| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Biophys J, July 2002, p. 371-381, Vol. 83, No. 1
and
Departments of *Chemistry and Physics, Yeungnam
University, Kyoungsan City 712-749, Republic of Korea
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
|
|---|
By utilizing circular and linear dichroism, the binding mode of meso-tetrakis(n-N-methylpyridiniumyl)porphyrin (n = 2, 3, 4) to various DNAs was studied in this work. 2-N-(methylpyridiniumyl)porphyrin(o-TMPyP), in which rotation of the periphery pyridinium ring is prevented, exhibits similar spectral properties when bound to DNA, poly[d(G-C)2] and poly[d(A-T)2], suggesting a similar binding mode. Close analysis of the spectral properties led us to conclude that o-TMPyP sits in the major groove. However, both 3-N- and 4-N-(methylpyridiniumyl)porphyrin (m- and p-TMPyP), of which the periphery pyridinium ring is free to rotate, intercalate between the basepairs of DNA and poly[d(G-C)2]. In the presence of poly[d(A-T)2], m-TMPyP exhibits a typical bisignate excitonic CD spectrum in the Soret band, while p-TMPyP shows two positive CD bands. The excitonic CD spectrum of the m-TMPyP-poly[d(A-T)2] complex and the positive CD band of the o-TMPyP-poly[d(A-T)2] complex were not affected by the presence of the minor groove binding drug, 4',6-diamidino-2-phenylindole (DAPI), indicating that this porphyrin is bound in the major groove. In contrast, two positive CD bands of the p-TMPyP-poly[d(A-T)2] complex altered in the presence of DAPI. From the changes in CD spectrum and other spectral properties, a few possible binding modes for p-TMPyP to poly[d(A-T)2] are suggested.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
|
|---|
The interaction of cationic porphyrins with
nucleic acids has been a subject of intensive investigation (for review
see Pasternack and Gibbs, 1996
; Pratviel et al., 1989) since the
pioneering work of Fiel and co-workers (Fiel et al., 1979), owing to
its potential applications in cancer chemotherapy and antiviral activity.
The three-binding mode for the porphyrin-DNA complex has been generally
accepted (Fiel, 1989; Marzilli, 1990), namely intercalation, outside
self-stacking, and outside random binding. These structures of the
porphyrin-DNA complexes have been extensively characterized using a
variety of physical techniques. NMR, equilibrium dialysis, flow
dichroism, and viscometry measurements on oligo and polynucleotides have supported that porphyrins intercalate into GC-rich regions and
that they bind in an outside manner at AT sites (for review see Fiel,
1989; Strickland et al., 1990). Subsequent NMR study has shown that the
intercalation occurs only at the 5'CG3' site, not at 5'GC3' or other
sites (Marzilli et al., 1986; Guliaev and Leontis, 1999). Compared to
the intercalation mode, outside self-stacking and outside random
binding are less well-characterized. Porphyrins that exhibit an outside
self-stacking mode include
meso-tetrakis(p-tri-N-methyl-pyridiniumyl)porphyrin (Marzilli et al., 1986; Banville et al., 1986; LeDoan et al., 1987),
meso-tetrakis[4-[(3-(trimethylammonio)prophyloxy)phenyl]porphyrin (T
OPP) (Mukundan et al., 1994, 1995),
trans-bis(N-methylpyridinium-4-yl)diphenyl porphyrin (Pasternack et al., 1998; Ismail et al., 2000), and some
copper(II) porphyrins (Pasternack et al., 2001). These porphyrins stack
outside DNA and induce DNA aggregation. The outside random binding
modes of porphyrins were also reported (Banville et al., 1986; Carvlin
and Fiel, 1983). In this binding mode, porphyrins interact with the
phosphate group of DNA through electrostatic interaction. This binding
mode is in competition with intercalation. Exceptions in the binding
mode have also been suggested for various porphyrins (Kuroda and
Tanaka, 1994; Sehlstedt et al., 1994; Schneider and Wang, 1994;
Lipscomb et al., 1996; Yun et al., 1998; Barnes and Schreiner, 1998).
When Cu(II)
[meso-tetra(N-methyl-4-pyridyl)porphyrin] forms
a complex with hexamer duplex [d(CGATCG)]2,
porphyrin intercalates: the cytosine base of 5'CGA3' flips out from DNA
(Lipscomb et al., 1996). This result from an x-ray structure was
criticized later by NMR study (Barnes and Schreiner, 1998). The groove
binding mode, mainly based on the extensive circular dichroism (CD) and linear dichroism (LD) study was also suggested (Kuroda and Tanaka, 1994; Sehlstedt et al., 1994; Schneider and Wang, 1994; Lipscomb et
al., 1996; Yun et al., 1998). The side of the porphyrin ring fits into
the minor groove of DNA or locates in the major groove by electrostatic
interaction between the negatively charged phosphate group of DNA and
the positively charged pyridinium ring of porphyrin.
The binding mode could be modulated by the nature of the metal ion and
the size and location of the substituent groups on the periphery of the
porphyrin. Generally, the free bases and square planner complexes such
as Ni2+ and Cu2+
intercalate between DNA basepairs (to the GC site). For the
porphyrin-metal complex, having axially bound ligands such as
Co3+, Mn3+, and
Fe3+ or those with bulky substituents on the
periphery on the structure, intercalation is blocked and "outside
binding" occurs. Recent studies showed that intercalation versus
outside binding may also be influenced by the charge on the porphyrin
core (Kuroda et al., 1990; Marzilli et al., 1992) and the ionic
strength of the medium, which affects self-association of the porphyrin
(Pasternack et al., 1993; Dixon and Steullet, 1998). For instance, when
the n-butyl group is attached to the periphery pyridinium
ring, porphyrin intercalates between the DNA basepairs (Sehlstedt et
al., 1994), while when replaced by the tri-methyl group,
porphyrin exhibits an outside self-stacking binding mode (Banville et
al., 1986), therefore indicating the importance of the steric effect of
the periphery pyridinium ring.
In this work we systematically investigated and classified the
binding geometry of
meso-tetrakis(n-N-methylpyridiniumyl)porphyrin (where n = 2, 3, 4; Fig.
1; referred to as o-TMPyP,
m-TMPyP, and p-TMPyP, respectively) to natural
calf thymus DNA (referred to as DNA) and
poly[d(A-T)2] and
poly[d(G-C)2] at a low porphyrin-to-DNA base
ratio using conventional spectroscopic methods including CD and LD. The
goal of this work is to study the influence of steric hindrance of the
periphery pyridinium ring on porphyrin-DNA binding. The rotation of the
periphery pyridinium ring of o-TMPyP requires high energy,
and was expected to prevent porphyrin intercalation (Carvlin et al.,
1982; Fiel, 1989; Marzilli, 1990), while this moiety is free to rotate
without any hindrance in m- and p-TMPyP. The
porphyrin-to-DNA base ratio was kept very low (lower than one porphyrin
molecule per 20 nucleobases) in this work because porphyrin is expected
to exhibit a homogeneous binding mode that results in well-defined
spectral properties at low mixing ratios (Sehlstedt et al., 1994; Yun
et al., 1998). Poly[d(A-T)2] and poly[d(G-C)2] were chosen because these two
polynucleotides exhibit the two representative binding modes for
p-TMPyP, namely intercalation between the GC basepair and
outside binding at the AT-rich site. From this study we also hope to
relate the known "outside stacking" and "outside random
binding" to recently suggested minor or major groove binding (Kuroda
and Tanaka, 1994; Sehlstedt et al., 1994; Schneider and Wang, 1994;
Lipscomb et al., 1996; Yun et al., 1998).
|
| |
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
|
|---|
Materials
Porphyrins were purchased from Midcentury (Chicago, IL) and used
without further purification. The extinction coefficients for the
o-TMPyP, m-TMPyP, and p-TMPyP were
determined spectrophotometrically to be, respectively, 
413
nm = 2.39 × 105
cm
1 M1,
417 nm = 2.78 × 105 cm1
M1, and 421 nm = 2.45 × 105 cm1
M1 in 5 mM cacodylate buffer at pH 7.0. Synthetic polynucleotides were purchased from Pharmacia (Seoul,
Korea) and DNA from Sigma (Seoul, Korea). Synthetic polynucleotides and
DNA were dissolved in 5 mM cacodylate buffer, containing 100 mM NaCl
and 1 mM EDTA at pH 7.0 by exhaustive stirring at 4°C, followed by
dialyzing against 5 mM cacodylate buffer, pH 7.0 at 4°C. The buffer
was changed six times at 5-h intervals and this buffer was used
throughout this work. The DNA concentrations were determined
spectrophotometrically using molar extinction coefficients:
258 nm = 6700 cm1
M1, 262 nm = 6600 cm1M1,
254 nm = 8400 cm1
M1 for DNA,
poly[d(A-T)2], and
poly[d(G-C)2], respectively. The DNA
concentrations given in this work thus indicate the concentration of
the nucleobases. The mixing ratio, R, is defined by the
ratio [porphyrin]/[nucleobase]. The samples with various
R ratios were prepared by adding aliquots of concentrated
porphyrin solution to DNA solution (typically 10-20 µl porphyrin
solution to 2 ml DNA solution) and the volume corrections were made.
Because the binding mode is affected by ionic strength and
porphyrin-DNA mixing ratio and stacking of porphyrin itself in solution
(Ismail et al., 2000), an extreme caution for the order of mixing and
the concentration of porphyrin stock solution was taken. All
measurements were performed at an ambient temperature.
Absorption and circular dichroism
Absorption spectra were recorded either on a Jasco V-550 or on a
Hewlett Packard 8452A diode array spectrophotometer using a 1 cm quartz
cell. Porphyrins do not possess any chiral center, but acquire an
induced CD spectrum upon binding to polynucleotides. The origin of
induced CD for the achiral drug-DNA complex, which appears in the drug
absorption region, is primarily an effect of coupling between the
transitions of drugs and the bases of the nucleic acid host (Lyng et
al., 1991, 1992). CD spectra were measured on a Jasco J-715
spectropolarimeter (displaying the CD in millidegrees ellipticity)
using a 1-cm cell. The CD spectrum was averaged over an appropriate
number of scans when necessary. The CD spectrum of the
m-TMPyP-poly[d(A-T)2] complex was
recorded right after mixing (only one scan) with extreme caution
because it changes with time (see below).
Reduced linear dichroism
Linear dichroism (LD) on the flow-aligned porphyrin-DNA
complexes was measured in a Couette cell as described by Nordén
and his co-workers (Nordén and Tjerneld, 1976; Nordén and
Seth, 1985; Nordén et al., 1992) on a Jasco J-500C
spectropolarimeter. The Measured LD spectrum is divided by the
isotropic absorption spectrum, Aiso,
to give LDr, which is related to an angle, 
,
between the light-absorbing transition dipole and the local helix axis
of DNA.
|
). An effective angle was calculated from the
LDr value observed for the in-plane 
*
transitions in the Soret band by assuming an effective angle of 86°
at 260 nm for the DNA bases with respect to the flow direction
(Matsuoka and Nordén, 1983). A larger or comparative
LDr magnitude in the Soret band compared to that
in the DNA absorption region is indicative of intercalation, whereas
small values are consistent with groove binding. In the present
case, in-plane * transitions are degenerated in the Soret
band. The tilt angle is then obtained by replacing
cos2 in the above equation by
1/2(cos2 
) (Härd and
Nordén, 1986).
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
|
|---|
Absorption spectra
In general, drug binding, including porphyrin, to DNA produces
hypochromism, a broadening of envelope and a red-shift in the drug
absorption region. These effects are particularly pronounced for
intercalators due to - stacking between the aromatic ring of
drugs and DNA bases. For the groove binders, a large red-shift in the
absorption band usually correlates with a drug's conformational change
or drug-drug interaction. The absorption spectra of DNA free
o-, m-, and p-TMPyP and those in the
presence of DNA, poly[d(A-T)2] and
poly[d(G-C)2] in the Soret region are depicted
in Fig. 2. When o-TMPyP is
bound to DNA and poly[d(A-T)2], the absorption maximum shifted from 414 nm to 417-418 nm and 8-9% hypochromism is
apparent (Fig. 2 A). The changes in the absorption spectrum are similar in the presence of DNA and
poly[d(A-T)2]. When bound to
poly[d(G-C)2] hypochromism reaches 34%,
although the red-shift is similar (3~4 nm). Restriction of the
rotation of the periphery pyridinium ring may prevent the
intercalation of o-TMPyP to polynucleotides and
exhibits outside binding or major groove binding (see below). The
binding of o-TMPyP may be similar in DNA and
poly[d(A-T)2], but is different from that of
poly[d(G-C)2]. The red-shift and hypochromism
are more pronounced in the m-TMPyP (Fig. 2 B)
case: red-shift is 11 nm, 16 nm, and 17 nm (from 417 nm of
polynucleotide-free m-TMPyP) and hypochromism is 31%, 51%,
and 55%, respectively, for the poly[d(A-T)2],
poly[d(G-C)2], and DNA. When p-TMPyP
(Fig. 2 C) is bound to
poly[d(A-T)2], red-shift is 8 nm (from its
polynucleotide-free absorption maximum at 422 nm). Those of
poly[d(G-C)2] and DNA are 17 nm and 22 nm.
Hypochromism is 11%, 46%, and 47%, respectively, for the
poly[d(A-T)2],
poly[d(G-C)2], and DNA. The shape of the absorption spectrum of the m- and
p-TMPyP-poly[d(G-C)2] complex is
intermediate, but resembles that complexed with DNA more closely than
with poly[d(A-T)2], which is in contrast with
o-TMPyP. However, the absorption spectra of both the
m- and p-TMPyP-DNA complexes are not a simple
combination of those complexed with
poly[d(A-T)2] and
poly[d(G-C)2].
|
CD spectra
Porphyrin induces CD bands in the Soret absorption region when
bound to polynucleotide. Recently, the CD behavior of
p-TMPyP complexed with various polynucleotides was
systematically recorded (Lee et al., 2001) when p-TMPyP
formed a complex with poly[d(A-T)2] and two
positive CD bands were apparent while, in the presence of
poly[d(G-C)2], p-TMPyP exhibits a
negative band in the Soret region. In the
p-TMPyP-poly[d(A-T)2] complex case
the band at longer wavelength was dominant (Lee et al., 2001).
The CD spectrum of o-TMPyP bound to DNA and to
poly[d(A-T)2] is essentially the same and
consists of a strong positive band at 416 nm and a positive shoulder
around 400 nm (Fig. 3 A),
indicating that the interaction of the electric transition moments of
porphyrin and those of DNA bases are similar in both complexes. The
intensity of induced CD in the same region as the
o-TMPyP-poly[d(G-C)2] complex is
smaller by a factor of ~3 and the maximum is at 413 nm. In contrast,
the CD spectra of m- (Fig. 3 B) and
p-TMPyP (Fig. 3 C) exhibit overall negative CD
band(s) when complexed with DNA and
poly[d(G-C)2]. The maximum intensity in the
negative CD band appears at 427 nm and 428 nm for the
m-TMPyP-poly[d(G-C)2] and the
m-TMPyP-DNA complex, while it appears at 441 nm and 436 nm for p-TMPyP associated with DNA and
poly[d(G-C)2]. The
m-TMPyP-poly[d(A-T)2] complex
exhibits a unique CD spectrum (Fig. 3 B): the CD spectrum consists of both a strong positive band at long wavelength (maximum at
430 nm), and a strong negative one (minimum at 420 nm), which is
typical for excitonic CD observed for the stacked porphyrins. The
intensity of excitonic CD decreases and disappears within an hour (see
below for change in CD spectrum with time). The
p-TMPyP-poly[d(A-T)2] complex
exhibits two positive bands centered at ~416 nm and ~435 nm (Fig. 3
C) as it was reported (32). The shape of the CD
spectrum is similar to that observed for the
MnTMPyP-poly[d(A-T)2] complex (Kuroda and
Tanaka, 1994). Overall, the CD spectrum of the porphyrin with freely
rotating methylpyridiniumyl rings is similar (but not identical) to
poly[d(G-C)2] and DNA (Fig. 3, B and
C), whereas o-TMPyP, in which the rotation of the
periphery pyridine ring is blocked, exhibits the identical CD spectra
for the DNA and poly[d(A-T)2] complex (Fig. 3
A).
|
Mixing ratio dependence of absorption and CD spectrum
It is well known that the binding mode of porphyrins to
various polynucleotides depends on the mixing ratio (R ratio
is the ratio of the concentration of porphyrin to DNA base). However, at a mixing ratio below 0.1, the spectroscopic properties are independent of the R ratio, indicating that the binding mode
at this low R ratio is homogeneous (Sehlstedt et al., 1994;
Lee et al., 2001). In this condition, the absorption and CD spectra are expected to obey the Beer-Lambert law. The CD spectrum of the m-TMPyP complexed with poly[d(G-C)2]
and poly[d(A-T)2] is depicted as examples in
Fig. 4, A and B).
The shapes of CD spectra of these complexes are identical at various
R ratios (R < 0.05). The absorbance and CD
intensity are proportional to the concentration of porphyrin (Fig. 4,
A and B, insets). Other complexes
exhibit similar independency of R ratio (data not shown),
indicating the binding mode at these low R ratios is
homogeneous.
|
Reduced linear dichroism
Reduced linear dichroism (LDr) is directly
related to the angle between the electric transition moment of the
polynucleotide-bound drug and the DNA helix axis. This property makes
LDr very useful to determine the binding mode of
drug relative to the polynucleotides (Nordén et al., 1992) and
has already been applied to some of the porphyrin-DNA complexes
(Geacintov et al., 1987; Sehlstedt et al., 1994; Yun et al., 1998; Lee
et al., 2001).
At a glance, the magnitude of LDr in the Soret
band is larger than that of the DNA absorption region for the
m- and p-TMPyP complexed with DNA (Fig.
5 C) and
poly[d(G-C)2] (Fig. 5 B). The larger
LDr magnitude in the drug absorption region
compared to that in the DNA absorption region is generally accepted as
an indication of drug intercalation, as it was reported for acridine
and ethidium derivatives (Kim et al., 1996; Tuite and Nordén,
1995). When porphyrin is intercalated between the DNA basepairs,
elongation and stiffening from unwinding of the DNA helix is expected,
hence an increase in the orientability of DNA in the flow, resulting in
an increase in the LDr magnitude in the DNA
absorption region compared to that of drug-free DNA. Both the
Bx and
By in-plane electric transition of
porphyrins relative to the DNA helix axis are expected to be
perpendicular with respect to the DNA helix axis that results in the
wavelength-independent (constant) LDr magnitude
in the Soret band. However, both the m- and
p-TMPyP-DNA and poly[d(G-C)2]
complexes exhibit a wavelength-dependent LDr
magnitude, and the magnitude of LDr in the DNA
absorption regions decreases. This LDr value
indicates that either the flexibilities of DNA and polynucleotides are
increased, or they are bent upon intercalation of m- and
p-TMPyP. o-TMPyP complexed with all
polynucleotides and m- and p-TMPyP complexed
poly[d(A-T)2] exhibit smaller
LDr magnitudes in the Soret band compared to
those in the DNA absorption region. Assuming an effective angle of
86° at 260 nm for the DNA bases with respect to the helix axis, the
angles and were calculated for these complexes and are
summarized in Table 1. Although there is
no definitive low limit for the angles and for intercalation,
the angle , which is lower than 70°, suggests that the possibility
of porphyrin intercalation in these complexes can conceivably be ruled
out.
|
|
CD spectra of the TMPyP-poly[d(A-T)2] complexes when the minor groove is blocked by DAPI
4',6-Diamidino-2-phenylindole (DAPI) has been well known
to bind strongly to the minor groove of
poly[d(A-T)2] and covers 4-5 AT basepairs
(Nordén et al., 1990; Eriksson et al., 1993), resulting in
the narrowing and reduced flexibility at the minor groove. This
property was utilized to probe the binding site of porphyrins in this
work, particularly for poly[d(A-T)2]. When DAPI
blocks the minor groove of poly[d(A-T)2],
spectroscopic properties of the TMPyP are expected to altered in great
extent if it is bound in the minor groove, while change would be small
if it binds at the surface or major groove of
poly[d(A-T)2].
The CD spectrum of the o-, m-, and p-TMPyP-poly[d(A-T)2] complex is compared in the presence and absence of DAPI, respectively, in Fig. 6, A-C). Here, the R ratio of DAPI was 0.12, which ensures that all available binding sites in the minor groove are blocked. A strong positive CD band at ~360 nm is apparent when DAPI is complexed with poly[d(A-T)2]. In Fig. 6, the CD spectrum of the poly[d(A-T)2]-DAPI complex was subtracted from the porphyrin-poly[d(A-T)2] complex for ease of comparison. However, it is noteworthy that a strong, positive-induced CD band of DAPI complex remained essentially the same even in the presence of porphyrin, indicating that the binding of porphyrin does not result in the release of DAPI. A positive CD band of o-TMPyP in the Soret band in the presence and absence of DAPI is similar (Fig. 6 A), suggesting that the effect of blocking the minor groove by DAPI to the binding mode of o-TMPyP is negligible. However, the CD spectrum of p-TMPyP is greatly altered by the presence of DAPI (Fig. 6 C): two positive bands at 416 nm and 435 nm blue-shift to ~398 nm and ~425 nm, and a new negative band appears at ~449 nm. For both o- and p-TMPyP, a new negative band at 360-370 nm is apparent, which may be attributed to partial release of DAPI due to porphyrin binding. The excitonic CD band of m-TMPyP remained the same even in the presence of DAPI, indicating that the binding mode of this porphyrin is unaffected. Because the CD spectrum of the m-TMPyP-poly[d(A-T)2] complex is time-dependent (see below), the CD spectrum in the presence and absence of DAPI was taken with great care to give the same time interval right after mixing.
|
Time dependence of the CD spectrum of the m-TMPyP-poly[d(A-T)2] complex
The initial CD spectrum of the
m-TMPyP-poly[d(A-T)2] complex is
compared with that after 2 h of mixing in Fig.
7. At this time, the reduction in CD
intensity reaches a plateau. The final CD spectrum in the Soret region
is symmetrical compared to that of the initial one, although the
intensity was reduced to less than a factor of 10. The reduction in CD
intensity for the
m-TMPyP-poly[d(A-T)2] does not
correspond to a simple first-order kinetic scheme (Fig. 7,
inset). Changes in the CD and absorption spectrum in
time were reported for TOPP to DNA and various polynucleotides
(Mukundan et al., 1995). However, the time scale for the change of
TOPP (24 h) and
m-TMPyP-poly[d(A-T)2] (<1 h) is
different. Although the unique dynamic aspect of m-TMPyP is
interesting, further coverage in this direction is out of the scope of
this work and will be published elsewhere.
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
|
|---|
Binding sites of o-TMPyP
The spectral properties of o-TMPyP complexed with
polynucleotides are summarized as a small red-shift and hypochromism in the absorption spectrum and positive CD band. The absorption and CD
spectra of the o-TMPyP-DNA and
o-TMPyP-poly[d(A-T)2] complexes are
almost identical, indicating that this porphyrin prefers to bind to
outside of the DNA at a low R ratio. The
LDr values for all o-TMPyP complexes
are negative. The angles of the Bx and
By in-plane electric transition of
porphyrins relative to the DNA helix axis correspond to 43-49° for
DNA, 44-47° for poly[d(A-T)2], and 60° for
the poly[d(G-C)2] complex relative to the DNA
helix axis (Table 1). From these angles it is evident that
o-TMPyP does not intercalate between the basepairs of DNA because of the steric hindrance of the periphery pyridinium ring, which
cannot rotate to form a planar molecule due to the methyl group. An
angle of ~45° for any drug that forms a complex with DNA is usually
attributed to the minor groove-binding (Nordén and Tjerneld,
1976; Nordén and Seth, 1985; Nordén et al., 1990, 1992;
Eriksson et al., 1993; Kim et al., 1996). In the case of porphyrin, it corresponds to the edgewise binding in the minor groove.
However, to exhibit this binding mode, planarity of the porphyrin
molecule is required because the minor groove is narrow. The
o-TMPyP cannot be a planner molecule. Furthermore, the
blocking of the minor groove by the minor groove-binder DAPI did not
significantly alter the CD spectrum. All these factors indicate that
the edge of the porphyrin molecule is not inserted in the narrow minor groove disregarding the angle of 45° between the transition moment and the DNA helix axis. Usually, a bisignate CD spectrum in the Soret
band is apparent for the outside stacking mode, which is not the case
for o-TMPyP. Invariance in spectral properties with respect
to the R ratio (from 0.01) is also against stacking of porphyrin. Rejecting the intercalation, minor groove edgewise-binding, and outside stacking binding modes, the only possible binding modes
left over to consider are outside random binding and face-on major
groove binding. The possibility of the outside random binding may be
ruled out because if porphyrin binds to the phosphate group by
electrostatic interaction in a random manner, then the magnitude of
LDr must be zero (LD requires orientation of the
sample), which is not the case. The small red-shift and monomeric CD
spectrum and the angle of 45° for both
Bx and
By transitions relative to the DNA
helix axis do not rule out the last possible binding mode, namely
face-on major groove binding, in which one of two lines connecting two
periphery pyridinium ring at opposite sides (45° away from both
Bx and
By transition) is parallel to the DNA
helix axis and the other perpendicular to that axis. The binding of o-TMPyP to poly[d(G-C)2] is less
clear. However, the absorption and CD spectrum are similar to those
complexed with poly[d(A-T)2] and DNA (although
hypochromism is larger and the CD is smaller), suggesting that the
binding mode in poly[d(G-C)2] is similar to other polynucleotides. However, the angle is significantly
different, being 60°, indicating that either porphyrin or DNA, or
both, are distorted, or porphyrin is rotated in the major groove. We do not have solid evidence to explain this distortion or rotation at present.
Intercalations of m- and p-TMPyP to GC-rich regions and native DNA
The optical spectroscopic properties of intercalated
p-TMPyP to poly[d(G-C)2] and DNA
have already been documented (Sehlstedt et al., 1994; Yun et al.,
1998). When it is intercalated between the basepairs of DNA or
poly[d(G-C)2], large (or larger)
LDr magnitude(s) in the Soret band compared with
the DNA absorption region, a negative CD band and large hypochromism
and red-shift in the absorption spectrum are produced. From the minimum
magnitude of LDr in the Soret band, the angle was calculated to be 76° for the p-TMPyP complex. The
larger LDr in the drug absorption region compared
to that in the DNA absorption region is often observed for
intercalators, resulting in an imaginary number for cos and cos (Sehlstedt et al., 1994; Tuite and Nordén., 1995; Kim et al.,
1996; Yun et al., 1998). This observation was attributed usually to a
kink or bend in DNA stem near the drug's binding site. These spectral
properties are confirmed for both p- and m-TMPyP
complexed with DNA and poly[d(G-C)2] in this work. At low mixing ratios (R < 0.05), spectral
properties of both the m- and
p-TMPyP-poly[d(G-C)2] complexes more
closely resemble those complexed with DNA than those with
poly[d(A-T)2], indicating that these porphyrins
prefer to intercalate between the nucleobase of native DNA. However, in
contrast with the o-TMPyP case, the absorption and CD
spectrum of the m- and
p-TMPyP-poly[d(G-C)2] complexes and
those complexed with DNA are not identical, indicating that the binding
of porphyrin to DNA cannot be explained by the simple combination of AT
and GC binding.
Outside binding of m- and p-TMPyP to AT-rich regions
The shapes of the CD spectra of both m- and
p-TMPyP complexed with poly[d(A-T)2]
are quite unique. The
p-TMPyP-poly[d(A-T)2] complex exhibits two positive CD bands in the Soret region, with intermediate hypochromism and red-shift in the absorption spectrum. Strong negative LDr, corresponding to the angle
of 61-74°, was also apparent. When the minor groove was blocked
by DAPI the CD spectrum changed, which is similar to that observed for
the MnTMPyP-poly[d(A-T)2] complex (Kuroda and
Tanaka, 1994). The two positive CD bands of MnTMPyP were attributed to
the different binding mode, namely major and minor groove binding. This
conclusion was based on the observation that intensity of the shorter
wavelength peak decreased and that of the longer wavelength increased
upon adding beneril or distamycin, which are minor groove binders. In
the p-TMPyP case, however, instead of a decrease or increase
in CD intensity, a blue-shift in the band is more pronounced with a
conceivable excitonic CD at the long wavelength edge. Although the
change in CD spectra in the presence of DAPI cannot be fully
understood, the facts that 1) both CD and absorption spectra are
directly proportional to the concentration of added p-TMPyP
(which supports the homogeneous binding mode); 2) the shape of CD
spectrum changes rather than disappears (which is against the edge of
porphyrin inserting fully into the minor groove because in this case
replacement is expected); 3) the long side chain, such as the
n-octyl group on the periphery of pyridine did not alter
spectral properties (Sehlstedt et al., 1994; Yun et al., 1998) (if the
porphyrin is inserted edgewise from the minor groove, the side chain is
expected to prevent binding, resulting in the alteration in spectral
properties); and finally 4) the angle (61-74°) being far greater
than 45° (which is expected from the porphyrin edgewise insertion to
minor groove) suggest that p-TMPyP conceivably locates near
the minor groove, but is not inserted from there. It seems to be bound
at the outside of poly[d(A-T)2] near the minor
groove. However, it is equally possible that p-TMPyP binds
at the major groove and changes its CD spectrum by the conformation
change of the polynucleotide due to DAPI binding in the minor groove.
From the simple CD result that we observed here, the possibility of the
coexistence of major and minor groove binding porphyrins (Kuroda and
Tanaka, 1994) also cannot be ruled out.
An excitonic CD is apparent for the
m-TMPyP-poly[d(A-T)2] complex,
indicating that m-TMPyP is stacked outside of
poly[d(A-T)2]. The presence of DAPI did not
affect the CD spectrum. Therefore, it is clear that m-TMPyP
stacks at the location where the change in minor groove has no effect.
Among the known binding modes, stacking at the major groove is most
conceivable. The angle is calculated to be 53-60°, indicating
that the plane of the porphyrin molecule is not parallel to DNA
basepairs (not perpendicular to the helix axis). Stacked
m-TMPyP seems to be tilted in the major groove. A change in
the position of the methyl group at the pyridine ring from a
para- to meta-position results in a very
different binding mode, particularly when complexed with
poly[d(A-T)2]. This variation in binding mode
cannot be explained by the ability of rotation of the pyridine moiety
because the pyridine ring can rotate freely in both the m-
and p-TMPyP case.
| |
CONCLUSIONS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
|
|---|
A systematic optical spectroscopic study for o-, m-, and p-TMPyP complexed with poly[d(A-T)2], poly[d(G-C)2], and DNA led us to the following conclusions: 1) preventing the rotation of the periphery methylpyridine ring alters the binding mode, as expected; o-TMPyP locates at the major groove; 2) both m- and p-TMPyP intercalate to poly[d(G-C)2] and DNA at a low R ratio; and 3) m-TMPyP is stacked in the major groove of poly[d(A-T)2], while various binding modes can be suggested for p-TMPyP.
| |
ACKNOWLEDGMENTS |
|---|
This work was supported by Korea Research Foundation Grant KRF 99-005-D00043.
| |
FOOTNOTES |
|---|
Address reprint requests to Seog K. Kim, 214 Dae-dong, Kyoungsan City, Kyoung-buk 712-749, Republic of Korea. Tel.: +82-53-810-2362; Fax: +82-53-815-5412; E-mail: seogkim{at}yu.ac.kr.
Submitted November 1, 2001, and accepted for publication March 7, 2002.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
|
|---|
Biophys J, July 2002, p. 371-381, Vol. 83, No. 1
© 2002 by the Biophysical Society 0006-3495/02/07/371/11 $2.00
This article has been cited by other articles:
 |
|
 |
|
  J.-O. Kim, Y.-A. Lee, B. H. Yun, S. W. Han, S. T. Kwag, and S. K. Kim Binding of meso-Tetrakis(N-methylpyridinium-4-yl)porphyrin to AT Oligomers: Effect of Chain Length and the Location of the Porphyrin Stacking Biophys. J., February 1, 2004; 86(2): 1012 - 1017. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-A. Lee, S. Lee, H. M. Lee, C.-S. Lee, and S. K. Kim Anomalous Excitonic CD of Meso-Tetrakis(3-N-Methylpyridiniumyl)porphyrin Bound to Poly[d(A-T)2] J. Biochem., March 1, 2003; 133(3): 343 - 349. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
(I(t) 
If)/(Ii 