e Design of Agen ts To Con tr ol DNA Meth yla tion
Ad d u cts. En h a n ced Ma jor Gr oove Meth yla tion of DNA
by a n N-Meth yl-N-n itr osou r ea F u n ction a lized P h en yl
Neu tr a l Red In ter ca la tor
Pratibha Mehta,† Kevin Church,‡ J onathan Williams,†,§ Fa-Xian Chen,†
Lance Encell,†,| David E. G. Shuker, and Barry Gold*,†,|
Eppley Institute for Research in Cancer and Allied Diseases and Department of Pharamceutical
Sciences, University of Nebraska Medical Center, Omaha, Nebraska 68198,
Department of Chemistry, University of Dayton, Dayton, Ohio 45469-2357, and
MRC Toxicology Unit, University of Leicester, Lancaster Road, Leicester LE1 9HN, United Kingdom
Received J anuary 17, 1996X
An N-methyl-N-nitrosourea (MNU) moiety [CH3N(NdO)C(dO)NH-] linked to the C4′-
position of the 5-substituted phenyl ring of phenyl neutral red (PNR), 2-methyl-3-amino-5-[p-
[[2-[(N-nitroso-N-methylcarbamoyl)amino]ethyl]carbamoyl]phenyl]-7-(dimethylamino)phena-
zenium chloride (MNU-PNR), has been synthesized as an approach to design a molecule that
will deliver alkylating agents with some preference to guanine (Gua) in the major groove of
DNA. The PNR nucleus was chosen because previous studies suggested the following: (1)
PNR binds with a slight preference for G/C rich sequences; and (2) PNR intercalates into DNA
from the major groove with the 5-phenyl ring pointing out into the major groove (Mu¨ller, W.,
Bu¨nemann, H., and Dattagupta, N. (1975) Eur. J . Biochem. 54, 279-291). It is demonstrated
that MNU-PNR yields 2.6 and 6.0 times more N7-methylguanine (7-MeGua) than MNU at
low salt (10 mM Tris buffer) and high salt (10 mM Tris buffer + 200 mM NaCl), respectively.
It is also shown that the ratio of 7-MeGua (a major groove adduct) to N3-methyladenine (a
minor groove adduct) is approximately 5 times higher for MNU-PNR than for MNU. The yield
of the 7-MeGua adduct is decreased by the coaddition of a nonmethylating analogue of MNU-
PNR or NaCl, but increased in the presence of the minor groove intercalator, ethidium bromide.
Using a 32P-end-labeled restriction fragment, the enhanced methylation by MNU-PNR at 7-Gua
is confirmed, and it is demonstrated that the sequence-dependent formation of 7-MeGua from
MNU-PNR is the same as that seen with MNU. UV, circular dichroism, and viscosity studies
are consistent with MNU-PNR binding to DNA via an intercalation-based process.
cells defective in a particular type of repair pathway
respond, in terms of cytotoxicity and/or mutagenicity, to
an alkylating agent (4-8). Both strategies have provided
remarkable insights into the consequences of DNA alky-
lation. The knowledge gained from such studies can be
useful in the design of antineoplastic drugs that are
cytotoxic but not mutagenic (2).
In tr od u ction
Research directed toward understanding the role(s) of
different types of DNA modifications in cytotoxicity and
mutagenicity is of ongoing interest (1, 2). Because most
alkylating agents, including carcinogens and antineo-
plastic agents, modify DNA at multiple sites, it is not
easy to dissect the contribution of individual DNA lesions
toward the two biological end points. The issue of adduct
complexity has generally been addressed in two ways.
The preparation of site-specifically alkylated DNA tar-
gets, which can be transfected into cells, provides infor-
mation on their miscoding (promutagenic) properties
upon in vivo replication (3). This approach requires that
the adduct be relatively stable so that it is compatible
with the chemical or biological procedures used to
introduce lesions into DNA, and the subsequent trans-
fection into cells. The other method is to determine how
We have taken a somewhat different approach to
elucidate the biological roles of different adducts in
cytotoxicity and mutagenicity, by designing compounds
that will sequence and/or groove selectively modify DNA
in vitro and in vivo. Previously, minor and major groove
equilibrium binding compounds appended with alkylat-
ing agents have been prepared that significantly alter
the agent’s “normal” alkylation pattern and groove
specificity (9-19). As a consequence of their unusual
alkylating properties, some of these compounds have
shown interesting biological activities (8). In the current
work, the goal was to design a methylating agent that
would preferentially increase the yield of the N7-meth-
ylguanine (7-MeGua),1 the chemically unstable major
groove lesion.
* To whom correspondence should be addressed: telephone, 402-
559-5148; FAX, 402-559-4651; Email, bgold@unmc.edu.
† Eppley Institute for Research in Cancer and Allied Diseases,
University of Nebraska Medical Center.
‡ University of Dayton.
Intercalation is one of the basic mechanisms by which
small molecules equilibrium bind to DNA (20). For
reasons not clear, the predominant route of entry of
intercalators into the DNA base pair stack is via the
minor groove. One of the proposed exceptions to this
§ Present address: Aston Molecules, 10 Hold Ct. S., Birmingham
B7 4EJ , U.K.
| Department of Pharmaceutical Sciences, University of Nebraska
Medical Center.
University of Leicester.
X Abstract published in Advance ACS Abstracts, J uly 15, 1996.
S0893-228x(96)00007-0 CCC: $12.00 © 1996 American Chemical Society

940 Chem. Res. Toxicol., Vol. 9, No. 6, 1996
Mehta et al.
Exp er im en ta l Section
Ca u tion . MNU-PNR and MNU should be considered toxic
and carcinogenic, and treated accordingly.
Gen er a l P r oced u r es. 1H NMR were recorded on a Varian
XL-300 spectrometer and 2D-NMR techniques (COSY and phase
sensitive DQCOSY) were used in structural assignments. Mass
spectral data were performed on an AEI MS-9 or Kratos MS-
50 spectrometer utilizing electron impact (EI-MS) or fast-atom-
bombardment (FAB-MS) ionization techniques. IR spectra were
obtained on an Mattson Alfa Centauri FT-IR, CD spectra on a
J asco J 710 spectropolarimeter, and UV spectra on a Varian
DMS 300 spectrophotometer. Analytical and preparative silica
TLC employed 0.25 and 2.0 mm Merck Kieselgel glass plates.
All flash column chromatography was performed with 40 µm
silica gel.
Ma ter ia ls. All chemical and biochemical reagents were
purchased from Aldrich Chemicals (Milwaukee, WI) and Sigma
Chemicals (St. Louis, MO) and used without further purification
unless stated otherwise. DMF was passed over a column of 4
Å molecular sieves and dried by sequential azeotropic distilla-
tion with benzene and reduced pressure distillation from BaO.
DMSO was freshly distilled from CaH2 after initial drying over
4 Å molecular sieves. Ethylenediamine was dried over KOH
and distilled. Salmon sperm DNA (Sigma) used in the DNA
methylation studies was dialyzed against 100 mM Tris-HCl (pH
8.0) buffer overnight.
Molecu la r Mod elin g. MNU-PNR was constructed and
minimized using the Dreiding force field (27). The coordinates
for the bis-intercalator ditercalinium complexed with 5′-
d(CGCG)2 were imported from the Brookhaven Data Base (28)
into Sybyl (Tripos Associates, St. Louis, MO). Ditercalinium
was chosen because it enters the DNA base pair stack from the
major groove (28, 29). The ditercalinium molecule was deleted
as was one of the two intercalation sites. The remaining d(CG)2
duplex and MNU-PNR were treated as separate aggregates in
all the calculations. The MNU-PNR was aligned with the
intercalation site and slowly moved into the DNA stack using
a distance constraint while monitoring van der Waals energies.
When the van der Waals energy of the system started to rise,
the analysis was halted. The resulting structure was then
minimized using the Kollman Unified-Atom force field (30) with
no restraints. This process was repeated to obtain the lowest
energy structure shown in Figure 2. A similar analysis was
done with MNU-PNR approaching the same intercalation site
from the minor groove (data not shown).
Syn th esis (see F igu r e 3). (A) N-(2-Am in oeth yl)-4-a m i-
n oben za m id e (3). Cyclohexene (10 mL) and Pd/C 10% (1.0 g)
were added to a stirred solution of nitro derivative 2 (2.09 g, 10
mmol) in absolute EtOH (50 mL), and the mixture was heated
to reflux for 72 h. The solvent was concentrated in vacuo to
afford compound 3 as a yellow oil (62%): TLC (MeOH/CH2Cl2/
AcOH, 1:4:trace) Rf 0.20; NMR (DMSO-d6) δ 2.62 (t, 2 H, CH2-
NH2), 3.20 (d-t, 2 H, CH2NHCO), 6.54 (d, 2 H, aryl CH), 7.60
(d, 2 H, aryl CH), 8.00 (t, 1 H, CONHCH2); UV (H2O) λmax (log
ꢀ) 276 nm (3.78); EI-MS m/z 180 (M + 1), 179 (M), 164 (M -
NH2), 121 (M - NHCH3), 93.
(B) N-(N-Ca r boben zoxy-2-a m in oeth yl)-4-a m in oben za -
m id e (4). A solution of 3 (1.6 g, 8.9 mmol) and Et3N (5 mL) in
MeOH (15 mL) was cooled to 0 °C, and benzyl chloroformate
(1.6 g, 9 mmol) was slowly added. After the addition at 0 °C,
the reaction was maintained at room temperature for 16 h. The
solvent was evaporated and the residue dissolved in CH2Cl2 (25
mL), washed with water (3 × 15 mL), dried (Na2SO4), and
concentrated to give crude product 4. The pure compound was
obtained after flash silica column chromatography (benzene/
CH2Cl2, 3:7) in 72% yield: mp 128-130 °C; TLC (MeOH/CH2-
Cl2, 1:19) Rf 0.6; NMR (DMSO-d6) δ 3.17 (d-t, 2 H, CH2NHCO),
3.29 (d-t, 2 H, CH2NHCO2), 5.05 (s, 2 H, OCH2Ar), 6.55 (d, 2 H,
aryl CH), 7.37 (m, 5 H, aryl CH), 7.59 (d, 2 H, aryl CH), 8.05 (t,
1 H, CONH); UV (MeOH) λmax (log ꢀ) 276 nm (4.05); FAB-MS
m/z 314 (M + 1), 313 (M+), 277, 241, 210, 185.
F igu r e 1. Structure of PNR, MNU-PNR, and ethidium.
generalization is the intercalating dye 2-methyl-3-amino-
5-phenyl-7-(dimethylamino)phenazenium chloride (phen-
yl neutral red, PNR) (see Figure 1 for structures) that
has been reported to associate with the major groove of
DNA (21-23). The width of PNR (∼9.5 Å from the
carbon of the exocyclic 2-methyl to the ring C-8) physi-
cally prevents minor groove access. In addition to its
proposed intercalation from the major groove, PNR binds
with some preference to G:C rich regions (21). Using this
knowledge on groove orientation and binding preference,
an N-methyl-N-nitrosourea [CH3N(NdO)NH-] function-
alized derivative of PNR was synthesized with the goal
of preferentially delivering a methylating agent to the
major groove of DNA. Since N-methylnitrosoureas (MNU)
hydrolyze at physiological pH to the highly reactive
alkylating agent methanediazonium ion (24-26), it was
anticipated that the regiospecific generation of meth-
anediazonium in the major groove would lead to an
increase in the yields of major groove adducts, e.g.,
7-MeGua, relative to minor groove adducts, e.g., N3-
methyladenine (3-MeAde). The association of this dye
with DNA is described along with its DNA methylating
characteristics.
1
Abbreviations: CD, circular dichroism; EB, ethidium bromide; EI-
MS, electron impact mass spectrometry; ELISA, enzyme-linked im-
munosorbent assay; FAB-MS, fast-atom-bombardment mass spectrom-
etry; 3-MeAde, N3-methyladenine; 7-MeGua, N7-methylguanine; MNU,
N-methylnitrosourea; MNU-PNR, 2-methyl-3-amino-5-[p-[[2-[(N-ni-
troso-N-methylcarbamoyl)amino]ethyl]carbamoyl]phenyl]-7-(dimethyl-
amino)phenazenium chloride; MU-PNR, 2-methyl-3-amino-5-[p-[[2-[(N-
m et h ylca r ba m oyl)a m in o]et h yl]ca r ba m oyl]ph en yl]-7-(dim et h yl-
amino)phenazenium chloride; PNR, 2-methyl-3-amino-5-phenyl-7-
(dimethylamino)phenazenium chloride (phenyl neutral red).
(C) N,N-Dim eth yl-4,4′-d ia m in od ip h en yla m in e Ch lor id e
(7). N,N-Dimethyl-p-phenylenediamine (5, 210 mg, 1 mmol)

MNU-PNR
Chem. Res. Toxicol., Vol. 9, No. 6, 1996 941
reaction mixture was cooled and concentrated and the crude
product purified by silica column chromatography (MeOH/CH2-
Cl2, 3:47) to give phenyl neutral red derivative 8 in 5% yield:
mp >300 °C; TLC (MeOH/CH2Cl2, 1:9) Rf 0.45; IR (KBr disc)
3833, 3771, 2983, 1700, 741 cm-1; NMR (DMSO-d6) δ 2.30 (s, 3
H, CH3Ar), 3.04 (s, 6 H, (CH3)2N), 3.25 (q, 2 H, CH2NHCO),
3.40 (q, 2 H, CH2NHCO2), 5.03 (s, 2 H, OCH2Ar), 5.62 (s, 1 H,
aryl CH), 5.92 (s, 1 H, aryl CH), 7.35 (m, 5 H, aryl CH), 7.40 (t,
1 H, CONH), 7.52 (d, 1 H, aryl CH), 7.74 (d, 2 H, aryl CH), 7.80
(s, 1 H, aryl CH), 8.03 (d, 1 H, aryl CH), 8.32 (d, 2 H, aryl CH),
8.90 (t, 1 H, CONH); UV (MeOH) λmax 551 nm; FAB-MS m/z
550 (M + 1), 416, 369, 135.
(E) 2-Meth yl-3-a m in o-5-[p-[(2-a m in oeth yl)ca r ba m oyl]-
p h en yl]-7-(d im eth yla m in o)p h en a zen iu m Ch lor id e (9). The
protected phenyl neutral red derivative 8 (40 mg, 72.4 µmol)
was dissolved in absolute EtOH (5 mL) containing anhydrous
HOAc (2 mL), Pd/C (10%, 20 mg), and cyclohexene (5 mL). The
reaction mixture was stirred for 48 h under Ar, the catalyst
removed by filtration, and the residue washed with EtOH (3 ×
5 mL). The combined filtrate and washings were concentrated
to give crude product which was purified by silica column
chromatography using MeOH with HOAc (0.1%) as eluant.
Amine 9 was obtained in 70% yield: mp >300 °C; TLC (MeOH/
CH2Cl2, 1:4), Rf 0.2; IR (KBr disc) 3524, 3431, 3137, 3076, 2952,
1607, 1514, 1344, 1190, 1020, 865, 757 cm-1; NMR (DMSO-d6)
δ 2.32 (s, 3 H, CH3Ar), 2.80 (t, 2H, CH2NH2), 3.07 (s, 6 H,
(CH3)2N), 3.80 (m, 2 H, CH2NHCO), 5.66 (s, 1 H, aryl CH), 5.95
(s, 1 H, aryl CH), 7.54 (s, 1 H, aryl CH), 7.76 (d, 2 H, aryl CH),
7.91 (s, 1 H, aryl CH), 8.07 (d, 1 H, aryl CH), 8.37 (d, 2 H, aryl
CH), 8.92 (t, 1H, CONH); UV (MeOH) λmax (log ꢀ) 549.9 nm
(4.21); FAB-MS m/z 416 (M + 1), 415 (M+), 353.
(F ) 2-Met h yl-3-a m in o-5-[p -[[2-[(N-m et h ylca r b a m oyl)-
am in o]eth yl]car bam oyl]ph en yl]-7-(dim eth ylam in o)ph en a-
zen iu m Ch lor id e (MU-P NR). A solution of 9 (50 mg, 120
mmol) in 50% aqueous dioxane (4 mL) and Et3N (12 mg, 120
µmol) was added to a solution of 1,2,2,2-tetrachloroethyl N-
methylcarbamate (31) (25 mg, 120 µmol) in dioxane (2 mL) at 0
°C. The reaction mixture was stirred for 8 h and concentrated
in vacuo, and product MU-PNR, after purification by silica
column chromatography (MeOH/CH2Cl2, 1:19), was obtained in
38% yield: TLC (MeOH/CH2Cl2, 2:23) Rf 0.40; IR (KBr disc)
3462, 3354, 2937, 1623, 1530, 1190, 587 cm-1; NMR (DMSO-
d6) δ 2.31 (s, 3 H, CH3Ar), 2.56 (d, 3 H, CH3NH), 3.06 (s, 6 H,
(CH3)2N), 3.24 (m, 4 H, CH2’s), 5.63 (s, 1 H, aryl CH), 5.95 (s, 1
H, aryl CH), 6.07 (q, 1 H, CONHCH3), 6.40 (t, 1 H, CH2NH),
7.53 (d, 1 H, aryl CH), 7.74 (d, 2 H, aryl CH), 7.87 (s, 1 H, aryl
CH), 8.02 (d, 1 H, aryl CH), 8.38 (d, 2 H, aryl CH), 9.08 (t, 1 H,
CONHCH2); UV (MeCN) λmax (log ꢀ) 548 nm (4.33), 214 (4.31);
FAB-MS m/z 473 (M + 1), 472 (M+), 458 (-CH3), 443 (-NHCH3).
(G) 2-Meth yl-3-a m in o-5-[p-[[2-[(N-n itr oso-N-m eth ylca r -
ba m oyl)a m in o]eth yl]ca r ba m oyl]p h en yl]-7-(d im eth yla m i-
n o)p h en a zen iu m Ch lor id e (MNU-P NR). A solution of 9 (25
mg, 60 mmol) in dry DMF (2 mL) containing diisopropylamine
(742 mg, 26 mmol, 50 µL) was added to a solution of 1,2,2,2-
tetrachloroethyl N-methyl-N-nitrosocarbamate (31) at -5 °C
under Ar and the solution stirred for 7 h at this temperature.
The solvent and volatile reaction byproducts were then evapo-
rated in vacuo at 35 °C, and the resulting solid material was
triturated with MeOH (0.5 mL) and product MNU-PNR isolated
by precipitation from MeOH with EtOAc and washing with the
same solvent. Yield, 41%: TLC (MeOH/CH2Cl2/HOAc, 3:47:
trace) Rf 0.30; IR (KBr disc) 3431, 2983, 2353, 1623, 1530, 1344,
1090, 1020 cm-1; NMR (DMSO-d6) δ 2.35 (s, 3 H, CH3Ar), 3.08
(s, 6 H, (CH3)2N), 3.12 (s, 3 H, CH3N), 3.60 (m, 4 H, CH2’s), 5.64
(s, 1 H, aryl CH), 5.98 (s, 1 H, aryl CH), 7.55 (d, 1 H, aryl CH),
7.78 (d, 2 H, aryl CH), 7.95 (s, 1 H, aryl CH), 8.08 (d, 1 H, aryl
CH), 8.35 (d, 2 H, aryl CH), 8.90 (b-s, 1 H, CONH), 9.08 (b-s, 1
H, CONH); UV (MeOH) λmax (log ꢀ) 549 nm (4.34); FAB-MS m/z
502 (M + 1), 472 (-NO), 443 (M - NNOCH3), 415 (M -
CONNOCH3); high resolution FAB-MS found 501.23650 (+0.4
ppm from value calculated for C26H29O3N8).
F igu r e 2. Relaxed stereoscopic views of MNU-PNR docked into
5′-d(CG)2 intercalation site from ditercalinium-DNA complex
(28): (a) from the major groove highlighting the position of the
C-2 methyl (filled-in atoms) of PNR and the backbone; (b) from
the major groove highlighting the position of the N7-(dimeth-
ylamino) group (filled-in atoms) of PNR and the backbone; (c)
looking down the helical axis at major groove complex; and (d)
major groove MNU-PNR intercalated complex viewed from the
major groove.
and o-toluidine hydrochloride (6, 144 mg, 1 mmol) were dissolved
in H2O (20 mL), and then Na2Cr2O7 (600 mg) in 5 mL of H2O
was added slowly with stirring. The green precipitate that
formed was collected and thoroughly washed with H2O and then
dried to afford 7 in 67% yield: mp >300 °C; TLC (MeOH/CH2-
Cl2, 1:4), Rf 0.5; NMR (DMSO-d6) δ 1.95 (s, 3 H, CH3Ar), 2.92
(s, 6 H, (CH3)2N), 6.80 (m, 7 H, aryl CH); FAB-MS m/z 241 (M
+ 1), 240 (M).
(D) 2-Meth yl-3-a m in o-5-[p-[[2-(N-ca r boben zoxya m in o]-
et h yl]ca r b a m oyl]p h en yl]-7-(d im et h yla m in o)p h en a zen i-
u m Ch lor id e (8). To a stirred suspension of 7 (241 mg, 1 mmol)
in water (25 mL) adjusted to pH 5.0 was added 4 (362 mg, 11.6
mmol) in MeOH (70 mL). A pH of 5.0 was maintained with
HCl and the mixture heated to boiling. After 10 min the color
had changed from green to blue and then to violet-purple. The
Ad d u ct An a lysis by HP LC/UV. Salmon sperm DNA (1 mM
phosphate) was reacted with MNU (1 mM) or MNU-PNR (1 mM)

942 Chem. Res. Toxicol., Vol. 9, No. 6, 1996
Mehta et al.
at room temperature for 24 h in Tris-HCl buffer (10 mM, pH
8.0), with or without 200 mM NaCl. The DNA was precipitated
from the reaction mixture with 3 M NaOAc and EtOH, and
cooling in ice. The precipitated DNA was washed with EtOH,
redissolved in 500 µL of 10 mM Tris buffer (pH 7.0), and then
heated at 90 °C for 30 min to release 3- and 7-alkylpurines (32).
The partially apurinic DNA was precipitated at 0 °C by adding
0.1 volume of 0.1 N HCl and the supernatant analyzed on HPLC
using a Whatman Partisil-10 SCX strong cation exchange
column with 175 mM ammonium formate buffer (pH 3.0) as the
eluant at a flow rate of 1 mL/min. The 7-MeGua and 3-MeAde
compounds were monitored by UV detection at 270 nm. Au-
thentic standards were used to confirm the retention time of
the two adducts and to calculate response factors. To correct
for DNA recovery, the yields of the two adducts are presented
as a ratio of adduct to unmodified Ade. Ade was measured by
removing an aliquot of the DNA immediately before the neutral
thermal hydrolysis (90 °C for 30 min at pH 7.0) and treating it
with 0.1 N HCl at 70 °C for 30 min to depurinate the DNA.
After precipitation of the apurinic DNA, the supernatant was
filtered and analyzed by HPLC to quantitate Ade using the same
conditions described above for the adducts.
3-MeAd e An a lysis by Com p etitive ELISA. The MNU or
MNU-PNR (1 mM) was incubated with salmon sperm DNA (100
µM phosphate) in 10 mM Tris-HCl buffer (pH 8.0) for 24 h at
room temperature. At the end of the incubation, the reactions
were heated at 90 °C for 30 min to release 3-MeAde (and other
heat labile adducts) from the DNA. After cooling in an ice bath,
the partially apurinic DNA was precipitated with ice cold 0.1
N HCl (32). The supernatant was concentrated in vacuo and
the residue dissolved in 10 mM Tris-HCl buffer (pH 7.4)
containing 0.9% NaCl, 2.5 mM MgCl2, and 0.1% bovine serum
albumin. Samples were assayed for 3-MeAde using a specific
antibody in a competitive ELISA (33). The N3-methyl-N6-[5-
(carboxypentyl)ovalbumin]adenine used as the coating antigen
was prepared as previously described (33). Standard ELISA
inhibition curves using authentic 3-MeAde adduct were gener-
ated simultaneously with the samples for each 96 well plate.
The standard curve was calculated using SigmaPlot 4.01 (J andel
Scientific, San Rafael, CA) by least squares fitting of the data
to the equation:
top of a 12% polyacrylamide (7.8 M urea) denaturing gel and
the gel run at 75 W (∼55 °C). The standard Maxam-Gilbert
Gua and Gua+Ade reaction lanes were included as sequence
markers (36). The gels were either: (a) exposed to Kodak
X-OMAT AR film at -70 °C and the resulting autoradiogram
analyzed using a Shimadzu CS-9000 scanning densitometer; or
(b) analyzed using a Molecular Dynamics PhosphorImager.
UV-Vis Stu d ies. The UV-vis spectra of MU-PNR in the
presence and absence of salmon sperm DNA were obtained in
10 mM Tris-HCl buffer (pH 8.0) with or without 200 mM NaCl
at room temperature.
CD Stu d ies. The CD spectra (220-330 nm) of salmon sperm
DNA (200 µM nucleotide) in the absence and presence of 76 µM
MU-PNR in 13 mM Tris-HCl buffer (pH 8.0), with and without
200 mM NaCl, were obtained (32 scans/run) at 20 °C.
DNA Viscosity Stu d ies. Viscometry experiments were
performed using a 10 mL Cannon-Ostwald No. 80-100 viscom-
eter. Calf thymus DNA (15 mg) (Sigma Type 1, highly polym-
erized) was suspended in 100 mL of 5 mM Tris-HCl buffer (pH
7.5) containing 1 mM NaCl and sonicated for 4 h. Final DNA
concentrations were determined by UV at 260 nm (ꢀ, 6600). The
melting temperature of this DNA was 55 °C, which agrees with
literature references (38). DNA solutions (350 µM phosphate)
were filtered into the viscometer, and the viscometer was
maintained at 27 °C by immersion in a thermostated water bath.
Flow times were measured with a stop watch and experiments
run in triplicate to give standard deviations of <0.1 s. The flow
times of the buffer and buffered DNA were 90.6 and 101.1 s,
respectively. The compounds (MU-PNR, ethidium bromide)
were dissolved in MeOH (5% final concentration by volume) and
added to the viscometer with a Hamilton syringe. In the
concentration range of compound studied, the MeOH had no
effect on DNA viscosity. The compound-DNA solutions were
mixed by gentle bubbling with air. The solutions reached
equilibrium within 3-4 min after addition of compound. Flow
times for buffer, DNA, and compound-DNA complex were fitted
to the equation for relative helix extension of Cohen and
Eisenberg (39, 40):
L/L0 ) {[tcompd - tbuffer(V)]/[tDNA - tbuffer(V)]}1/3
where L0 and L are the apparent length of DNA in the absence
and presence of compound, respectively, tcompd is the flow time
for the compound-DNA complex, tDNA is the flow time for pure
DNA, and tbuffer is the flow time for buffer with a given volume
V. All plots were forced through the origin intersection corre-
sponding to L/L0 in the absence of compound.
P ) YMIN + (YMIN - YMAX)(In/Kn)/(1 + In/Kn)
where P is the % inhibition, I is the amount of inhibitor (3-
MeAde), YMIN and YMAX are the y-axis values where the curve
plateaus, n is the Hill coefficient, and K is the dissociation
constant (34). The adduct levels are expressed as 3-MeAde/Ade
where Ade is based on the 100 µM phosphate DNA concentration
used in the incubation.
Resu lts
Mod elin g. The intercalation complex of MNU-PNR
with DNA was modeled using the atomic coordinates
from the Brookhaven Protein Data Bank of an intercala-
tion site derived from the crystal structure of diter-
calinium-d(CGCG)2 (28, 29). The ditercalinium was
deleted and remaining DNA used as a host for MNU-
PNR. A final minimum energy structure (the 5′-CG-3′:
3′-GC-5′ region of the DNA that is not involved in the
intercalation complex has been removed for viewing) is
shown in Figure 2, and it is clear that the 2-methyl
(Figure 2a) and N7-(dimethylamino) (Figure 2b) groups
prevent the dye from penetrating any further into the
base pair stack. In the perspective looking down the
helix axis (Figure 2c), it is apparent that the dye can
interact with the π-systems of the Cyt that are on
opposing strands in the CpG intercalation site. It should
be noted that stabilization of the dye with DNA via π-π
stacking interactions is not well modeled in Sybyl. The
same interstrand base pair overlap is seen in at a
modeled (dG)2:(dC)2 dimer: the 5′-base from the two
complementary strands is stacked with the PNR nucleus
Meth yla tion of Restr iction F r a gm en t. A 5′-end-labeled
576 bp fragment was prepared from pBSCPV A1 plasmid
containing the promotor region for the coat protein of the canine
parvovirus (35) by sequential NcoI endonuclease digestion,
treatment of the linearized DNA with alkaline phosphatase,
phosphorylation with [γ-32P]ATP in the presence of T4 poly-
nucleotide kinase, and restriction with HindIII endonuclease
(36). The labeled fragment was purified on 5% polyacrylamide
gel prior to use. The restriction fragment (80 000 cpm) and
sonicated calf thymus DNA (final concentration, 100 µM phos-
phate) were dissolved in 10 mM sodium cacodylate buffer (pH
7.8) containing the desired concentration of NaCl or DNA
affinity binder. This DNA solution was incubated at 37 °C for
a specified time with freshly prepared solutions of MNU or
MNU-PNR. The reactions were terminated by cooling in ice
water and precipitation of the DNA with NaOAc and EtOH. The
DNA was washed with cold 70% EtOH and dried in vacuo.
Strand breaks in the modified DNA were generated by neutral
thermal hydrolysis (37) as previously described (14). The DNA
was suspended in loading buffer (80% deionized formamide, 50
mM Tris-borate (pH 8.3), 1 mM EDTA, 0.1% xylene cyanol,
and 0.1% bromphenol blue) and denatured by heating at 90 °C
for 1 min and cooling in ice. The DNA was placed into wells on

MNU-PNR
Chem. Res. Toxicol., Vol. 9, No. 6, 1996 943
equilibrium binding studies. Alternatively, the conden-
sation of 9 with 1,2,2,2-tetrachloroethyl N-methyl-N-
nitrosocarbamate, prepared by the nitrosation of 1,2,2,2-
tetrachloroethyl N-methylcarbamate, gave the desired
nitrosourea product MNU-PNR in a similar 40% yield.
The purity of the compound is estimated to be greater
1
than 95% based on H NMR analysis (data not shown).
This route allows for the addition of the unstable ni-
trosourea functionality in the final step of the synthesis
and is amenable to the synthesis of [3H or 14C]methyl-
labeled material (41).
Sequ en ce Selective Meth yla tion of DNA. The
autoradiogram of the 576 bp fragment sequencing gel
shows that there is a dose response for the 50 and 100
µM concentrations of MNU-PNR (Figure 4A, lanes e and
f). The time course for the reaction of MNU-PNR with
DNA is shown in Figure 4A (lanes f-i) along with the
uncorrected densitometry values for the 24 h time points
(Figure 4B). The band corresponding to G271 was used
in the time course analysis because it is well resolved
and isolated from other G’s. A similar analysis using the
band for G277 affords equivalent results.
The cleavage of DNA by 100 µM MNU-PNR is dimin-
ished (15% at the 24 h time point, Figure 4B) by the
coaddition of 500 µM MU-PNR, the nonmethylating
analogue (Figure 4A, lane j). In contrast, 500 µM
ethidium bromide enhances the intensity of the bands
at the 24 h time point by 30% (Figure 4A, lane n; Figure
4B). The effect of 200 mM NaCl is to decrease DNA
methylation by 50% at 24 h (Figure 4A, lanes r vs f;
Figure 4B).
The 7-MeGua pattern for 750 µM MNU after a 24 h
incubation is also shown (Figure 4A, lanes v-y). If the
intensity of the G271 band for MNU, determined by
densitometry, is divided by 7.5 to correct for the differ-
ence in doses between MNU (750 µM) and MNU-PNR
(100 µM), the results indicate that on an equimolar basis
MNU-PNR yields ∼2.6 times more 7-MeGua than MNU.
The dose response for the formation of 7-MeGua by MNU
is linear under the experimental conditions used from
50 µM through 2 mM (26, 42) Direct comparison of the
same concentration of MNU-PNR and MNU in the
restriction fragment confirms that DNA methylation at
7-Gua by MNU-PNR is 2-7 times that seen with MNU
at the 500 and 1000 µM concentrations (Figure 5). This
F igu r e 3. Synthesis of MNU-PNR: a, NH2CH2CH2NH2, ∆; b,
Pd/C; c, PhCH2OCOCl; d, Na2Cr2O7; e, pH 5.0, ∆; f, Pd/C; g,
Cl3CCH(Cl)OC(dO)NHCH3 for 10; h, Cl3CCH(Cl)OC(dO)N-
(NdO)CH3 for 11.
(data not shown). This is also true if the MNU-PNR is
flipped around so the C2-methyl of the compound is closer
to the Cyt rather than the Gua (the latter being the
arrangement shown in Figure 2). During the energy
minimization a slight twisting of the PNR nucleus with
respect to the duplex base pairs occurred because of the
C2-methyl group. The opposite stacking scenario would
occur if the dye entered the helix via the minor groove;
however, due to the steric effect of the C2-methyl and
7-(N,N-dimethylamino) groups, it is not possible to dock
the MNU-PNR molecule into the DNA stack from the
minor groove side. As shown in the different perspectives
in Figure 2, the N-alkyl-N-nitrosourea moiety is in an
extended conformation and somewhat outside the cylin-
der defined by the long axis of the helix. Obviously, the
aliphatic linkage between the nitrosourea and the aro-
matic system is somewhat flexible and can move into the
major groove.
Syn th esis. Compounds MU-PNR and MNU-PNR
were prepared as outlined in Figure 3 using an approach
based, in part, on a previously described method for the
synthesis of PNR analogues (23). As anticipated, the
formation of the phenazenium ring system (8) by the non-
regiospecific condensation of 4 with 7 was a particularly
low yield step, and purification of the final product from
side products was difficult. Numerous efforts to improve
this step in the synthesis were unsuccessful. The de-
rivatization of 9 with 1,2,2,2-tetrachloroethyl N-methyl-
carbamate, obtained by the reaction of CH3NH3+Cl- with
1,2,2,2-tetrachloroethyl chloroformate, gave the urea
derivative MU-PNR in ∼40% yield. This compound was
used as a stable surrogate for MNU-PNR in the DNA
analysis is based on data using G271 and G277
. The
enhanced methylation for MNU-PNR is higher at the
lower dose, suggesting that PNR equilibrium binding
sites are saturated at the 1 mM concentration. This
determination of the relative yield of 7-MeGua from
MNU-PNR and MNU compares favorably to that calcu-
lated directly by HPLC analysis of the adducts after
hydrolysis of DNA treated with the two agents (see
below). MU-PNR (500 µM) inhibits MNU-mediated
strand breaks by ∼30%, while 500 µM ethidium bromide
and 200 mM NaCl inhibit by ∼50% and 70%, respectively
(Figures 4A, lanes w-z, and 4B).
The sequence selectivity for the formation of 7-MeGua
is basically the same for MNU-PNR and MNU. This is
most obvious at the G281-283 run (Figure 4A). Coincuba-
tion of MU-PNR or NaCl with MNU-PNR does not alter
the cleavage pattern.
Meth yl Ad d u ct An a lysis. The yields of 7-MeGua and
3-MeAde, based on HPLC/UV analysis, from the reaction
of MNU and MNU-PNR with salmon sperm DNA, in the
absence and presence of 200 mM NaCl, are shown in
Table 1. The adducts were released from the DNA by

944 Chem. Res. Toxicol., Vol. 9, No. 6, 1996
Mehta et al.
F igu r e 5. (A) Reactions of MNU-PNR with 576 bp restriction
fragment. Autoradiogram of polyacrylamide sequencing gel:
lane a, Gua; lane b, Gua+Ade; lane c, control; lanes d-i, 10,
50, 100, 200, 500, and 1000 µM MNU; lanes j-o, 10, 50, 100,
200, 500, and 1000 µM MNU-PNR. (B) Uncorrected densitom-
etry data for 7-MeGua formation from MNU (lanes h and i) and
MNU-PNR (lanes n and o) at Gua271 and Gua277 from panel A.
neutral thermal hydrolysis, and separated and identified
by HPLC analysis using UV detection (270 nm) that was
optimized to analyze for the minor 3-MeAde adduct (43).
The results are expressed as a ratio of adduct to unmodi-
fied Ade to correct for any difference in DNA recovery
between experiments. The highest level of DNA modi-
fication is <1 adduct/20 Ade, so there is no saturation of
potential alkylation sites.
The ratios of the yields of 7-MeGua to 3-MeAde from
MNU are 7.8 and 6.7 at low and high salt, respectively
(Table 1). These values compare well with previous
studies that report 7-MeGua/3-MeAde ratios averaging
∼7.5 (32). The ratio of 7-MeGua/3-MeAde from MNU-
PNR increases ∼5-fold to 37.2 at low salt. The increase
at high salt is ∼2-fold. The results also show that MNU-
PNR generates 2.8 times more 7-MeGua than MNU
based on an equimolar dose. This outcome is very close
to the difference calculated from the sequencing gel data
F igu r e 4. (A) Reactions of MNU-PNR with 576 bp restriction
fragment. Autoradiogram of polyacrylamide sequencing gel:
lane a, Gua; lane b, Gua+Ade; lane c, control; lane d, 10 µM
MNU-PNR; lane e, 50 µM MNU-PNR; lanes f-i, 100 µM MNU-
PNR with 24, 12, 6, and 3 h incubation times, respectively; lanes
j-m, 100 µM MNU-PNR + 500 µM MU-PNR with 24, 12, 6,
and 3 h incubation times, respectively; lanes n-q, 100 µM MNU-
PNR + 500 µM ethidium bromide with 24, 12, 6, and 3 h
incubation times, respectively; lanes r-u, 100 µM MNU-PNR
+ 200 mM NaCl with 24, 12, 6, and 3 h incubation times,
respectively; lanes v-y, 750 µM MNU alone or with 500 µM
MU-PNR, 500 µM ethidium bromide, or 200 mM NaCl, respec-
tively, with 24 h incubation. (B) Quantitation of 7-Gua methy-
lation of DNA at 24 h time point by 100 µM MNU-PNR in the
absence or presence of 500 µM MU-PNR, 500 µM ethidium
bromide, or 200 mM NaCl: uncorrected densitometric analysis
of bands corresponding to Gua271 from panel A (lanes f, j, n,
and r). Also shown are the concentration corrected (see text)
intensities of the bands for 7-MeGua formation from MNU in
the absence or presence of 500 µM MU-PNR, 500 µM ethidium
bromide, or 200 mM NaCl from panel A (lanes v-y).

MNU-PNR
Chem. Res. Toxicol., Vol. 9, No. 6, 1996 945
Ta ble 1. Yield s of 7-MeGu a a n d 3-MeAd e Ad d u cts fr om
th e Rea ction of Sa lm on Sp er m DNA w ith MNU a n d
MNU-P NR
NaCl
7-MeGua/
compd
MNUb
MNU-PNRb HPLC
method (mM) 7-MeGuaa 3-MeAdea 3-MeAde
HPLCc
0
200
0
0
200
200
1.63
0.20
4.51
4.01
1.20
1.19
0.21
0.03
0.12
0.11
0.08
0.08
8
7
38
37
15
15
MNUd
ELISAe
0
200
0
ndf
nd
nd
nd
2.28
0.31
0.86
0.63
MNU-PNRd ELISA
200
a
b
Yields are expressed as mmol of adduct/mol of Ade. [DNA]
) 1 mM; [methylating agent] ) 1 mM. c HPLC analysis with UV
d
detection. [DNA] ) 0.1 mM; [methylating agent] ) 1 mM.
e Antibody/ELISA analysis. f Not determined.
(see above). In the presence of 200 mM NaCl, the
effectiveness of MNU-PNR over MNU to methylate 7-Gua
increases to 6-fold using the HPLC/UV data from the
DNA digestion or the sequencing gels.
F igu r e 6. CD spectra of 7.6 µM MU-PNR with 200 µM
(phosphate) salmon sperm DNA in 13 mM pH 8.0 Tris-HCl
buffer: (s) DNA; (- -) DNA + 200 mM NaCl; (- - -) DNA +
MU-PNR; (- ‚ -) DNA + MU-PNR + 200 mM NaCl.
Because the levels of 3-MeAde were at the limit of UV
detection, studies were done using a more sensitive
monoclonal antibody to 3-MeAde in a competitive ELISA
determination. These studies were done using 1 mM
MNU or MNU-PNR and 0.1 mM DNA (1/10 the concen-
tration of DNA used in the UV analysis) in order to obtain
a clear signal with the ELISA. The results (Table 1)
verify that MNU generates approximately 2.5 times more
3-MeAde than MNU-PNR. The data also show that DNA
methylation at 3-Ade by MNU is much more sensitive to
the inhibitory effect of salt. There is a 70% decrease in
yield of the 7-MeGua adduct from MNU-PNR in the
presence of salt, which is much larger decrease than that
observed for 3-MeAde (25% decrease). In the ELISA
studies the ratio of methylating agent to DNA was 10:1
vs 1:1 for the HPLC/UV analysis. Also, in the HPLC/
UV analysis 3-MeAde levels are expressed as adduct/
unmodified Ade, where Ade was experimentally verified
by HPLC analysis. In the ELISA work, 3-MeAde is
expressed as adduct/Ade, where Ade levels were calcu-
lated from the initial DNA concentration used in the
incubation. It is for this reason that the apparent levels
of adduct are 10-fold higher in the antibody data (Table
1).
UV-Vis a n d CD Sp ectr a . In order to obtain physical
evidence for the association of MNU-PNR with DNA, the
stable urea derivative MU-PNR was incubated with DNA
and the visible spectrum followed as a function of
increasing the [DNA]/[compound] (data not shown). The
effect of the addition of DNA to MU-PNR, with or without
200 mM NaCl, is a red shift of 5 nm from 555 to 560 nm
with a concomitant 22% decrease in the absorbance at
the highest compound to DNA ratios. The addition of
200 mM NaCl to MU-PNR did not alter the UV-vis
spectrum.
Ta ble 2. Effect of Eth id iu m Br om id e (EB) a n d MU-P NR
on th e Viscosity of Ca lf Th ym u s DNA
compd
NaCl (mM)
slopea
rc
EB
0
200
0.1814
0.1704
0.9970
0.9959
MU-PNR
0
200
0.1207
0.1295
0.9969
0.9975
EB + MU-PNR
0
0.1066
0.9856
a
Calculated from formula shown in Experimental Section.
b Correlation coefficient. c Mixture of equimolar amounts of EB and
MU-PNR added to DNA.
of compound is not changed by the addition of salt. MU-
PNR also induces a small red shift of approximately 3
nm in the positive and negative bands at both low and
high ionic strength.
DNA Viscosity Stu d ies. The results for the viscom-
etry studies using calf thymus DNA are shown in Table
2. The slopes of the lines (L/L0 vs [compound]/[DNA-
phosphate]) indicate that ethidium bromide (slope )
0.1814, r ) 0.9970) (r, correlation coefficient) is more
effective at lengthening the DNA than MU-PNR (slope
) 0.1207, r ) 0.9969). The coaddition of 200 mM NaCl
with either compound has no significant effect on the
slopes: ethidium bromide (0.1704, r ) 0.9959) and MU-
PNR (0.1295, r ) 0.9975). The addition of the same
concentrations of a 1:1 mixture of MU-PNR and ethidium
bromide to the DNA results in a decrease in the slope
(0.1066, r ) 0.9856) as compared to either compound
alone.
Discu ssion
P NR/DNA Com p lex. The proposed orientation of
intercalated PNR (Figure 2) is rare; most small interca-
lating molecules enter the DNA stack from the minor
groove. This placement of PNR was originally inferred
from DNA binding studies (in phosphate buffer contain-
ing 180 mM NaCl) using a number of PNR analogues
and is based on both steric and electronic arguments (21,
23). It was also shown, using DNA’s with different G:C/
A:T content, that PNR has a modest preference for G:C
rich DNA (21). The modeling of PNR with a diter-
The CD spectrum of 200 µM salmon sperm DNA in 13
mM Tris buffer shows a clear B-form conformation, and
the coaddition of 7.6 µM MU-PNR increases the intensi-
ties of the positive (∼40%) and negative (∼25%) Cotton
bands (Figure 6). When the spectrum was run in the
presence of 200 mM NaCl, the intensity of the long
wavelength band increased by 20% in the presence of
MU-PNR, but there was virtually no change in the 247
nm band. The CD spectrum of the DNA in the absence

946 Chem. Res. Toxicol., Vol. 9, No. 6, 1996
Mehta et al.
calinium-DNA derived intercalation site (28) confirms
that the PNR nucleus is too wide, because of the 2-methyl
and N7-(dimethylamino) groups, to penetrate into duplex
DNA from the minor groove and attain any stacking with
the flanking bases (Figure 2).
seen with MNU-PNR can also be explained by inhibition
of nonspecific electrostatic binding of the cationic meth-
anediazonium ion and/or the cationic PNR dye to DNA
(44).
At the present time, we cannot differentiate between
the two mechanisms for salt inhibition. However, the
observation that DNA methylation at 7-Gua and 3-Ade
by MNU-PNR is less sensitive to changes in ionic
strength than MNU is consistent with the delivery of
methanediazonium ion to DNA via an intercalated
precursor. The bias in the methylation pattern to major
groove adduction, i.e., the 7-MeGua/3-MeAde ratio of
>30:1 vs to 8:1 for MNU, is also fitting with the MNU-
PNR/DNA complex shown in Figure 2 since a priori there
is no reason to assume that electrostatic interactions of
PNR with DNA would favor major groove complexation
and/or adduction.
The viscosity, CD, and UV data are consistent with the
intercalation of MNU-PNR into DNA, although the exact
structure of the MU-PNR/DNA complex can only be
inferred from the molecular modeling studies and other
modes of groove binding and cannot be ruled out. The
viscosity and spectroscopic studies show that the effects
of MU-PNR are independent of ionic strength, which is
also congruous with a hydrophobic interaction of the
compound with DNA. Previously, an acrylate linker has
been substituted onto the C4′-position of PNR in order
to exploit the preference of PNR for G:C sequences in an
affinity column strategy to selectively bind G:C rich
DNA’s (23). The acrylamide analogue of PNR has a
similar binding affinity and G:C base pair preference as
PNR.
DNA Meth yla tion . The hydrolytic decomposition of
MNU-PNR and MNU yields the same methanediazonium
ion intermediate, and this is consistent with the similar-
ity between the 7-Gua methylation pattern from MNU-
PNR and MNU. As mentioned above, the absolute yield
of 7-MeGua from MNU-PNR is approximately 2.5- and
6-fold higher than for MNU, at low and high salt,
respectively (Table 1). Moreover, the 7-MeGua to 3-MeAde
ratio, which reflects adduction in the major vs minor
groove, is approximately 5- and 2-fold higher at low and
high salt, respectively. We propose that the primary
difference between MNU-PNR and MNU, in terms of
DNA methylation, is the regioselective generation of
methanediazonium ion in, or near, the major groove,
because of the formation of an intercalation complex. It
is also possible that a nonspecific electrostatic association,
and the ∼2-fold preference of PNR for G:C over A:T
regions, can enhance the formation of Gua adducts (21,
23, 45).
The effect of ionic strength provides an important
insight into the route of delivery of reactive intermediate
to DNA. The methylation of DNA at 7-Gua by both MNU
and MNU-PNR is decreased by the addition of salt even
though salt does not affect PNR intercalation as mea-
sured by viscosity, UV, or CD. This result can be
explained by two nonexclusive mechanisms. First, a
decrease in methylation by methanediazonium ion is
expected at high salt based on previous studies with
MNU and other charged alkylating agents (14, 26, 42,
46-53). This phenomenon has been attributed to the
competitive inhibition of the electrostatic attraction
between DNA and positively charged alkylating agents
by inorganic and equilibrium binding cations. However,
neither the nature of the cation nor the ionic strength
qualitatively affects the DNA alkylation pattern of MNU
(26, 42), and the same result is seen with MNU-PNR. A
second possibility is that at low ionic strength there is a
nonspecific electrostatic association between charged
PNR dye and DNA, and this association can increase the
concentration of methanediazonium ion generated in the
vicinity of DNA (44). Along this line, a chlorambucil
nitrogen mustard has been conjugated to a polyamine
(54). This polycationic conjugate cross-links DNA through
major groove sites 104 times more efficiently than chloram-
bucil, and it is assumed that the electrostatic attraction
between the polyamine and the DNA contributes to this
increased reactivity. Therefore, the inhibitory salt effect
The increase in the MNU-PNR-mediated formation of
7-MeGua upon coaddition of ethidium bromide and the
decrease with MU-PNR are additional evidence that the
MNU-PNR complex with DNA is not simply electrostatic
since, if that were the case, both cationic intercalators
would have the same effect on DNA methylation. It
should be noted that ethidium bromide inhibits the yield
of 7-MeGua with MNU by ∼50% (Figure 4A, lanes r and
s). We suggest that MU-PNR, because of its structural
similarity to MNU-PNR, is a true competitive inhibitor
of MNU-PNR intercalation, while ethidium, because it
enters the helix from the minor groove (55, 56), is not.
In fact, it is possible, albeit speculative, that the enhanced
methylation with MNU-PNR in the presence of ethidium
is because ethidium, by entering via the minor groove,
can provide a major groove intercalation site for the
former without any additional significant conformational
change in the DNA. The observation that an equimolar
combination of MU-PNR and ethidium bromide (while
maintaining the same [total compound]/[DNA] ratios)
affords less DNA lengthening than the individual mol-
ecules (Table 2) could be explained by the suggested
synergistic mode of binding with the ethidium association
dominating the viscosity results. In the co-occupation
model, it is anticipated that the slope would be close to
a value of 0.1 or approximately 50% of that measured
when ethidium bromide is incubated with DNA alone at
the same compound/DNA ratio (slope ≈ 0.18). The initial
formation of the intercalation site by the ethidium from
the minor groove followed by entrance of the PNR from
the major groove is reasonable since ethidium has a log
KD of -5.7 vs -3.6 for PNR and the kinetics of ethidium
binding, as measured by NMR, is slow relative to PNR
(45). Other possible explanations of the viscosity data
that require a decrease in available compound through
some interaction between the ethidium and PNR mol-
ecules are not consistent with the observed increase in
DNA methylation when the two compounds are added
together. We are not aware of any evidence for the
cooperativity model for major and minor groove interca-
lators; however, there is an analogy for this process if
one assumes that a stacked bulged base crudely mimics
an intercalator. The preferential intercalation of meth-
idiumpropyl-EDTA-Fe(II) and 9-aminoacridine at bulge
sites has been reported (57, 58).
A well-known example of an intercalator that is used
to deliver a DNA damaging agent to DNA is the meth-
idium nucleus appended with an iron chelating EDTA
(59). This reagent, which is thought to bind in the minor
groove, generates oxygen-mediated DNA cleavage with

MNU-PNR
Chem. Res. Toxicol., Vol. 9, No. 6, 1996 947
structural and biological effects of DNA-damaging agents. Chem.
Res. Toxicol. 1, 1-18.
(4) Karran, P., Lindahl, T., Ofsteng, I., Evensen, G., and Seeberg, E.
(1980) E. coli mutants deficient in 3-methyladenine-DNA gly-
cosylase. J . Mol. Biol. 140, 101-127.
(5) Boiteaux, S., Huisman, O., and Laval, J . (1984) 3-Methyladenine
residues in DNA induce the SOS function sfiA in Escherichia coli.
EMBO J . 3, 2569-2573.
(6) Xiao, W., and Samson, L. (1993) In vivo evidence for endogenous
DNA alkylation damage as a source of spontaneous mutation in
eukaryotic cells. Proc. Natl. Acad. Sci. U.S.A. 90, 2117-2121.
(7) McCready, S. (1994) Repair of 6-4 photoproducts and cyclobutane
pyrimidine dimers in rad mutants of Saccharomyces cerevisiae.
Mutat. Res., DNA Repair 315, 261-273.
(8) Engelward, B., Dreslin, A., Christensen, J ., Huszar, D., Kurahara,
C., and Samson, L. (1996) Repair deficient 3-methyladenine DNA
glycosylase homozygous mutant mouse cells have increased
sensitivity to alkylation induced chromosome damage and cell
killing. EMBO J . 15, 945-952.
(9) Krowicki, K., Balzarini, J ., De Clercq, E., Newman, R. A., and
Lown, J . W. (1988) Novel DNA groove binding alkylators: design,
synthesis, and biological evaluation. J . Med. Chem. 31, 341-345.
(10) Konakahara, T., Wurdeman, R. L., and Gold, B. (1988) Synthesis
of an N-methyl-N-nitrosourea linked to a methidium chloride
analogue and its reactions with 32P-end-labeled DNA. Biochem-
istry 27, 8606-8613.
(11) Baker, B. F., and Dervan, P. B. (1989) Sequence-specific cleavage
of DNA by N-bromoacetyldistamycin. Product and kinetic analy-
ses. J . Am. Chem. Soc. 111, 2700-2712.
(12) Arcamone, F. M., Animati, F., Barbieri, B., Configliacchi, E.,
DAlessio, R., Geroni, C., Giuliana, C., Lazzari, E., Menozzi, M.,
Mongelli, N., Penco, S., and Verini, M. A. (1989) Synthesis, DNA-
binding properties, and antitumor activity of novel distamycin
derivatives. J . Med. Chem. 32, 774-778.
(13) Prakash, A. S., Denny, W. A., Gourdie, T. A., Valu, K. K.,
Woodgate, P. D., and Wakelin, L. P. G. (1990) DNA-directed
alkylating ligands as potential antitumor agents: sequence
specificity of alkylation by intercalating aniline mustards. Bio-
chemistry 29, 9799-9807.
(14) Church, K. M., Wurdeman, R. L., Zhang, Y., Chen, F.-X., and Gold,
B. (1990) N-(2-Chloroethyl)-N-nitrosoureas covalently bound to
nonionic and monocationic lexitropsin dipeptides. Synthesis, DNA
affinity binding characteristics, and reactions with 32P-end-labeled
DNA. Biochemistry 29, 6827-6838.
(15) Zhang, Y., Chen, F.-X., Mehta, P., and Gold, B. (1993) Groove-
and sequence-selective alkylation of DNA by sulfonate esters
tethered to lexitropsins. Biochemistry 32, 7954-7965.
(16) Iversen, B. L., and Dervan, P. B. (1988) Non-enzymatic sequence-
specific methyl transfer to single-stranded DNA. Proc. Natl. Acad.
Sci. U.S.A. 85, 4615-4619.
(17) Meyer, R. B., J r., Tabone, J . C., Hurst, G. D., Smith, T. M., and
Gamper, H. (1989) Efficient, specific cross-linking and cleavage
of DNA by stable, synthetic complementary oligonucleotides. J .
Am. Chem. Soc. 111, 8517-8519.
(18) Kido, K., Inoue, H., and Ohtsuka, E. (1992) Sequence-dependent
cleavage of DNA by alkylation with antisense oligodeoxynucle-
otides containing a 2-(N-iodoacetylaminoethyl)thio-adenine. Nu-
cleic Acids Res. 20, 1339-1344.
(19) Alley, S. C., and Hopkins, P. B. (1994) DNA-DNA interstrand
cross-linking by 2,5-bis(1-aziridinyl)-3,6-bis(carbethoxyamino)-1,4-
benzoquinone: covalent structure of the dG-to-dG cross-links in
calf thymus DNA and a synthetic DNA duplex. Chem. Res.
Toxicol. 7, 666-672.
little sequence specificity and is widely used as a foot-
printing agent (60). In other elegant studies, aniline
mustards have been linked to the 9-position of the
intercalator 9-aminoacridine so that the mustard and
acridine become covalently linked to DNA (13, 61, 62).
The results of this work show that the DNA-acridine
interactions, presumably through a minor groove interac-
tion, affect the sequence selectivity of DNA alkylation by
the nondiffusible mustard group. Unlike the typical
nitrogen mustards that alkylate Gua, these intercalator-
linked mustards react significantly with Ade sites, al-
though complete characterization of the Ade adducts has
not been reported.
Selective methylation of DNA at 7-Gua has also been
described. Iversen and Dervan prepared deoxyuracil
triphosphate appended with a methylthioether group and
introduced the substituted base into ss-DNA using Kle-
now (16). The ss-DNA was hybridized to a complemen-
tary strand, and the methylthioether functionality acti-
vated by the addition of cyanogen bromide. The resulting
cyanomethanesulfonium ion methylated the complemen-
tary strand selectively at 7-Gua. In a more straightfor-
ward approach, synthetic oligomers have been site-
specifically modified with 7-MeGua by the incorporation
of 7-Me-dGTP into a gap using DNA polymerase (63).
While these methods sequence selectively afford 7-MeGua
lesions, they obviously cannot be used to generate this
lesion in vivo.
In summary, the data presented show that the use of
PNR to deliver a small diffusible alkylating agent to DNA
can cause a significant change in the methylation pattern
with a 5-fold increase in the ratio of major to minor
groove lesions. The data also demonstrate that an
intercalation complex between MNU-PNR and DNA is
most likely responsible for the change in the methylation
pattern due to an increase in the amount of methanedia-
zonium ion formed near the major groove of DNA. One
of the reasons that the groove specificity for DNA
methylation is not higher from MNU-PNR is due to the
involvement of the “diffusible” methanediazonium ion.
The employment of bimolecular methylating agents, e.g.,
methyl sulfonate esters (15), should increase the major
to minor groove methylation ratio. Other future work
will involve the analysis of how the nature and length of
the tether between the alkylating agent and intercalator
affect equilibrium binding and methylation, and the
extension of the molecular design to include potential
antineoplastic molecules derivatized with cross-linking
agents.
(20) Bergman, H. M., and Young, P. R. (1981) The interaction of
intercalating drugs with nucleic acids. Annu. Rev. Biophys.
Bioeng. 10, 87-114.
(21) Mu¨ller, W., Bu¨nemann, H., and Dattagupta, N. (1975) Interac-
tions of heteroaromatic compounds with nucleic acids. 2. Influence
of substituents on the base and sequence specificity of intercalat-
ing ligands. Eur. J . Biochem. 54, 279-291.
(22) Kosturko, L. D., Dattagupta, N., and Crothers, D. M. (1979)
Selective repression of transcription by base sequence specific
synthetic polymers. Biochemistry 18, 5751-5756.
Ack n ow led gm en t. This work was supported by PHS
Grants R01 CA29088 and Core Grant P30 CA36727 from
the National Cancer Institute and by SIG-16 grant from
the American Cancer Society. Mass spectra were run at
the Eppley Institute by Mr. David Babcook (Eppley
Institute) and at the Midwest Center for Mass Spectrom-
etry (University of NebraskasLincoln).
(23) Bu¨nemann, H., Dattagupta, N., Schuetz, H. J ., and Mu¨ller, W.
(1981) Synthesis and properties of acrylamide-substituted base
pair specific dyes for deoxyribonucleic acid template mediated
synthesis of dye polymers. Biochemistry 20, 2864-2874.
(24) Kriek, E., and Emmelot, P. (1964) Methylation of deoxyribonucleic
acid by diazomethane. Biochim. Biophys. Acta 91, 59-66.
(25) Smith, R. H., J r., Koepke, S. R., Tondeur, Y., Denlinger, C. L.,
and Michejda, C. J . (1988) The methanediazonium ion in water:
competition between hydrolysis and proton exchange. J . Chem.
Soc., Chem Commun. 1985, 936-937.
Refer en ces
(1) Henry-Amar, M., and Dietrich, P. Y. (1993) Acute leukemia after
the treatment of Hodgkin’s disease. Hematol. Oncol. Clin. North
Am. 7, 369-387.
(2) Povirk, L. F., and Shuker, D. E. (1994) DNA damage and
mutagenesis induced by nitrogen mustards. Mutat. Res. 318,
205-226.
(3) For review: Basu, A. K., and Essigmann, J . M. (1988) Site-
specifically modified oligodeoxynucleotides as probes for the
(26) Wurdeman, R. L., Church, K. M., and Gold, B. (1989) DNA
methylation by N-methyl-N-nitrosourea, N-methyl-N′-nitro-N-

948 Chem. Res. Toxicol., Vol. 9, No. 6, 1996
Mehta et al.
nitrosoguanidine, N-nitroso(1-acetoxyethyl)methylamine, and dia-
zomethane: mechanism for the formation of N7-methylguanine
in sequence-characterized 5′-32P-end-labeled DNA. J . Am. Chem.
Soc. 111, 6408-6412.
(47) McCalla, D. R. (1968) Reaction of N-methyl-N′-nitro-N-ni-
trosoguanidine and N-methyl-N-nitroso-p-toluenesulfonamide with
DNA in vitro. Biochim. Biophys. Acta 155, 114-120.
(48) J ensen, D. E., and Reed, D. J . (1978) Reaction of DNA with
alkylating agents. Quantitation of alkylation by ethylnitrosourea
of oxygen and nitrogen sites on poly[dA-dT] including phospho-
triester formation. Biochemistry 17, 5098-5107.
(49) Briscoe, W. T., and Cotter, L.-E. (1985) DNA sequence has an
effect on the extent and kinds of alkylation of DNA by a potent
carcinogen. Chem.-Biol. Interact. 56, 321-331.
(50) Rajalakshmi, S., Rao, P. M., and Sarma, D. S. R. (1978) Studies
on carcinogen chromatin-DNA interaction: inhibition of N-
methyl-N-nitrosourea-induced methylation of chromatin-DNA by
spermidine and distamycin A. Biochemistry 17, 4515-4518.
(51) Mattes, W. B., Hartley, J . A., and Kohn, K. W. (1986) DNA
sequence selectivity of guanine-N7 alkylation by nitrogen mus-
tards. Nucleic Acids Res. 14, 2971-2987.
(52) Hartley, J . A., Mattes, W. B., Vaughan, K., and Gibson, N. W.
(1988) DNA sequence specificity of guanine N7-alkylations for a
series of structurally related triazines. Carcinogenesis 9, 669-
674.
(53) Liang, G., Encell, L., Nelson, M. G., Switzer, C., Shuker, D. E.
G., and Gold, B. (1995) The role of electrostatics in the sequence
selective reaction of charged alkylating agents with DNA. J . Am.
Chem. Soc. 117, 10135-10136.
(54) Cullis, P. M., Merson-Davies, L., and Weaver, R. (1995) Conjuga-
tion of a polyamine to the bifunctional alkylating agent chloram-
bucil does not alter the preferred cross-linking site in duplex DNA.
J . Am. Chem. Soc. 117, 8033-8034.
(27) Mayo, S. L., Olafson, B. D., and Goddard, W. A., III (1990)
Dreiding: a generic force field for molecular simulations. J . Phys.
Chem. 94, 8897-8909.
(28) Gao, Q., Williams, L. D., Egli, M., Rabinovich, D., Chen, S.-H.,
Quigley, G. J ., and Rich, A. (1991) Drug-induced DNA repair:
X-ray structure of a DNA-ditercalinium complex. Proc. Natl.
Acad. Sci. U.S.A. 88, 2422-2426.
(29) Williams, L. D., and Gao, Q. (1992) DNA-ditercalinium interac-
tions: Implications for recognition of damaged DNA. Biochemistry
31, 4315-4324.
(30) Weiner, S. J ., Kollman, P. A., Case, D. A., Singh, U. C., Ghio, C.,
Alagona, G., Profeta, S., J r., and Weiner, P. (1984) A new force
field for molecular mechanical simulation of nucleic acids and
proteins. J . Am. Chem. Soc. 106, 765-784.
(31) Barcelo, G., Senet, J .-P., and Sennyey, G. (1987) 1,2,2,2-Tetra-
chloroethyl carbamates: versatile intermediates for the synthesis
of N-nitrosoureas. Synthesis 11, 1027-1029.
(32) Beranek, D. T., Weis, C. C., and Swenson, D. H. (1980) A
comprehensive quantitative analysis of methylated and ethylated
DNA using high pressure liquid chromatography. Carcinogenesis
1, 595-605.
(33) Prevost, V., Shuker, D. E. G., Bartsch, H., Pastorelli, R., Stillwell,
W. G., Trudel, L. J ., and Tannenbaum, S. R. (1990) The deter-
mination of urinary 3-methyladenine by immunoaffinity chro-
matography-monoclonal-antibody-based ELISA: use in human
biomonitoring studies. Carcinogenesis 11, 1747-1751.
(34) Cantor, C. R., and Schimmel, P. R. (1980) Ligand interactions at
equilibrium. In Biophysical Chemistry, Part III, p 864, W. H.
Freeman and Co., New York.
(55) Tsai, C.-C., J ain, S. C., and Sobell, H. M. (1977) Visualization of
drug-nucleic acid interactions at atomic resolution. I. Structure
of an ethidium/dinucleotide monophosphate crystalline complex,
ethidium:5-iodouridylyl(3′-5′) adenosine. J . Mol. Biol. 144, 301-
315.
(35) Rhode, S. L., III (1985) Nucleotide sequence of the coat protein
gene of canine parvovirus. J . Virol. 54, 630-633.
(36) Maxam, A. M., and Gilbert, W. (1980) Sequencing end-labeled
DNA with base-specific chemical cleavages. Methods Enzymol. 65,
499-560.
(37) Lawley, P. D., and Brookes, P. (1963) Further studies on the
alkylation of nucleic acids and their constituent nucleotides.
Biochem. J . 89, 127-138.
(38) Cory, M., McKee, D., Kagan, J ., Henry, D., and Miller, J . (1985)
Design, synthesis, and DNA binding properties of bifunctional
intercalators. Comparison of polymethylene and diphenyl ether
chains connecting phenanthridine. J . Am. Chem. Soc. 107, 2528-
2536.
(39) Cohen, G., and Eisenberg, H. (1966) Conformation studies on the
sodium and cesium salts of calf thymus deoxyribonucleic acid
(DNA). Biopolymers 4, 429-440.
(56) J ain, S. C., Tsai, C.-C., and Sobell, H. M. (1977) Visualization of
drug-nucleic acid interactions at atomic resolution. II. Structure
of an ethidium/dinucleotide monophosphate crystalline complex,
ethidium:5-iodocytidylyl(3′-5′) guanosine. J . Mol. Biol. 114, 317-
331.
(57) Williams, L. D., and Goldberg, I. H. (1988) Selective strand
scission by intercalating drugs at DNA bulges. Biochemistry 27,
3004-3011.
(58) Woodson, S. A., and Crothers, D. M. (1988) Binding of 9-ami-
noacridine to bulged-base DNA oligomers from a frame-shift hot
spot. Biochemistry 27, 8904-8914.
(59) Hertzberg, R. P., and Dervan, P. B. (1984) Cleavage of DNA with
methidiumpropyl-EDTA-iron(II): Reaction conditions and product
analyses. Biochemistry 23, 3934-3945.
(60) Van Dyke, M. W., Hertzberg, R. P., and Dervan, P. B. (1982) Map
of distamycin, netropsin, and actinomycin binding sites on
heterogeneous DNA: DNA cleavage-inhibition patterns with
methidiumpropyl-EDTA‚Fe(II). Proc. Natl. Acad. Sci. U.S.A. 79,
5470-5474.
(61) Valu, K. K., Gourdie, T. A., Boritzki, T. J ., Gravatt, L., Baguley,
B. C., Wilson, W. R., Wakelin, L. P. G., Woodgate, P. D., and
Denny, W. A. (1990) DNA-directed alkylating agents. 3. Structure-
activity relationships for acridine-linked aniline mustards: con-
sequence of varying the length of the linker chain. J . Med. Chem.
33, 3014-3019.
(40) Cohen, G., and Eisenberg, H. (1969) Viscosity and sedimentation
study of sonicated DNA-proflavine complexes. Biopolymers 8,
45-55.
(41) Mehta, P., Konakahara, T., and Gold, B. (1992) A convenient
method to synthesize N-[3
H]methyl-N-methyl-N-nitrosocarbamate
transfer reagents. J . Labelled Compd. Radiopharm. 31, 925-931.
(42) Wurdeman, R. L., and Gold, B. (1988) The effect of DNA sequence,
ionic strength, and cationic DNA affinity binders on the methy-
lation of DNA by N-methyl-N-nitrosourea. Chem. Res. Toxicol.
1, 146-147.
(43) Singer, B., and Grunberger, D. (1983) Molecular Biology of
Mutagens and Carcinogens, p 299, Plenum Press, New York.
(44) Zimmermann, H. W. (1986) Physicochemical and cytochemical
investigations on the binding of ethidium and acridine dyes to
DNA and to organelles in living cells. Angew. Chem., Inter. Ed.
Engl. 25, 115-130.
(62) Prakash, A. S., Denny, W. A., and Wakelin, L. P. G. (1990)
Differences in sequence selectivity of DNA alkylation by isomeric
intercalating aniline mustards. Chem.-Biol. Interact. 76, 241-
248.
(45) Feigon, J ., Denny, W. A., Leupin, W., and Kearns, D. R. (1984)
Interactions of antitumor drugs with natural DNA: 1H NMR
study of binding mode and kinetics. J . Med. Chem. 27, 450-465.
(46) Kriek, E., and Emmelot, P. (1963) Methylation and breakdown
of microsomal and soluble ribonucleic acid from rat liver by
diazomethane. Biochemistry 2, 733-740.
(63) Ezaz-Nikpay, K., and Verdine, G. L. (1992) Aberrantly methylated
DNA: Site-specific introduction of N7-methyl-2′-deoxyguanosine
into the Dickerson/Drew dodecamer. J . Am. Chem. Soc. 114,
6562-6563.
TX960007N