Structural Features of Two Distinct Molecular Complexes of

Copper(II) Cationic Porphyrin and Deoxyribonucleotides

PETER MOJZES,1 PETR PRAUS,1 VLADIMIR BAUMRUK,1 PIERRE-YVES TURPIN,2 PAVEL MATOUSEK,3
MICHAEL TOWRIE3
1
Institute of Physics, Charles University, Ke Karlovu 5, Prague 2, CZ-12116, Czech Republic

2
L.P.B.C., Université P. & M. Curie, 4 Place Jussieu, Paris Cedex 05, F-75252, France

3
Central Laser Facility, CLRC, Rutherford Appleton Laboratory, Chilton, Didcot, OX11 0QX, United Kingdom

Received 8 September 2001; revised 7 November 2001; accepted 19 November 2001

Published online 25 April 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/bip.10098

ABSTRACT: The associations of the water-soluble cationic copper(II)-5,10,15,20-meso-

tetrakis(4-N-methylpyridyl) porphyrin (CuP) with d(pT)9 oligothymidylate and its
building blocks deoxythymidine (dT) and deoxythymidine 5⬘-monophosphate (dTMP)
were investigated by spectrophotometric titration [absorption, nanosecond transient
resonance Raman (ns-RR) and picosecond time-resolved resonance Raman (ps-TR3)
spectroscopies] to elucidate the structural requirements for the CuP exciplex formation
in molecular complexes with unchained mononucleotides. In the d(pT)9 a factor anal-
ysis and global fit of the CuP absorption spectra revealed the formation of a single
spectral species attributable to a 1 : 1 CuP 䡠 d(pT)9 complex throughout a wide range of
d(pT)9/CuP ratios (0 –10). Using ps-TR3 spectroscopy, the CuP 䡠 d(pT)9 complex was
shown to be fully responsible for exciplex formation. In contrast, CuP mixed with dTMP
([dTMP]/[CuP] ⬍ 3000) yielded two spectroscopically distinct types of molecular com-
plexes with 1 : 1 (CuP 䡠 dTMP) and 1 : 2 (CuP 䡠 (dTMP)2) (or even higher for [dTMP]/
[CuP] ⬎ 3000) stoichiometry, the latter being spectroscopically identical to the CuP
䡠 d(pT)9 and providing a microenvironment favorable for exciplex formation to the same
extent as the oligothymidylate. On the other hand, the 1 : 1 CuP 䡠 dTMP complex
(prevailing for [dTMP]/[CuP] ⬍ 100) yielded no exciplex features. Similar behavior was
observed for the CuP complexed with dT. To explain the difference in the ability of the
CuP 䡠 dTMP and CuP 䡠 (dTMP)2 species to form the exciplex, two types of molecular
complexes were suggested and discussed, differing in the orientation of the thymine
planes with respect to the porphyrin macrocycle. © 2002 Wiley Periodicals, Inc. Biopolymers
(Biospectroscopy) 67: 278 –281, 2002

Keywords: Raman spectroscopy; copper porphyrins; nucleic acids; exciplex forma-

tion; intermolecular complexes

INTRODUCTION

Correspondence to: P. Mojzes (mojzes@karlov.mff.cuni.cz). Water-soluble cationic (metallo)porphyrins com-

Contract grant sponsor: European Commission on Frame
work V IHP Large Scale Facility Access Programme; contract
grant number: USEV11C3/00.
Contract grant sponsor: Ministry of Education, Czech Re-
public; contract grant numbers: VS97113, ME271.
Contract grant sponsor: Grant Agency, Czech Republic;
contract grant number: 203/01/1013.
Biopolymers (Biospectroscopy), Vol. 67, 278 –281 (2002)
© 2002 Wiley Periodicals, Inc.
278
plexed with nucleic acids are widely investigated
for their possible use in the photodynamic ther-
apy of cancer. Besides promising medical applica-
tions, cationic porphyrins are also interesting as
specific probes of local nucleic acid base pairs
and/or sequence composition, structure, confor-
mation, and dynamics.1,2
 STRUCTURAL FEATURES OF CuP/DEOXYRIBONUCLEOTIDE COMPLEXES
Among porphyrins studied extensively com-
plexed with nucleic acids, the copper(II) deriva-
tive of 5,10,15,20-meso-tetrakis(4-N-methylpyri-
dyl) porphyrin (CuP) attracted particular interest
because of its ability in its excited (d– d) state to
form a transient axially coordinated exciplex with
an oxygen-containing ligand in close proximity.1,2
Because a CAO group of thymine (or uracil in
RNA) was identified as the CuP axial ligand,1
formation of the CuP*CAO exciplex can serve as
an indicator of the proper distance (⬃ 0.2 nm) and
appropriate orientation between the CAO and
the center of CuP (i.e., the more or less perpen-
dicular arrangement of the CuP macrocycle and
the thymine plane that excludes a stacking inter-
action between them),1 thus providing structural
information about the CuP microenvironment.
The exciplex formation can be conveniently fol-
lowed by resonance Raman (RR) spectroscopy, be-
cause under high power pulsed excitation specific
exciplex Raman bands appear at ⬃ 1346 (␯*4) and
1550 (␯*2) cm⫺1 in pairs with their ground-state
counterparts located at ⬃ 1366 (␯4) and 1570
(␯2) cm⫺1, respectively (Fig. 1). Besides the
CuP*CAO species, a CuP*H2O exciplex (having a
Figure 1. Normalized transient ns-RR spectra of the
CuP mixed with dTMP. The nucleotide/porphyrin ra-
tios are ⬃ 0 (spectrum a), ⬃ 1 (spectrum b), ⬃ 50
(spectrum c), ⬃ 100 (spectrum d), ⬃ 200 (spectrum e),
⬃ 500 (spectrum f), ⬃ 1000 (spectrum g), and ⬃ 3000
(spectrum h).
279
water molecule as its fifth axial ligand) was also
formed in aqueous solutions containing free
CuP.2 Because both exciplex species exhibit vir-
tually identical Raman features,2 they cannot be
discerned from each other by nanosecond tran-
sient (ns)-RR spectra in the systems containing
fractions of both complexed and free CuP. Never-
theless, the CuP*H2O and CuP*CAO species can
be distinguished by their picosecond time-re-
solved RR (ps-TR3) spectra due to different relax-
ation kinetics, because their corresponding life-
times differ by orders of magnitude (⬃ 10 ps and
⬃ 1–3 ns, respectively).2
Previously, exciplex formation was studied
preferentially with double-stranded polynucleoti-
des and natural dsDNA having regular double-
helical structure. Apart from studies1–3 showing
that the CuP exciplex can also be formed effec-
tively with single-stranded oligo- and polythymi-
dylates, no systematic investigation was carried
out with building blocks of nucleic acids. Al-
though molecular complexes of ground-state CuP
with various nucleosides and nucleotides were
investigated by spectrophotometric titrations,4 no
correspondent study is yet published for the elec-
tronically excited CuP. The aim of the present
study is to clarify whether a microenvironment
favorable for exciplex formation can be created
effectively by interaction with unchained thymine
nucleotides or whether some regular structure of
the oligonucleotide is required.
EXPERIMENTAL
Titration of the CuP solution [pH ⬃ 7; ionic
strength (␮) ⬃ 0.6M] by deoxythymidine (dT),
deoxythymidine 5⬘-monophosphate (dTMP), and
oligothymidylate d(pT)9 was monitored by ab-
sorption in the CuP Soret band. Keeping the total
concentration of the CuP fixed (⬃ 3 ⫻ 10⫺5M), the
nucleotide/porphyrin ratio (R: [dTMP]/[CuP],
[d(pT)9]/[CuP]) was varied stepwise to cover the
range in which substantial changes of the CuP
absorption are detectable [R ⬃ 0–7000 and
⬃ 0 – 40 for dT/dTMP and d(pT)9, respectively].
Exciplex formation in the studied mixtures was
followed by ns-RR spectroscopy (␭exc ⫽ 425 nm,
20-ns pulse duration).1,2 To eliminate contribu-
tions from the fast relaxing CuP*H2O species, the
efficiency of the CuP*CAO exciplex formation
was evaluated from the pump and probe ps-TR3
spectra (⌬t ⫽ ⫹50 ps) using a single color ar-
rangement (␭pump ⫽ ␭probe ⫽ 420 nm, 1-ps pulse
280 MOJZES ET AL.

K2
¡ CuP 䡠 共dTMP兲2
CuP 䡠 dTMP ⫹ dTMP ¢
O

关CuP 䡠 共dTMP兲2 兴
K2 ⫽ (2)
关CuP 䡠 dTMP兴关dTMP兴

For [dTMP]/[CuP] ⬍ 3000 and [d(pT)9]/[CuP]

⬍ 10, an excellent congruence of the global fits
with the presumed models was obtained (giving
stability constants K 1 ⫽ 500 ⫾ 100 M ⫺1 , K 2
⫽ 250 ⫾ 50 M ⫺1 , and K ⫽ 1.2 ⫾ 0.1 ⫻ 10 6
M ⫺1 , respectively). Nevertheless, some devia-
tions appeared at extremely high nucleotide load-
ing. The discrepancies may be explained by
Figure 2. The distribution of the CuP complexes (de-
postulating that there is formation of spectroscop-
termined from the absorption spectra) as a function of
the dTMP/CuP ratio versus the effectiveness of the
ically indiscernible complexes having stoichiome-
CuP*CAO exciplex formation as determined from the tries higher than the presumed ones. It is worth-
ps-TR3 spectra; ⌬t ⫽ ⫹50 ps. while to note that the extracted absorption
spectrum of the 1 : 2 CuP 䡠 (dTMP)2 complex is
very close to that of CuP 䡠 d(pT)9.
As illustrated by the set of normalized ns-RR
duration).5 Undesirable RR scattering from the spectra (Fig. 1), the increase of the dTMP/CuP
pump pulses was rejected by means of a Kerr gate ratio in the mixtures results in the appearance of
(3-ps resolution).5 The ns-RR and ps-TR3 spectra the exciplex Raman features (␯*2 and ␯*4 bands),
were normalized by using an internal intensity whose relative intensity increases with R. Essen-
standard (⬃982 cm⫺1 Raman band of 0.2M SO42⫺ tially, the same dependence was observed in the
added to the samples). pump and probe ps-TR3 spectra (⌬t ⫽ ⫹50 ps).
Under the experimental conditions used in the
present study, the extent of the spectral changes
RESULTS AND DISCUSSION in the ps-TR3 spectra (i.e. overall decrease of the
total Raman signal and intensity ratio between
Factor analysis (singular value decomposition)6 of
the sets of absorption spectra revealed the pres-
ence of two spectroscopically distinct types of CuP
molecular complexes formed in mixtures with dT
and dTMP, whereas mixtures with d(pT)9 were
shown to provide only one complex species. The
results of the factor analysis were fitted6 to equi-
librium equations indicating the formation of a
single CuP 䡠 d(pT)9 species (Eq. 1) and 1 : 1 CuP 䡠
dTMP and 1 : 2 CuP 䡠 (dTMP)2 complexes [Eqs. (1)
and (2)]. The spectra of the pure species were
extracted (not shown here) and their relative frac-
tions evaluated as a function of the R (see Figs. 2
and 3).

K1
CuP ⫹ dTMP O
¡ CuP 䡠 dTMP
¢ Figure 3. The distribution of the CuP complexes (de-
termined from the absorption spectra) as a function of
the d(pT)9/CuP ratio versus the effectiveness of the
关CuP 䡠 dTMP兴 CuP*CAO exciplex formation as determined from the
K1 ⫽ (1)
关CuP兴关dTMP兴 ps-TR3 spectra; ⌬t ⫽ ⫹50 ps.
 STRUCTURAL FEATURES OF CuP/DEOXYRIBONUCLEOTIDE COMPLEXES
the exciplex and ground-state features) for the
CuP complexed with dTMP at R ⬎ 3000 was
found to be nearly identical to that of CuP com-
plexed with d(pT)9, which was [d(pT)9]/[CuP] ⬎ 1.
Thus, at extremely high nucleotide loading the
CuP formed molecular complexes with a microen-
vironment similar to that generated by the oligo-
nucleotide having a regular secondary structure.
To identify complex species responsible for the
CuP*CAO exciplex formation in mixtures with
dTMP, the effectiveness of the exciplex formation
(evaluated directly from the ps-TR3 spectra by
factor analysis) was confronted with the relative
fractions of the complex species with 1 : 1 (CuP
䡠 dTMP) and 1 : 2 (CuP 䡠 (dTMP)2) stoichiometries,
evaluated from the absorption spectra.
As shown in Figure 2, only the CuP 䡠 (dTMP)2
species contributes to the CuP*CAO exciplex for-
mation. Nevertheless, slight but evident devia-
tion of the exciplex effectiveness from the CuP
䡠 (dTMP)2 distribution curve seems to suggest
that there is a substantial participation of
complexes with stoichiometries higher than 1 : 2.
In contrast, in the d(pT)9 a single spectral spe-
cies attributable to a 1 : 1 CuP 䡠 d(pT)9 complex
was revealed by spectrophotometric titration. As
shown in Figure 3, the same complex is fully
responsible for the CuP exciplex formation
throughout a wide range of d(pT)9/CuP ratios.
The dT with no charged phosphate group showed
behavior similar to that of dTMP. This finding is
fully consistent with the suggestion that the van
der Waals’ and hydrophobic types of interactions
are more important in stabilizing the CuP com-
plexes with nucleosides and mononucleotides
than electrostatic attraction between the phos-
phate groups and the cationic porphyrins.4
CONCLUSIONS
By means of Kerr gated ps-TR3 spectroscopy, it
was shown that a CuP*CAO exciplex can be ef-
fectively formed in molecular complexes with
dTMP and dT. Nevertheless, among the two spec-
troscopically distinct complex species formed,
only the complexes with stoichiometries of at
least 1 : 2 seem to provide a microenvironment
favorable for exciplex formation. Considering con-
temporary thoughts about structural require-
ments for axial coordination that gives rise to an
exciplex1,2 (i.e., assuming an appropriate orienta-
tion between the CAO group and the center of
CuP, where a near perpendicular arrangement of
281
the porphyrin macrocycle and the thymine plane
excludes the stacking interaction between them),
some structural features of the CuP molecular
complexes with monomer units may be suggested.
The 1 : 1 species can be characterized by a near
“parallel” arrangement of the porphyrin macro-
cycle and the thymine plane, thus stabilized
mainly by stacking interactions between them. It
is formed in the CuP mixtures with dT and dTMP
at low nucleotide/porphyrin ratios (R ⬍ 100);
and, because of the unsuitable orientation of the
CAO group with respect to the porphyrin center,
it is not favorable for the CuP*CAO exciplex for-
mation.
The 1 : 2 and higher species are characterized
by a more or less “perpendicular” arrangement of
the porphyrin macrocycle and self-stacked thy-
mine planes. It is stabilized by van der Waals’
hydrophobic interactions, of a type other than
stacking, between the CuP macrocycle and the
self-stacked thymines. It is formed in the CuP
mixtures with dT and dTMP at extremely high
nucleotide/porphyrin ratios. Its structure is close
to that of CuP complexed with oligothymidylates,
and it is thus favorable for the CuP*CAO exci-
plex formation.
The ps-TR3 experiments were performed within the
CLF, CLRC Rutherford Appleton Laboratory under the
European Commission Framework V IHP Large Scale
Facility Access Programme. The financial support
awarded by the Ministry of Education and Grant
Agency of the Czech Republic is acknowledged.
REFERENCES
1. Mojzes, P.; Chinsky, L.; Turpin, P.-Y. J Phys Chem
1993, 97, 4841– 4847.
2. Kruglik, S. G.; Mojzes, P.; Mizutani, Y.; Kitagawa,
T.; Turpin, P.-Y. J Phys Chem B 2001, 105, 5018 –
5031.
3. Chirvony, V. S.; Galievsky, V. A.; Sazanovich, I. V.;
Turpin, P.-Y. J Photochem Photobiol 1999, 52, 43–
50.
4. Pasternack, R. F.; Gibbs, E. J.; Gaudemer, A.; An-
tebi, A.; Bassner, S.; De Poy, L.; Turner, D. H.;
Williams, A.; Laplace, F.; Lansard, M. H.; Merienne,
C.; Perrée-Fauvet, M. J Am Chem Soc 1985, 107,
8179 – 8186.
5. Matousek, P.; Towrie, M.; Stanley, A.; Parker, A. W.
Appl Spectrosc 1999, 53, 1485–1489.
6. Hanus, J.; Chmelova, K.; Stepanek, J.; Turpin, P.-
Y.; Bok, J.; Rosenberg, I.; Tocik, Z. J Raman Spec-
trosc 1999, 30, 667– 676.