Induction of B- to Z-DNA transition by copper and zinc complexes with C(15) substituted macrocyclic pentaaza ligands

Article abstract of DOI:10.1039/b500317b

The new macrocyclic ligand 15-fluoro-15-methyl-l,4,7,10,13- pentaazacyclohexadecan-14,16-dione (2) was synthesised and its crystal structure determined together with the ones of the known analogues of 2, 15-fluoro-1,4,7,10,13-pentaazacyclohexadecan-14,16-dione (1) and 15,15-difluoro-1,4,7,10,13-pentaazacyclohexadecan-14,16-dione (3). The binding behaviour of all three ligands to copper and zinc was studied in the solid state. They can bind to the metal centre by either triple coordination (N3) with all secondary amines or after double deprotonation of the two amides with all five nitrogen atoms (N5). The N5 coordination mode is favoured by the presence of one or two fluorine substituents at the C(15) position and by a high pH in the case of aqueous solutions. Circular dichroism titrations of poly d(GC) with the metal complexes showed that only 4 and 5, that is the copper complexes of 1 and 2, induced a complete B- to Z-DNA transition. The degree of cooperativity of the transition was found to be 3.4 and 7.3 for 4 and 5 respectively. As a possible hypothesis to explain this difference, the additional methyl group in 5 compared with 4 may be involved in a hydrophobic interaction with the DNA. Ligand 2, the copper complex 6 of the bis fluoro substituted ligand 3, and the zinc complex 7 of ligand 1 did not induce any change in the direction of Z-DNA. In the case of 6, the CD spectrum of the DNA actually showed no change at all, indicating that the complex was even not interacting with the B form of DNA. Therefore it is assumed that the bis fluoro substitution is causing the complex to be in the neutral N5 coordination mode at the experimental conditions of pH 7. The electrostatic contribution together with the shielding effect of the ligand might explain the absence of any interaction with the DNA. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b500317b

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Induction of B- to Z-DNA transition by copper and zinc complexes
with C(15) substituted macrocyclic pentaaza ligands†
Bernhard Spingler* and Chiara Da Pieve
University of Z u¨ rich, Winterthurerstr. 190, CH 8057, Z u¨ rich, Switzerland.
E-mail: spingler@aci.unizh.ch; Fax: +41 44 635 68 03; Tel: +41 44 635 46 56
Received 10th January 2005, Accepted 11th March 2005
First published as an Advance Article on the web 1st April 2005
The new macrocyclic ligand 15-fluoro-15-methyl-1,4,7,10,13-pentaazacyclohexadecan-14,16-dione (2) was
synthesised and its crystal structure determined together with the ones of the known analogues of 2, 15-fluoro-
1
,4,7,10,13-pentaazacyclohexadecan-14,16-dione (1) and 15,15-difluoro-1,4,7,10,13-pentaazacyclohexadecan-14,16-
dione (3). The binding behaviour of all three ligands to copper and zinc was studied in the solid state. They can bind
to the metal centre by either triple coordination (N3) with all secondary amines or after double deprotonation of the
two amides with all five nitrogen atoms (N5). The N5 coordination mode is favoured by the presence of one or two
fluorine substituents at the C(15) position and by a high pH in the case of aqueous solutions. Circular dichroism
titrations of poly d(GC) with the metal complexes showed that only 4 and 5, that is the copper complexes of 1 and 2,
induced a complete B- to Z-DNA transition. The degree of cooperativity of the transition was found to be 3.4 and 7.3
for 4 and 5 respectively. As a possible hypothesis to explain this difference, the additional methyl group in 5 compared
with 4 may be involved in a hydrophobic interaction with the DNA. Ligand 2, the copper complex 6 of the bis fluoro
substituted ligand 3, and the zinc complex 7 of ligand 1 did not induce any change in the direction of Z-DNA. In the
case of 6, the CD spectrum of the DNA actually showed no change at all, indicating that the complex was even not
interacting with the B form of DNA. Therefore it is assumed that the bis fluoro substitution is causing the complex to
be in the neutral N5 coordination mode at the experimental conditions of pH 7. The electrostatic contribution
together with the shielding effect of the ligand might explain the absence of any interaction with the DNA.
Introduction
DNA is a flexible and dynamic molecule that can assume
a variety of interconverting forms. Each form has its own
biological role in the regulation of the life of the cell. Selective
stabilisation of one of these forms can help to discover and
better understand their biological roles. Genetic information is
provided by the DNA in at least two different ways. First, the
sequence of its nucleotides determines the primary structure of
the proteins. Second, DNA can regulate gene expression through
1
its shape.
Since the left-handed Z-DNA is distinguished from the other
DNA conformations by its own very characteristic geometry,
Fig.
1 Left: complex NiHFGN synthesised by Rokita and
14
2
co-workers. Right: structure of the complexes studied (see Table 1).
it can be recognised due to its geometrical features and to a less
extent by the sequences of the nucleobases. Aliphatic, including
naturally occurring, polyamines are able to induce the formation
Table 1 Numbering scheme of the complexes studied
3,4
of Z-DNA. There have been several reports about synthetic,
rigid organic compounds which favor the formation of Z-
M
R
1
= F, R
2
= H
R
1
= F, R
2
= Me
R
1
= F, R = F
2
5,6
—
Cu
Zn
1
4
7
2
5
—
3
6
8
DNA. Some transition metals such as copper or nickel are also
7–9
known to induce Z-DNA, but this process is strongly inhibited
10,11
in the presence of only 10 mM sodium chloride.
Copper ions
were found to bind to N(7) of guanine when crystals of a hexamer
12
in the Z-DNA form were soaked with CuCl
2
.
Furthermore
We investigated how different metals on the one hand
and various substituents at the C(15) position of 1,4,7,10,13-
pentaazacyclohexadecan-14,16-dione on the other hand influ-
ence the ability of the resulting metal complexes to promote the
transition from B- to Z-DNA relative to the original complex
NiHFGN. The substitutents were chosen to have very different
electronic but similar sterical properties.
the complex of copper with diethylenediamine but not copper
13
chloride on its own was shown to induce Z-DNA. Rokita and
co-workers synthesised the small complex NiHFGN (Fig. 1) and
showed that it promotes the conversion of poly d(GC) from the
14
B to the Z form in a molar ratio complex : nucleotides = 1 : 3.
The basic core of HFGN (1,4,7,10,13-pentaazacyclohexadecan-
4,16-dione) had been introduced by Kimura and co-workers
1
when studying the corresponding copper and nickel complexes
intended to inhibit superoxide dismutase and work as an oxygen
activator respectively.
Experimental
15,16
Chemicals and methods
Chemicals were purchased from Aldrich or Fluka and used
without further purification. Diethyl fluoromalonate was pur-
chased from Apollo Scientific Limited, and diethyl difluoroma-
lonate from ABCR. Tetraethylene pentamine was isolated from
†
Electronic supplementary information (ESI) available: ORTEP plots
and crystallographic information tables for 1·2HCl, 2·2HCl, and 3.
CD spectrum of 2 added to poly d(GC). See http://www.rsc.org/
suppdata/dt/b5/b500317b/
17
tetraethylene pentamine pentahydrochloride in 96% yield.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
D a l t o n T r a n s . , 2 0 0 5 , 1 6 3 7 – 1 6 4 3
1 6 3 7

View Article Online
Table 2 Crystallographic data of 4, 10, 7, and 5
4
10
7
5
Empirical formula
M/g mol−
C
474.98
15
H
30CuFN
5
O
7
C
354.88
Monoclinic
P2
8.7165(5)
11.8274(8)
14.3976(8)
90
101.960(6)
90
1452.08(15)
4
1.534
11
H
22CuFN
5
O
3
C
476.81
15
H
30FN
5
O
7
Zn
C
489.01
Monoclinic
P2 /n
10.2427(5)
13.4133(10)
16.2893(8)
90
100.645(6)
90
2199.4(2)
4
1.048
16 5 27
H32CuFN O
1
Crystal system
Monoclinic
I2/a
15.0078(6)
9.2569(5)
29.9742(12)
90
100.063(5)
90
4100.1(3)
8
Monoclinic
I2/a
15.2026(8)
9.1232(4)
30.4076(14)
90
100.810(6)
90
4142.6(3)
8
Space group
1
/n
1
a/A˚
b/A˚
c/A˚
◦
a/
b/
◦
◦
c /
V/A˚
3
Z
−
1
l/mm
1.121
1.241
Reflections collected
Independent reflections [Rint
Final R indices [I > 2r(I)]
23587
4937 [0.0796]
43906
]
6266 [0.0547]
= 0.0455 wR
R
1
= 0.0511 wR
2
= 0.1286
2
= 0.1358
R
1
2
= 0.1258
2
1
5-Fluoro-1,4,7,10,13-pentaazacyclohexadecan-14,16-dione (1)
See http://www.rsc.org/suppdata/dt/b5/b500317b/ for cry-
stallographic data in CIF format.
18
was synthesised according to the literature. All reactions were
performed under a nitrogen atmosphere. The reactions were
monitored by HPLC or thin layer chromatography (TLC). TLC
was carried out on 0.25 mm Merck silica gel aluminium plates
Syntheses of mononucleating ligands and complexes
25
2
-Fluoro-2-methyl-diethyl malonic ester (9). Diethyl fluo-
(
60 F254) or aluminium oxide pre-coated plastic sheets (alox
19
romalonate (4.5 ml, 22.5 mmol) was added dropwise to a
suspension of NaH (2.7 g, 112.5 mmol) in 60 ml of dry
THF. The mixture was stirred at r.t. until the gas evolution
ceased. Iodomethane (1.43 ml, 22.9 mmol) was added dropwise
to the mixture, which was stirred at r.t. for 12 h. The suspension
was then filtrated and the solvent was removed under vacuum.
N/UV254) using UV light or the Schlittler staining solution
as the visualising agent.
Column chromatography was performed on silica gel (particle
size 0.040–0.063 mm) or aluminium oxide (basic, 0.05–0.15 mm,
pH 9.5 ± 0.5). Ion exchange chromatography was performed on
DOWEX 2 X8 20–50 or AMBERLYST A26-OH.
Electrospray ionisation (ESI) mass spectra were recorded on
a Merck Hitachi M-8000 spectrometer. Electron impact (EI)
and fast atom bombardment (FAB) mass spectra were recorded
on a Finnigan MAT mass spectrometer model 8230 equipped
with an ION Tech ionisator (Teddington England). In the case
of FAB-MS the samples were dissolved in ethylene glycol and
para-nitrobenzyl alcohol (p-NBA) was used as a matrix.
Gas chromatography-mass spectra (GC-MS) were recorded
The residue was dissolved in water and extracted with CH
2
Cl
2
.
The organic phase was dried over Na SO and evaporated.
2
4
The residue was purified by chromatography on silica gel. The
elution was started with hexane 100% and then changed to 5 : 1
hexane : CH
66.5%). R
= 0.1 (100% hexane). H NMR (300 MHz, CDCl
), 1.75 (d, J = 23.7 Hz, 3 H,
). C NMR (75 MHz,
), 62.50
2
Cl
2
to obtain a colourless viscous oil. Yield 2.87 g
1
(
f
3
):
d = 1.28 (t, J = 7.2 Hz, 6 H, CH
3
1
3
CFCH
CDCl
CH
3
), 4.25 (q, J = 7.2 Hz, 4 H, CH
2
3
): d = 13.76 (CH
3
), 20.48 (d, J = 23 Hz, CFCH
3
on
equipped with a Saturn 2000 electron impact mass spectrometer
EI-MS). The samples were carried by He through a CP-
a Varian-Chrompack CP-3800 gas chromatographer
(
2
), 92.30 (d, J = 194 Hz, CF), 166.90 (d, J = 25 Hz, C=O).
+
MS (GC-MS): R.T. = 9.14 min; m/z (%) = 193 (100) [M] .
(
SIL 8 CB-MS column (30 m × 0.25 mm). UV/vis spectra
were measured on a Varian Cary 50 spectrometer. IR spectra
were recorded on a Perkin Elmer BX FT-IR spectrometer
using samples in KBr pellets. NMR spectra were determined
with the help of a Varian Gemini 200 or 300 MHz and
a Bruker DRX 500 MHz spectrometer. The chemical shifts
are relative to residual solvent protons as reference. Circular
dichroism measurements were performed using a Jasco J-810
spectropolarimeter equipped with a Jasco PFD-4255 Peltier
temperature controller. Elemental analyses were carried out on
a Leco CHNS-932 elemental analyser.
1
5-Fluoro-15-methyl-1,4,7,10,13-pentaazacyclohexadecan-14,
1
6-dione (2). A solution of 2-fluoro-2-methyl-diethyl malonic
ester (9) (908 mg, 4.73 mmol) in 2 ml of dry ethanol was
added to a solution of 0.89 g of tetraethylenepentamine
4.7 mmol) dissolved in 62 ml of dry ethanol. The solution
was refluxed for 2 days. The solvent was then removed
under vacuum and the crude residue was recrystallised from
0 : 1 CH CN : MeOH as a white solid. Since the product
was partially a hydrochloride salt, it was purified by ion-
exchange chromatography (AMBERLYST A-26 OH). The
AMBERLYST was washed with bidistilled water and with
MeOH, and was then activated with 1 M NaOH. The product
(
2
3
Crystallographic data were collected on a Stoe IPDS diffrac-
tometer at 183(2) K using a graphite-monochromated Mo-Ka
was recovered by using water as eluent. Yield 177 mg (42%).
˚
radiation (k = 0.71073 A). Suitable crystals were covered with
1
R
f
= 0.17 (80% CH
2
Cl
2
, 18% MeOH, 2% conc. ammonia). H
Paratone N oil, mounted on top of a glass fibre and immediately
transferred to the diffractometer. Eight thousand reflections
distributed over the whole limiting sphere were selected by the
program SELECT and used for unit cell parameter refinement
NMR (300 MHz, D
2
O): d = 1.62 (d, J = 23.4 Hz, 3 H, CH
.53–2.64 (m, 12 H, CH ), 3.05–3.65 (m, 4 H, CH NHCO).
), 36.47
3
13
),
C
2
2
2
NMR (75 MHz, D
2
O): d = 18.99 (d, J = 23.1 Hz, CH
3
(
1
(
1
CH
2
), 44.52 (CH
2
), 44.58 (CH
2
), 44.67 (CH
2
), 93.40 (d, J =
20
with the program CELL. Data were collected for Lorentz
and polarisation effects as well as for absorption (numerical).
96 Hz, CF), 167.79 (d, J = 22.7 Hz, C=O). MS (ESI): m/z
+
+
%) = 290 (100) [M + H] , 272 (42) [M − F + 2H] . IR (KBr):
2
1
Structures were solved with direct methods using SHELXS-97
−1
691 cm m C=O.
22
or SIR97 and were refined by full-matrix least-squares methods
on F with SHELXL-97. Crystallographic data is presented in
2
23
15,15-Difluoro-1,4,7,10,13-pentaazacyclohexadecan-14,16-
26
Table 2. ORTEP plots were drawn with the program ORTEP-
dione (3). Diethyl difluoromalonate (1 ml, 5.924 mmol)
was added to a solution of 1.12 g of tetraethylenepentamine
(5.920 mmol) in 110 ml of dry ethanol. The solution was refluxed
for 24 h. The solvent was then removed under vacuum and the
24
3
for Windows at a probability of 50%; non-acidic hydrogen
atoms were omitted for clarity.
CCDC reference numbers 260290–260295 and 265759.
1
6 3 8
D a l t o n T r a n s . , 2 0 0 5 , 1 6 3 7 – 1 6 4 3

View Article Online
crude residue was recrystallised from 10 : 1 CH
to give a white solid. Yield 692.5 mg (40%). R
CH Cl
500 MHz, MeOH-d
m, 4 H, H(3,11)), 3.45–3.47 (m, 4 H, H(2,12)). C NMR
125 MHz, MeOH-d
3
CN : MeOH
a solution of 39.9 mg of zinc acetate·2H
2
O (0.182 mmol) in
15 ml of dry ethanol. The solution was refluxed for 12 h and
f
= 0.76 (80%
1
2
2
, 18% MeOH, 2% conc. ammonia; on alox). H NMR
the solvent removed under vacuum to obtain a white product.
1
(
(
(
4
): d = 2.69 (s, 8 H, H(5,6,8,9)), 2.77–2.79
Yield 83 mg (quantitative). H NMR (500 MHz, MeOH-d
4
):
). EA calcd
O: C, 36.49; H, 6.53; N, 14.19; found:
1
3
d = 1.96 (s, 6 H, CH
3
), 2.40–3.57 (m, 16 H, CH
2
4
): d = 39.26 (C(2,12)), 48.69 (C(5,9)),
for C15
H
28FN
5
O
6
Zn·2H
2
−
1
4
1
9.50 (C(6,8)), 49.76 (C(3,11)), 110.33 (t, J = 261 Hz, C(15)),
64.47 (t, J = 24.4 Hz, C(14,16)). MS (ESI): m/z (%) = 293
: C,
5.02; H, 7.21; N, 23.88; found: C, 45.81; H, 7.10; N, 23.84. IR
C, 36.37; H, 6.52; N, 14.16. IR (KBr): 1700 cm m C=Oligand
,
−
1
1576 cm m C=Oacetate
.
+
+
(
4
100) [M] , 274 (48) [M − F] . EA calcd for C11
H
21
F
2
O
2
N
5
Zn(II) complex of 15,15-difluoro-1,4,7,10,13-pentaazacyclo-
hexadecan-14,16-dione (8). 15,15-Difluoro-1,4,7,10,13-penta-
azacyclohexadecan-14,16-dione (3) (50 mg, 0.171 mmol) was
−
1
(
KBr): 1695 cm m C=O.
Cu(II) complex of 15-fluoro-1,4,7,10,13-pentaazacyclohexa-
added to a solution of 37.4 mg of zinc acetate·2H
2
O
decan-14,16-dione (4). 15-Fluoro-1,4,7,10,13-pentaazacyclo-
hexadecan-14,16-dione (1) (50 mg, 0.1818 mmol) was added
to a solution of 36.3 mg of copper acetate monohydrate
0.1818 mmol) in 15 ml of dry ethanol. The solution was stirred at
r.t. for 3 h. The solvent was removed under vacuum to yield a blue
product. Violet crystals were obtained from vapour diffusion of
(0.171 mmol) in 10 ml of dry ethanol. The solution was refluxed
for 3 h and the solvent removed under vacuum to obtain a white
product. Yield 81 mg (quantitative). MS (FAB): m/z (%) =
+
1
(
356 (100) [M] . H NMR (200 MHz, MeOH-d
4
): d = 1.95 (s,
−1
6 H, CH
3
), 2.49–3.45 (m, 16 H, CH ). IR (KBr): 1715 cm
m C=Oligand, 1597 cm m C=Oacetate
2
−
1
.
hexane into an ethanolic solution of the complex. Yield 82 mg
+
(
quantitative). MS (FAB): m/z (%) = 336 (100) [M] . EA calcd
CD titration of poly d(GC) with 1, 4, 5, 6, 7 and Cu(OAc)
2
for C15
H
28CuFN
5
O
6
·2H
2
O: C, 36.57; H, 6.55; N, 14.22; found:
1
A 0.1 mM poly d(GC) solution in 1 mM sodium cacodylate
buffer (pH = 7) was titrated with various aliquots of 1 mM
solutions of the metal complexes 4, 5, 6, 7, the ligand 1, or
−
C, 36.63; H, 6.32; N, 14.40. IR (KBr): 1696 cm m C=Oligand
,
−
1
−1
1
570 cm m C=Oacetate (blue complex). IR (KBr): 1618 cm
m C=Oligand (violet complex). UV-Vis (10 M, EtOH): kmax
−
3
=
of Cu(OAc)
2
in the same buffer. The samples were warmed
−
1
−1
−3
5
80 nm (e = 107 cm M ). UV-Vis (10 M, bi-distilled water,
◦
◦
to 60 C for 5 min and then cooled down to 25 C for the
CD measurements. The CD spectra were smoothed by adjacent
averaging.
−
1
−1
−3
pH = 5): kmax = 606 nm (e = 135 cm M ). UV-Vis (10 M,
1
1
mM sodium cacodylate buffer, pH = 7): kmax = 585 nm (e =
−1 −1 −3
36 cm M ). UV-Vis (10 M, water adjusted with 0.1 M
NaOH to a pH of 9.5): kmax = 577 nm (e = 136 cm− M ).
1
−1
Results and discussion
Cu(II) complex of 15-fluoro-15-methyl-1,4,7,10,13-pentaaza-
cyclohexadecan-14,16-dione (5). 15-Fluoro-15-methyl-1,4,7,10,
The new ligand 2 was synthesised analogously to the known
18
26
ligands 1 and 3 in a yield of 42%. The crystal structures
of 1·2HCl, 2·2HCl, and 3 were determined and are shown in
the ESI.† The presence of the hydrochlorides in 2 originates
from HCl abstraction of the solvent used (dichloromethane)
to grow the crystals. In the case of 3 the use of only partially
neutralised pentahydrochloride salt of tetraethylene pentamine,
1
0
3-pentaazacyclohexadecan-14,16-dione·1.5HCl (2) (50 mg,
.145 mmol) was added to a green solution of 31.82 mg of
copper acetate monohydrate (0.145 mmol) in 18 ml of dry
ethanol. The dark blue solution was stirred at r.t. for 3 h. The
ethanol was removed under vacuum to yield a blue residue,
which was passed through an ion-exchange column (DOWEX2
X8 20–50, previously equilibrated with a saturated solution of
ammonium acetate) with water. Yield 65 mg (95%). MS (ESI):
17
following the published procedure, caused the presence of the
HCl molecules.
+
−1
m/z (%) = 350 (100) [M − 2 acetate] . IR (KBr): 1692 cm
−
1
−3
Complexes of ligand 1
m C=Oligand, 1564 cm m C=Oacetate. UV-Vis (10 M, bi-distilled
−
1
−1
−3
water, pH = 5): kmax = 614 nm (e = 92 cm M ). UV-Vis (10 M,
The reaction of ligand 1 with either one equivalent of copper(II)
or zinc(II) acetate at r.t. in ethanol gave the blue copper(II)
complex 4 or the colourless zinc(II) complex 7, respectively.
The blue Cu compound 4 has an amide carbonyl IR vibration
1
8
mM sodium cacodylate buffer, pH = 7): kmax = 606 nm (e =
−1 −1 −3
9 cm M ). UV-Vis (10 M, water adjusted with 0.1 M NaOH
to a pH of 9.5): kmax = 565 nm (e = 120 cm− M ).
1
−1
−
1
at 1696 cm (in KBr) and an UV-Vis absorption maximum
dependent on the solvent and on the pH of the solution (Table 3).
It can also be seen that the kmax of absorbance of complex 4 moves
to lower values in more alkaline solutions, which corresponds
to the formation of the penta-nitrogen coordinated complex
Cu(II) complex of 15,15-difluoro-1,4,7,10,13-pentaazacyclo-
hexadecan-14,16-dione (6). CuCl
was dissolved in 4 ml of absolute ethanol. Silver triflate (87.6 mg,
.341 mmol) was added to the green solution, which became light
2
·H
2
O (29 mg, 0.171 mmol)
0
blue while a precipitate of AgCl formed. The mixture was stirred
for 30 min and was then filtrated. 15,15-Difluoro-1,4,7,10,13-
pentaazacyclohexadecan-14,16-dione (3) (50 mg, 0.17 mmol)
was added to the solution. The colour immediately turned dark
blue. The solution was stirred at r.t. for 3 h and the solvent
removed under vacuum to give a glassy blue product. Yield
(
Fig. 2).
This change in the coordination mode can also be seen from
the colour of the solution of the complex, which changes from
blue (at pH = 5, tri-amino coordination mode) to violet (at
pH = 9.5, penta-nitrogen coordination mode). Two different
crystal types were obtained depending on the solvent system
used to grow them. The blue crystals of 4 were obtained by
1
11 mg (quantitative). MS (FAB): m/z (%) = 338 (29%) [M −
F + H] , 356 (100) [M] . EA calcd for C13 : C,
3.83; H, 3.23; N, 10.69; found: C, 23.91; H, 3.57; N, 10.46. IR
+
+
H
21CuF
8
N
5
O
8
S
2
vapour diffusion of pentane in CH
2
2
Cl . The copper(II) centre is
2
−
1
−1
−1
in a distorted trigonal bipyramidal conformation (Fig. 3). The
three amine nitrogen atoms coordinate meridionally, while the
(
KBr): 1718 cm m C=Oligand, 1256 cm m S=Otriflate, 1031 cm
−
1
−3
m S–Otriflate, 639 cm d O–S–Otriflate. UV-Vis (10 M, EtOH):
−
1
−1
−3
k
max = 604 nm (e = 122 cm M ). UV-Vis (10 M, bi-distilled
−1 −1 −3
Table 3 Dependency of the kmax on solvent and pH for complexes 4, 5,
and 6
water, pH = 5): kmax = 596 nm (e = 84 cm M ). UV-Vis (10 M,
1
1
mM sodium cacodylate, buffer pH = 7): kmax = 587 nm (e =
−1 −1 −3
02 cm M ). UV-Vis (10 M, water adjusted with 0.1 M
Complex
Ethanol
pH = 5
pH = 7
pH = 9.5
NaOH to a pH of 9.5): kmax = 577 nm (e = 104 cm− M ).
1
−1
4
5
6
580
—
604
606
614
596
585
606
587
577
565
577
Zn(II) complex of 15-fluoro-1,4,7,10,13-pentaazacyclohexa-
decan-14,16-dione (7). 15-Fluoro-1,4,7,10,13-pentaazacyclo-
hexadecan-14,16-dione (1) (50 mg, 0.182 mmol) was added to
D a l t o n T r a n s . , 2 0 0 5 , 1 6 3 7 – 1 6 4 3
1 6 3 9

View Article Online
Table 4 Selected bond lengths of complexes 4, 5, 10, and 7
4
(M = Cu) 5 (M = Cu) 10 (M = Cu) 7 (M = Zn)
M(1)–N(1)
M(1)–N(4)
M(1)–N(7)
M(1)–N(10) 2.063(3)
M(1)–N(13)
1.960(3)
2.077(3)
2.030(3)
2.072(5)
2.026(4)
2.071(5)
2.058(3)
2.077(3)
2.030(3)
2.063(3)
2.185(2)
2.0876(19)
2.201(2)
M(1)–O(31) 2.158(3)
M(1)–O(33) 1.995(2)
2.199(4)
1.966(3)
2.0042(16)
2.0103(15)
Fig. 2 Equilibrium between the tri-amino and the bis deprotonated
penta-nitrogen coordinated metal complexes.
Fig. 4 ORTEP plot of the violet complex 10, solvent molecule omitted.
character of the carbonyl bonds. An enolate is not formed,
though this would be doubly stabilised by the adjacent carbonyl
groups. The pyramidalisation of C(15), the bond lengths of
C(15)–C(14) as well as C(15)–C(16) in 4 compared with 10,
and the existence of the experimentally found (but not refined)
hydrogen atom on C(15) exclude this possibility.
Fig. 3 ORTEP plot of the blue complex 4, solvent molecule omitted.
remaining positions are occupied by two acetate groups. One
acetate group is asymmetrically coordinated to the copper, since
the oxygen atom O(34) is only weakly bonded to the metal with
The colourless Zn(II) complex 7 shows an IR amide bond
˚
an O–Cu distance of 2.877(3) A. The other acetate is mono-
−1
absorption at 1700 cm ; again, this indicates that the metal ion
coordinated via O(31) to the metal centre, its second carbonyl
is not coordinated to the amide nitrogen atoms of the ligand.
In the crystal obtained from vapour diffusion of hexane into a
solution of chloroform, the Zn(II) centre is coordinated to the
three amine nitrogen atoms of the ligand and to the two acetate
molecules in a distorted trigonal bipyramidal mode (Fig. 5).
Each acetate group acts as a mono-dentate ligand, which is very
similar to the case of the copper complex 4. The fluorine atom
is axial relative to the plane of the azamacrocycle.
oxygen O(32) forming a weak hydrogen bond to N(7) (2.943
˚
(
6) A). The fluorine atom is in an axial position relative to the
macrocyclic ring.
By vapour diffusion of hexane into a solution of ethanol, we
obtained the violet crystals of 10 in which the copper(II) centre is
coordinated to the five nitrogen atoms of the ligand in a distorted
square pyramidal conformation due to previous deprotonation
of the amide nitrogen atoms (Fig. 4). The overall geometry,
with the exception of the additional fluorine atom present in 10,
closely resembles the analogue double deprotonated Ni complex
2
+
27
LM in Fig. 2 (M = Ni , R
1
= R
2
= H). The geometrical
similarity is not restricted to the molecular level: both complexes
crystallised in the same space group, with the unit cells of the
two crystal structures being almost identical. Obviously, the
extra fluorine atom in 10 does not disrupt the crystal packing
compared with Kimura’s structure. The complex 10 shows an
−
1
amide carbonyl IR vibration at 1618 cm (in KBr) compared
−
1
with 1696 cm in the case of 4. This is a clear sign that the metal
ion in 10 is coordinated to the three secondary amines and to the
two deprotonated amide nitrogen atoms of the ligand. This shift
of the IR frequencies of the amide bonds has also been observed
28
in the analogue nickel complexes by Rokita et al.
Comparing the carbonyl bond lengths of the two different
copper crystals 4 and 10, one can notice that they are longer in
˚
the purple complex 10 (1.267(4) A) than in the blue crystals 4
˚
(
1.227(4) A). On the other hand, the corresponding C(16/14)–
N(1/13) and C(15/15)–C(14/16) bond lengths are the same in
both complexes. Therefore the negative charge is only delocalised
onto the oxygen atoms, which indicates a stronger single bond
Fig. 5 ORTEP plot of the Zn complex 7 without solvent molecule.
1
6 4 0
D a l t o n T r a n s . , 2 0 0 5 , 1 6 3 7 – 1 6 4 3

View Article Online
Complexes of ligand 2
A titration of poly d(GC) with 2 was carried out to determine
if the ligand itself was able to induce B- to Z-DNA transition.
The CD spectrum of the poly d(GC) after the addition of the
product did not change indicating that the DNA was still in the
B-form (Fig. S4).†
The blue copper(II) complex 5 was synthesised by reacting
2
·2HCl with Cu(II)acetate and the chloride was exchanged by
acetate ions. As observed for 4, the UV-Vis absorption maximum
of the complex is dependent on the pH of the solution (Table 3).
The amide carbonyl IR vibration is at 1692 cm .
The interaction of Cu(OAc) with poly d(GC) was studied as
2
−
1
a control as to whether copper ions on their own can promote
the B- to Z-DNA transition. The CD spectrum showed that
even after the addition of 0.6 equivalents of copper versus DNA
phosphate groups the transition to Z-DNA had not yet reached
its midpoint (Fig. 7). Interestingly, Zacharias et al. mentioned
that copper(II) chloride was not able to induce the B- to Z-DNA
The X-ray structure of the blue crystals of 5, obtained by
vapour diffusion of pentane into a solution of dichloromethane,
shows that the Cu(II) centre is coordinated to the three amine
nitrogen atoms of the ligand and to the two acetate molecules
in a distorted square pyramidal mode (Fig. 6). Each acetate
group contributes as a mono-dentate ligand. A water molecule
is linked through a hydrogen bond to oxygen O(34). There are
furthermore twoweakhydrogen bonds betweenN(3) andoxygen
O(32) as well as between O(34) and N(4).
30
transition at all. This might be another case where chloride
31
anions inhibit the transition metal driven induction of Z-DNA.
The interaction of complex 7 with poly d(GC) is also shown in
Fig. 7. There is a change exclusively in the positive band of
the spectrum while the negative band is untouched. Thus the
zinc complex 7 binds to the B-DNA, but does not promote the
transition from B- to Z-DNA.
Fig. 6 ORTEP plot of the blue complex 5 without solvent molecule.
Fig. 7 CD spectra of poly d(GC) before and after the addition of 0.6
Complexes of ligand 3
equivalents of Cu(OAc) or 0.57 equivalents of 7.
2
Ligand 3 was reacted with either Cu(II)triflate or Zn(II)acetate
in ethanol at r.t. to form the complexes 6 and 8, respectively. The
blue copper(II) complex 6 was synthesised by reacting the ligand
The copper complexes 4, 5, and 6 behave in different ways
Fig. 8). In the case of 6, because of the strong influence of two
fluorine substituents on the acidity of the amide hydrogen atoms,
the equilibrium in Fig. 2 is shifted to the right side in favour of the
penta-nitrogen coordinated copper. The neutral complex is thus
not able to interact with the DNA and to promote its transition
from B- to Z-form.
(
3
with Cu(II)triflate, which had been previously obtained from
CuCl and silver triflate. The complex has a UV-Vis absorption
2
maximum dependent on the solvent and on the pH of the
solution (Table 3), and an amide carbonyl IR absorption at
−
1
1
718 cm . It was not possible to grow crystals of the complex.
Table 3 shows that the kmax of absorbance of complex 6 moves
to lower values as the pH of the solution increases, which
corresponds to the formation of the penta-nitrogen coordinated
complex. In this case the difference of kmax of absorbance between
the solution at pH = 5 and 9 is smaller than in the case of complex
4
. The second fluorine atom, with its electron withdrawing
action, enhances the acidity of the amide hydrogen atoms. Once
in aqueous solution, these hydrogen atoms are easily removed,
leading to the penta-nitrogen coordination of the metal centre.
Induction of Z-DNA by copper acetate, ligand 2, and complexes
4, 5, 6, and 7
The capability of the complexes 4, 5, 6, and 7 to induce the
B- to Z-DNA transition was studied. Aliquots (2–5 ll) of a
1
0
mM solution of each complex were added to an approximate
.1 mM solution of poly d(GC) in 1 mM sodium cacodylate
◦
buffer at pH = 7. The samples were warmed to 60 C to induce
a temporary and partial melting of the DNA. This ensured the
29
completion of the otherwise slow B- to Z-DNA transition.
Before measuring the spectra, the samples were cooled down to
Fig. 8 CD spectra of poly d(GC) after the addition of 0.66 equivalents
of 4, 0.64 equivalents of 5, and 0.67 equivalents of 6.
◦
25 C for 5 min.
D a l t o n T r a n s . , 2 0 0 5 , 1 6 3 7 – 1 6 4 3
1 6 4 1

View Article Online
The other two copper complexes 4 and 5 have an amazingly
different effect on poly d(GC). Looking at the CD signal at a
wavelength of 255 nm, one can easily follow the change from the
negative signal of B-DNA to the positive one of Z-DNA. The
two titrations of poly d(GC) with 4 and 5 are shown in Figs. 9
and 10, respectively. In the case of complex 4 the midpoint is
reached after the addition of 0.37 equivalents of the complex
versus DNA phosphates and the transition from B- to Z-DNA
occurs over a wide concentration range of added 4. In the case
of 5, the transition starts suddenly after the addition of 0.35
equivalents and the midpoint is achieved after the addition of
Conclusions
The new macrocyclic trisaza bisamide ligand 2 was synthesised
analogously to the known ligands 1 and 3 in a yield of 42%.
The used ligands can either bind to the metal centre by triple
1
8
26
17
nitrogen coordination (N3) with the secondary amines or after
double deprotonation of the two amides with all five nitrogen
atoms (N5). Here, the overall charge of the complex is zero
and the ligand occupies five out of six coordination sites of a
distorted octahedron. The N5 coordination mode is favoured if
one or two fluorine substituents at the C(15) position are present
and if the pH is high in the case of aqueous solutions. When 15-
fluoro-1,4,7,10,13-pentaazacyclohexadecan-14,16-dione (1) was
reacted with nickel acetate, all three secondary amines plus one
oxygen of an amide group were found to coordinate to the
27
0
.45 equivalents of the complex versus DNA phosphates. Figs. 9
and 10 additionally show the simulation of the experimental
curves using the Hill equation for an equilibrium between two
32,33
states:
26,28
nickel in the crystal structure.
never observed during our studies. All our crystal structures
of complexes in the N3 mode for LH
overall geometry of the nickel acetate complex with 1,4,7,10,13-
Such an NNNO mode was
y
0
− y
∞
x
y = y
∞
+ 1
ꢀ
ꢁ
n
2+
2
M
rather resemble the
+
K
1
7
pentaazacyclohexadecan-14,16-dione.
where y is the spectroscopic response, x is the concentration
of the added reagent, y is the
Circular dichroism titrations of poly d(GC) with the metal
complexes showed the ability of the systems studied to promote
the B- to Z-DNA transition. Ligand 2, the copper complex 6
with the bis fluoro substituted ligand 3 and the zinc complex
0
is the response when x = 0, y
∞
maximal response when x = ∞, K is the concentration of x at the
midpoint, and n is the degree of cooperativity that determines
the slope of the curve. The fitting was achieved by varying y
0
,
7
were not able to induce any change in the direction of Z-
y
∞
, and K to minimise the least square deviations of yexperimental
DNA. In the case of 6, the CD spectrum of the DNA actually
showed no change at all, indicating that the complex was
not even interacting with the B form of DNA. Therefore we
assume that the bis fluoro substitution causes the complex at
the experimental conditions of pH 7 to be in the neutral LM
form. The electrostatics together with the shielding effect of the
ligand might explain the absence of any interaction with the
DNA. Copper acetate generated a partial transition. This is
remarkable since copper chloride was inactive in generating Z-
versus ycalculated. For the complexes 4 and 5 the cooperativity (n)
was found to be 3.4 and 7.3, respectively. It seems that the in-
troduction of the neither sterically demanding nor electronically
very influential methyl group has an unexpectedly large influence
upon the cooperativity of the transition. Lacking a molecular
structure of the DNA to copper adducts, we can only assume
that the additional methyl group is involved in a hydrophobic
interaction with the DNA.
30
31
DNA, most likely due to chloride inhibition. Both complexes
and 5 induced Z-DNA with a midpoints of 0.37 and 0.45
4
equivalents of the complex versus DNA phosphates, respectively.
The cooperativities of 4 and 5 were found to be 3.4 and 7.3,
respectively. A possible hypothesis to explain this remarkable
difference might be that the additional methyl group of 5 is
involved in a hydrophobic interaction with the DNA.
In the future the amount of metal complex needed for the B-
to Z-DNA transition should be reduced. The use of dinuclear
metal complexes might help to achieve this goal. We are currently
exploring this hypothesis.
Acknowledgements
This work was supported by the Swiss National Science Foun-
dation and the Forschungskredit of the University of Z u¨ rich.
We thank Professor Dr R. Alberto, Professor Dr R. Sigel,
Dr J. K. Pak, Dr B. Knobloch and Ph. Kurz for valuable
discussions. We further thank B. Spichtig for performing the
FAB-MS analyses, B. and H. Spring for doing the elemental
analyses.
Fig. 9 Titration of poly d(GC) with 4 monitored by CD at 255 nm.
References
1
2
A. Herbert and A. Rich, J. Biol. Chem., 1996, 271, 11595.
A. Rich, A. Nordheim and A. H.-J. Wang, Annu. Rev. Biochem.,
1
984, 53, 791.
3
4
5
6
M. Behe and G. Felsenfeld, Proc. Natl. Acad. Sci. USA, 1981, 78,
1
619.
R. V. Gessner, C. A. Frederick, G. J. Quigley, A. Rich and A. H.-J.
Wang, J. Biol. Chem., 1989, 264, 7921.
X. Qu, J. O. Trent, I. Fokt, W. Priebe and J. B. Chaires, Proc. Natl.
Acad. Sci. USA, 2000, 97, 12032.
Y. Xu, Y. X. Zhang, H. Sugiyama, T. Umano, H. Osuga and K.
Tanaka, J. Am. Chem. Soc., 2004, 126, 6566.
7
8
G. V. Fazakerley, Nucleic Acids Res., 1984, 12, 3643.
S. M. Middendorf and H. Diebler, Ber. Bunsen-Ges. Phys. Chem.,
1
996, 100, 641.
Fig. 10 Titration of poly d(GC) with 5 monitored by CD at 255 nm.
9 T. Schoenknecht and H. Diebler, J. Inorg. Biochem., 1993, 50, 283.
1
6 4 2
D a l t o n T r a n s . , 2 0 0 5 , 1 6 3 7 – 1 6 4 3

View Article Online
1
1
1
1
1
1
1
1
1
0 J. H. van de Sande, L. P. McIntosh and T. M. Jovin, EMBO J., 1982,
, 777.
1 B. Spingler, P. Antoni, B. Wild and G. Viola, manuscript in
preparation.
2 B. H. Geierstanger, T. F. Kagawa, S.-L. Chen, G. J. Quigley and P. S.
Ho, J. Biol. Chem., 1991, 266, 20185.
3 A. Woisard, G. V. Fazakerley and W. Guschlbauer, J. Biomol. Struct.
Dyn., 1985, 2, 1205.
22 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C.
Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori and R.
Spagna, J. Appl. Crystallogr., 1999, 32, 115.
1
23 SHELXL97-2: Program for the Refinement of Crystal Structures,
University of G o¨ ttingen, Germany, 1997.
24 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
25 N. Ishikawa, A. Takaoka and M. K. Ibrahim, J. Fluorine Chem.,
1984, 25, 203.
26 C.-C. Cheng, PhD Dissertation, State University of New York, Stony
Brook, 1993.
27 R. Machida, E. Kimura and Y. Kushi, Inorg. Chem., 1986, 25, 3461.
28 C.-C. Cheng, J. Gulia, S. E. Rokita and C. J. Burrows, J. Mol. Cat.
A: Chem., 1996, 113, 379.
29 H.S. Basu, H. C. A. Schwietert, B. G. Feuerstein and L. J. Marton,
Biochem. J., 1990, 269, 329.
4 H.-C. Shih, N. Tang, C. J. Burrows and S. E. Rokita, J. Am. Chem.
Soc., 1998, 120, 3284.
5 E. Kimura, A. Sakonaka and M. Nakamoto, Biochim. Biophys. Acta,
1
981, 678, 172.
6 E. Kimura, A. Sakonaka, R. Machida and M. Kodama, J. Am. Chem.
Soc., 1982, 104, 4255.
7 Z. Zhang, J. L. Petersen and A. M. Stolzenberg, Inorg. Chem., 1996,
3
5, 4649.
30 W. Zacharias, J. E. Larson, J. Klysik, S. M. Stirdivant and R. D.
Wells, J. Biol. Chem., 1982, 257, 2775.
31 B. Spingler, Inorg. Chem., 2005, 44, 831.
8 C.-C. Cheng, S. E. Rokita and C. J. Burrows, Angew. Chem., Int. Ed.,
993, 32, 277.
1
1
2
2
9 E. Schlittler and J. Hohl, Helv. Chim. Acta, 1952, 35, 29.
0 STOE-IPDS Software package, Darmstadt, Germany, 1998.
1 G. M. Sheldrick, Acta Crystallogr. Sect. A: Fundam. Crystallogr.,
32 C.-W. Chen, R. H. Knop and J. S. Cohen, Biochemistry, 1983, 22,
5468.
33 A. De Lean, P. J. Munson and D. Rodbard, Am. J. Physiol., 1978,
1
990, A46, 467.
235, E97.
D a l t o n T r a n s . , 2 0 0 5 , 1 6 3 7 – 1 6 4 3
1 6 4 3