Self-aggregation inhibits the photonuclease activity of porphyrins.

Article abstract of DOI:10.1248/cpb.51.899

A series of tentacle porphyrins having four aminoalkyl groups at the periphery was synthesized, and the DNA binding properties were investigated by absorption and circular dichroism (CD) spectroscopic methods. The aminopropyl chain was found to facilitate binding, and bisignate induced CD spectra revealed that the porphyrins are self-stacked on the DNA surface. The photonuclease activity of the tentacle porphyrins was also studied, and the aminopropylporphyrin showed the highest activity. The activity increased in proportion to the porphyrin load, but higher loads resulted in the decrease of activity. This inhibitory step corresponded to aggregation of the porphyrin. Thus, the aggregation was suggested to shield the inner porphyrin from the solvent, the production of active oxygen species being suppressed.

Full text of DOI:10.1248/cpb.51.899

August 2003
Chem. Pharm. Bull. 51(8) 899—903 (2003)
899
Self-Aggregation Inhibits the Photonuclease Activity of Porphyrins
Katsumasa AOKI, Yoshinobu ISHIKAWA, Miyuki OYAMA, Yoshikazu TOMISUGI, and Tadayuki UNO*
Faculty of Pharmaceutical Sciences, Kumamoto University; Oehonmachi, Kumamoto 862–0973, Japan.
Received December 16, 2002; accepted June 13, 2003; published online June 13, 2003
A series of tentacle porphyrins having four aminoalkyl groups at the periphery was synthesized, and the
DNA binding properties were investigated by absorption and circular dichroism (CD) spectroscopic methods.
The aminopropyl chain was found to facilitate binding, and bisignate induced CD spectra revealed that the por-
phyrins are self-stacked on the DNA surface. The photonuclease activity of the tentacle porphyrins was also stud-
ied, and the aminopropylporphyrin showed the highest activity. The activity increased in proportion to the por-
phyrin load, but higher loads resulted in the decrease of activity. This inhibitory step corresponded to aggrega-
tion of the porphyrin. Thus, the aggregation was suggested to shield the inner porphyrin from the solvent, the
production of active oxygen species being suppressed.
Key words photonuclease; porphyrin; aggregation; DNA; active oxygen
The development of efficient and DNA-specific photosen- Experimental
1
Instrumentation The H-NMR spectra were recorded on a JEOL GX-
400 or JNM-A-500 spectrometer. Matrix-assisted laser desorption ionization
time-of-flight (MALDI-TOF) mass spectra were measured on a Bruker RE-
FLEX. The UV–visible absorption measurement was performed on a Beck-
sitizers attracts much attention in the growing field of cancer
1
,2)
photodynamic therapy (PDT).
In these applications, the
photosensitizers bind to DNA specifically, generate active
oxygen species upon illumination, and subsequently cleave man DU640 spectrophotometer. The CD spectra were recorded on a JASCO
J-720 spectropolarimeter. Elemental analyses were performed at the Center
for Instrumental Analysis, Kumamoto University.
Materials and Methods Tetrakis(4-aminophenyl)porphine (TAPP) and
reagents for the synthesis of tentacle porphyrins were purchased from Tokyo
DNA in malignant cells. The DNA damage will disturb the
genetic events and thus the proliferation of tumor cells is
suppressed. Therefore, both tight binding to DNA and effi-
cient generation of active oxygen species are prerequisites
for an effective PDT reagent.
Chemical Industry, Wako Pure Chemicals, and Dojin. Calf thymus (ct) DNA
was purchased from Sigma, and pUC18 plasmid DNA from Stratagene. The
DNA solution was quantitated spectrophotometrically using e260ϭ13.2
Porphyrins can be effective photosensitizers, and anionic
Photofrin, a purified fraction of an oligomeric hematopor-
Ϫ1
Ϫ1
mM cm , and the concentrations of DNA were expressed in terms of base
pairs.
3
,4)
phyrin derivative (HPD), is now clinically used in cancer
Synthesis of Porphyrins Twenty millimoles of amino acid (4-amino-n-
5
)
PDT. From the view point of DNA targeting, cationic por- butyric acid, 6-aminohexanoic acid, and w-aminocaprylic acid) was dis-
phyrins have been studied extensively, because of their po- solved in water (11 ml) and triethylamine (30 mmol), and dioxane solution
6
—8)
(11 ml) of 2-BOC-thio-4,6-dimethylpyrimidine (22 mmol, BOCϭbutyloxy-
carbonyl) was poured. After stirring for 26 h at room temperature, 30 ml of
water was poured and unreacted materials were removed with ethyl acetate.
The water layer was brought to pH 2 with HCl on ice, and BOC-protected
tential for electrostatic binding to DNA.
Three major
modes have been proposed for porphyrin binding: intercala-
tion, outside binding in the groove, and outside binding with
9
)
self-stacking along the DNA surface. These binding modes amino acid was extracted with ethyl acetate. The extract was washed with
can be revealed by various techniques such as UV–visible ab- ice-cold 5% HCl and then with saturated NaCl, and dried over anhydrous
9
—11)
9—12)
13,14)
sodium sulfate. The solvent was removed and product was obtained as oil
(BOC-amino-n-butyric acid and BOC-aminohexanoic acid) or as powder
sorption,
NMR,
circular dichroism (CD),
fluorescence,
spectroscopies.
We have recently prepared zinc complexes of cationic por-
1
5,16)
17,18)
and resonance Raman
(BOC-aminocaprylic acid). The yield commonly amounted to more than 80%.
The BOC-amino acid (3 mmol) and 1-ethyl-3-(3-dimethylaminopropyl)-
phyrins, and shown that metal insertion into the porphyrin carbodiimide (3 mmol) were dissolved in dimethylformamide (DMF) and
stirred for 1 h. TAPP (0.15 mmol) and 1-hydroxybenzotriazole (3 mmol)
core drastically enhances the photonuclease activity of the
1
9)
were added and stirred overnight. Chloroform (200 ml) was poured into the
mixture, washed three times with water, and the organic layer was dried over
anhydrous sodium sulfate. In the case of BOC-aminocaprylic acid, unreacted
porphyrin. The evidence suggests strongly that the active
1
9)
species is singlet oxygen, since the activity was propor-
tional to the singlet oxygen productivity of the porphyrins,
and scavengers of this molecule greatly reduced the activity.
As well as the enhanced generation of the active oxygen,
refined control of the mode of binding is required in order to
maximally enhance the photonuclease activity of porphyrins.
The relationship between the photonuclease activity and the
binding modes is not yet fully understood. In the course of
materials were separated by silica gel chromatography (methanol/CHCl ϭ
3
50/50). The solvent was removed and trifluoroacetic acid (TFA) was poured
onto the residue. The amino group was deprotected stirring for 1 h on ice.
H N ( )n
2
H
N
H
N
( )n
NH2
O
O
1
9—21)
N
N
intensive studies on porphyrin photonuclease activities,
we have found that highly aggregative porphyrins tend to
have lower activities. We therefore synthesized a series of
tentacle porphyrins having four aminoalkyl chains (Fig. 1).
By correlating the binding properties with photonuclease ac-
tivities, we have shown that self-aggregation is an inactive
mode of binding. This result provides a clue to the future de-
velopment of porphyrin-based photosensitizers applicable to
PDT.
NH
HN
O
O
N
H
N
H
H2N ( )n
( ) NH2
n
C3, n = 3
C4, n = 4
C5, n = 5
Fig. 1. Tentacle Porphyrins Synthesized in This Study
To whom correspondence should be addressed. e-mail: unot@gpo.kumamoto-u.ac.jp © 2003 Pharmaceutical Society of Japan
∗

9
00
Vol. 51, No. 8
The product was precipitated by pouring diethylether, filtered, and dissolved
in minimal amount of methanol. When the solution was green, triethylamine
was added until the color turns to purple. Addition of diethylether to the so-
lution and slow diffusion gave a purple powder, which was collected and
dried.
0
.9
C3 + ctDNA
1
2
3
4
5
6
0.6
1, 0
µM
1
2, 0.83 µM
3, 1.48 µM
4, 2.13 µM
5, 3.10 µM
6, 5.46 µM
Tetrakis[4-(4-aminobutanamido)phenyl]-porphine (C3): Yield: 28%. H-
NMR (CD OD) d: 2.15 (8H, m), 2.75 (8H, t), 3.15 (8H, t), 8.10 (16H, dd),
3
0
.3
.0
8
.89 (8H, br). UV lmax (DMF) nm (e): 423 (478000), 519 (19600), 556
ϩ
(
16400), 594 (6590), 652 (9120). MALDI-TOF-MS m/z: 1015.3 (M , Calcd
for C H N O : 1015.2). Anal. Calcd for C H N O ·4CF COOH: C,
60
62 12
4
60 62 12
4
3
0
5
5.50; H, 4.53; N, 11.43. Found: C, 55.50; H, 4.81; N, 11.70.
Tetrakis[4-(4-aminohexanamido)phenyl]-porphine (C5): Yield: 64%. H-
NMR (CD OD) d: 1.61 (8H, m), 1.80 (8H, m), 1.89 (8H, m), 2.62 (8H, t),
1
3
00
400
500
600
700
3
3
.02 (8H, t), 8.09 (16H, dd), 8.91 (8H, br). UV lmax (DMF) nm (e): 423
Wavelength/nm
(
408000), 520 (18200), 558 (15800), 595 (6720), 652 (8100). MALDI-TOF-
ϩ
Fig. 2. Absorption Spectral Change of C3 on the Addition of CtDNA
MS m/z: 1127.2 (M , Calcd for C H N O : 1127.4). Anal. Calcd for
6
8
78 12
4
The porphyrin concentration was 1.72 mM. The spectra were recorded in 10 mM
sodium phosphate (pH 7.0) and 30% DMF. Concentrations of DNA were as follows: 1,
mM; 2, 0.83 mM; 3, 1.48 mM; 4, 2.13 mM; 5, 3.10 mM; 6, 5.46 mM.
C H N O ·4CF COOH: C, 57.65; H, 5.22; N, 10.61. Found: C, 57.50; H,
68
78 12
4
3
5
.02; N, 10.28.
Tetrakis[4-(4-aminooctanamido)phenyl]-porphine (C7): Yield: 32%. H-
NMR (CD OD) d: 1.53 (24H, m), 1.73 (8H, m), 1.88 (8H, m), 2.62 (8H, t),
0
1
3
2
.97 (8H, t), 8.40 (16H, dd), 8.81 (8H, s). UV lmax (DMF) nm (e): 423
1
.0
(
493000), 519 (18500), 556 (15600), 594 (5890), 651 (8090). MALDI-TOF-
ϩ
MS m/z: 1239.9 (M , Calcd for C H N O : 1239.7). Anal. Calcd for
7
6
94 12
4
0.8
0.6
C H N O ·4CF COOH: C, 59.50; H, 5.83; N, 9.91. Found: C, 59.13; H,
76
94 12
4
3
6
.29; N, 9.94.
Spectral Measurements Aliquots of ctDNA solution were added to the
porphyrin solution (ca. 2 mM), and absorption spectra were measured at
5 °C in 10 mM sodium phosphate and 30% DMF (pH 7.0). Since the por-
0
.4
0.2
.0
C3
C5
C7
2
phyrin solubility was low in water, addition of DMF was necessary. The
amount of 30% DMF does not affect the secondary structure of DNA, as
judged by CD spectroscopy. The binding constant (K) and the number of
binding sites per base pair (n) were estimated from the spectral changes, fol-
0
0
2
4
6
8
22—24)
[Base Pair]/µM
lowing Scatchard analysis.
The induced CD spectra of the porphyrins
(
ca. 2 mM) were measured at selected concentrations of DNA, and twenty in-
Fig. 3. Normalized Absorbance Change of Tentacle Porphyrins at 421 nm
by the Addition of CtDNA
dependent spectra were averaged. Conversely, aliquots of the porphyrin solu-
tion were added to a DNA solution (ca. 30 mM), and four independent CD
spectra in the UV range were averaged. The CD spectra were baseline cor-
rected and smoothed.
Closed squares, C3; open circles, C5; closed circles, C7. Theoretical solid lines were
drawn with the K and nAB values in Table 2.
Photonuclease Assay Photoillumination was performed at 25 °C using
a Hitachi 650-60 fluorescence spectrophotometer equipped with a 150 W Xe Table 1. Spectral Comparison among Tentacle Porphyrins Bound to
19—21)
lamp.
A sample containing supercoiled pUC18 plasmid DNA and a CtDNA
porphyrin was illuminated in 10 mM sodium phosphate (pH 7.0) and 30%
DMF at 420 nm. After illumination, DNA was analyzed by 0.8% agarose gel
electrophoresis. The DNA was stained by ethidium bromide and the DNA
bands were visualized by fluorescence over a UV lamp. The densitometric
a)
b)
c)
Porphyrin
Dl/nm
H/%
Induced CD/nm
C3
C5
C7
Ϫ12
Ϫ16
Ϫ11
68
72
73
393 (Ϫ17.5)
393 (Ϫ11.2)
389 (Ϫ1.6)
405 (ϩ24.7)
405 (ϩ18.1)
404 (ϩ4.0)
2
5)
data of the bands were obtained with NIH Image software. The staining
intensity of relaxed plasmid DNA was found to be 1.5 times that of super-
26)
coiled DNA. The band density was corrected based on this difference.
a) Bathochromic shift. b) The hypochromicity (H/%) was determined by the
equation Hϭ(e Ϫe )/e ϫ100, where e and e represent the molar absorptivities of free
and bound porphyrins, respectively, which were determined at the respective Soret
f
b
f
f
b
Results
maxima. c) Induced CD peaks. The values in parentheses are molar ellipticities
4
2
Ϫ1
Absorption Spectra In order to evaluate the DNA bind- ([q]/10 deg cm dmol ).
ing properties of the porphyrins, we first titrated the por-
phyrin with ctDNA and recorded the absorption spectral Table 2. Comparison of Binding Parameters among Tentacle Porphyrins
change. A typical set of spectra for C3 is shown in Fig. 2. It
Ϫ1
nABa)
nCDb)
Porphyrin
C3
K/mM
96
is clear that the strong Soret peak at 421 nm decreases
greatly upon addition of DNA, indicating that the porphyrin
p electrons are perturbed considerably by the DNA binding.
Since one set of isosbestic points are clearly observed, the
binding appears to take place in a single step. The main Soret
peak is blue-shifted with a shoulder at about 430 nm in the
presence of large amount of DNA. The spectral properties of
0.43
0.41
0
.17
0.67
.15
0.84
C5
24
0.64
0.84
0
C7
25
a) The nAB values obtained from the absorbance changes. b) The nCD values ob-
C3, C5, and C7 are summarized in Table 1. The large tained from the elliptic changes.
hypochromicity (H), as well as the large blue shift (Dl) of
the Soret band is distinct from any other porphyrins thus far fully (Table 2). Theoretical curves using these K and nAB val-
2
4)
examined.
ues were found to fit nicely to the absorbance changes actu-
Normalized absorbance changes are plotted against the ally observed (Fig. 3). It is clear that the C3 porphyrin has
concentration of base pairs in Fig. 3. Since the spectral very large K value, while nAB value increases as the alkyl
change takes place in a single step, the binding constant (K) chain becomes long. The large K value suggests that the
and the number of binding sites (n) can be estimated success- chain length of C3 is appropriate for the binding to DNA.

August 2003
901
1
0
5
0
5
10
5
(a)
1
2
3
4
,
,
,
,
0
µM
6.4 µM
7.3 µM
8.2 µM
5, 10.0 µM
, 11.8 µM
6
0
6
5
4
3
2
1
-5
10
C3 + ctDNA
ctDNA + C3
-
-
2
50
270
290
310
330
350
5
0
(b)
Wavelength/nm
Fig. 4. CD Spectral Changes of CtDNA upon Addition of C3
DNA concentration was 27.9 mM. The spectra were recorded in 10 mM sodium phos-
phate (pH 7.0) and 30% DMF. The concentrations of C3: 1, 0 mM; 2, 6.4 mM; 3, 7.3 mM;
C5 + ctDNA
-5
4
, 8.2 mM; 5, 10.0 mM; 6, 11.8 mM.
5
0
(
c)
C7 + ctDNA
1.0
0.5
0.0
nCD = 0.41
C3
350
400
450
500
nCD = 0.17
Wavelength/nm
Fig. 6. Induced CD Spectra of Tentacle Porphyrins by the Addition of
CtDNA
1.0
0.5
0.0
nCD = 0.67
The spectra were recorded in 10 mM sodium phosphate (pH 7.0) and 30% DMF. (a)
C3, 2.94 mM; DNA, 7.42 mM (rϭ0.40, solid line) and 15.0 mM (rϭ0.20, dashed line). (b)
C5, 2.00 mM; DNA, 3.00 mM (rϭ0.67, solid line) and 15.7 mM (rϭ0.13, dashed line). (c)
C7, 3.17 mM; DNA, 3.85 mM (rϭ0.82).
C5
nCD = 0.15
1.0
0.5
0.0
from the theoretical curve (Fig. 5), and the CD changes de-
scribe a second binding step which is not detected by the ab-
sorbance changes (Fig. 3). When two binding steps are dis-
tinct and can be treated separately, two n values are obtained
independently from the break points (shown by arrows) of
C7
0
5
10
15
20
25
[
Porphyrin]/µM
Fig. 5. Normalized Changes in Ellipticity of CtDNA at 260 nm upon Ad- the dashed lines in Fig. 5. The n values (nCD) thus estimated
dition of Tentacle Porphyrins
are summarized in Table 2.
In the case of C3, the larger n value (n ϭ0.41) was quite
similar to that obtained from the absorbance change (n ϭ
0.43). Therefore, the nϭ0.17 step for C3 is sensitive only to
The base pair concentrations were 27.9 mM for C3, 25.5 mM for C5, and 23.4 mM for
C7. Closed squares, C3; open circles, C5; closed circles, C7. The solid lines were
drawn with the K and nAB values in Table 2. The dashed lines were drawn with the lin-
ear part of the ellipticity changes. The break points were marked with arrows.
CD
AB
the conformational change in DNA. This situation is also ap-
2
4)
CD Spectra As we pointed out previously, absorption plicable to C5, where the n ϭ0.15 step is detected only by
CD
spectroscopy occasionally fails to detect a binding step if the CD.
associated spectral change is small. As an additional ap-
Induced CD Spectra To clarify the binding modes of
9
,27)
proach we employed CD spectroscopy which is sensitive to the porphyrin, induced CD spectra
were measured at
conformational change in DNA. The CD spectral changes varying concentrations of the DNA. As seen in Fig. 6, bisig-
upon C3 binding are shown in Fig. 4. A strong signal arising nate CD peaks were commonly observed for the porphyrins,
from the solvent, DMF, overlaps in the wavelength range suggesting that the porphyrin binds externally with self-
below 250 nm, thus the spectra were measured between stacking on the DNA surface. In the case of C3 and C5, two-
2
50—350 nm. In this range, the spectral change was small, step binding is suggested by the ellipticity change (Fig. 5),
with no obvious shift being observed. Thus, any conforma- but the peaks diminish only slightly at large DNA loads cor-
tional change in the DNA is likely to be small, i.e., the bound responding to the smaller nCD steps, indicating that the self-
porphyrins interact with DNA without greatly perturbing its stacked porphyrins still dominate in these steps.
structure.
The most obvious difference among the porphyrins is the
The changes in ellipticity were normalized and plotted signal intensity of the induced CD peaks. The positive and
against the concentration of porphyrins (Fig. 5). The solid negative peaks in the C3 spectra are almost twice as strong
lines illustrate theoretical plots obtained using the K and nAB as those of C5, and more than ten-times stronger than those
1
7,24)
values obtained from the absorbance data (Table 2). The the- of C7. As we pointed out previously,
the intensity of the
oretical curve for C7 is in good agreement with the observed bisignate induced signals is proportional to the n value for
changes in ellipticity. This indicates that binding proceeds in cationic porphyrins. In the present case, however, the trend is
a single step, and that the absorbance and CD changes in- reversed, and an unusual mode of binding appears to be tak-
form on the same binding process. In the case of C3 and C5, ing place.
however, the observed ellipticity plots deviate considerably
DNA Cleavage The photonuclease activity for C3, C5,

9
02
Vol. 51, No. 8
Through the absorption and CD measurements, the spectral
characteristics of these DNA-bound tentacle porphyrins were
found to be distinct from any other cationic porphyrins; 1)
They exhibit very large hypochromicity (Fig. 2), 2) they per-
turb DNA structure minimally (Fig. 4), and 3) they produce
inverted bisignate induced CD pattern (Fig. 6). Thus, we con-
clude these porphyrins engage in a novel mode of DNA bind-
ing as discussed below.
(a)
Form II
Form I
0
2
4
6
8
10 12 14
Illumination Time/min
1
00
(b)
8
0
0
0
0
C3
C5
C7
r = 0.18
The large hypochromicity indicates that the p electrons of
the DNA-bound porphyrins are highly perturbed. In accord
with this, bisignate CD signals were induced, indicating that
the porphyrins are self-stacked; a negative signal is expected
when porphyrins intercalate between the DNA base pairs,
6
4
2
0
9
,27)
and groove-bound porphyrins show a positive signal.
The
0
5
10
15
20
bisignate signals were, however, inverted when compared
Illumination Time/min
1
7,24)
with “normal” cationic porphyrins.
Positive/negative
(
ϩ/Ϫ) peaks normally appear at shorter/longer wavelengths,
9
(c)
C3
C5
C7
and our tentacle porphyrins showed Ϫ/ϩ pattern. Self-
stacked porphyrins have been assumed to be aligned along
the DNA surface, with their peripheral cationic groups in the
DNA groove (edge-on). Thus, the greater the number of por-
phyrins bound (larger n value), the stronger the induced CD
6
3
0
1
7,24)
signals.
In our case, however, the signal intensity de-
creased in the order C3ϾC5ϾC7 as the nAB value increases.
Since the hypochromicities among the tentacle porphyrins
are nearly the same (Table 1), the degree of stacking is ex-
pected to be similar. Thus, DNA-bound porphyrins are sug-
gested to be highly disordered (aggregation), especially for
0
0.2
0.4
0.6
r
Fig. 7. Photonuclease Activity of Tentacle Porphyrins
(
a) Agarose gel electrophoresis of plasmid DNA treated with 5.3 mM C3. Cleavage
conditions: 30.0 mM pUC18 plasmid DNA in 10 mM sodium phosphate (pH 7.0) and C7.
0% DMF, illumination at 420 nm at 25 °C. Form I and form II denote the supercoiled
and nicked-circular forms of plasmid DNA, respectively. (b) Time course of the devel-
3
The tentacle porphyrins are achiral, but CD signals are in-
opment of form II. Molar fraction of the form II is plotted against the illumination time. duced when they bind to DNA which of course is chiral.
Porphyrin concentration was 5.3 mM (rϭ0.18). Closed squares, C3; open circles, C5;
Thus, the orientation of the bound porphyrin relative to the
closed circles, C7. (c) Cleaving rate of tentacle porphyrins plotted against r. Closed
2
8)
squares, C3; open circles, C5; closed circles, C7.
DNA axis is a major determinant of the CD signature. Al-
though speculative, we propose here that the tentacle por-
and C7 was examined using supercoiled pUC18 plasmid phyrins stacked with their planar surfaces facing the DNA
DNA. A mixture of porphyrin and plasmid (30 mM) was illu- duplex. This binding mode well accounts for the spectral
minated in air at 420 nm. After illumination, conversion of characteristics of our tentacle porphyrins.
supercoiled DNA (form I) to nicked-circular DNA (form II)
When cationic porphyrins are stacked edge-on, the posi-
was investigated by agarose gel electrophoresis. On genera- tively charged peripheral groups protrude into the DNA
tion of a single-strand break, form I is converted to form II, groove, and interact directly with base pairs. Accordingly,
which migrates more slowly on the gel.
CD signals in the UV range, where DNA bases are the main
As seen in Fig. 7a, form II DNA is clearly produced fol- contributors to the spectrum, are usually affected by por-
lowing prolonged illumination. The molar fraction of form II phyrin binding. In our tentacle porphyrins, however, the
was quantitated by densitometry, and plotted against illumi- structural change in the DNA is minimal, suggesting that the
nation time in Fig. 7b. It is clear that the molar fraction of the amino groups do not interact with base pairs. The amino
form II increases linearly with time and thus with the light group is expected to be protonated at neutral pH, and hence
dose. Highly efficient cleavage by C3 is readily apparent, able to interact favorably with the negatively charged phos-
while C5 and C7 are rather poor photonucleases. The con- phate groups. The longer hydrocarbon stem of C7 is more
version rates were determined from the slopes of the lines, flexible, and this allows a disordered mode of binding to
and these are plotted against the porphyrin load in Fig. 7c. DNA. The hydrophobic nature of the stem should facilitate
The activity of C5 and C7 remains very low even at rϭ0.6 (r aggregation (or pile-up) with large n values. The piled por-
is [porphyrin]/[DNA] ratio). On the other hand, the rate in- phyrins may be located away from the DNA axis, and con-
creases initially to the C3 load, indicating that C3 catalyzed tribute only weakly to the induced CD.
the DNA cleavage. When larger amount of C3 is loaded,
Photonuclease Activity C3 was an efficient photonucle-
however, the rate decreased and an inhibitory effect of bound ase, and its activity depended linearly on the light dose and
C3 is clear.
C3 concentration (Fig. 7). The high efficiency may be attrib-
uted in part to the high affinity of C3 for DNA (Table 2). The
short hydrocarbon stem of C3 is suitable for DNA binding.
Discussion
DNA Binding We have synthesized a series of novel However, C5 and C7 have moderate affinities, and their low
porphyrins having aminoalkyl tentacles, and analyzed their activities have to be explained by other mechanism(s).
DNA binding properties and photonuclease activities.
The nuclease activity of C3 approaches a maximum at

August 2003
903
2
3
)
)
Ali H., van Lier J. E., Chem. Rev., 99, 2379—2450 (1999).
Fiel R. J., Datta-Gupta N., Mark E. H., Howard J. C., Cancer Res., 41,
543—3545 (1981).
Praseuth D., Gaudemer A., Verlhac J.-B., Kraljic I., Sissoeff I., Guille
E., Photochem. Photobiol., 44, 717—724 (1986).
Dougherty T. J., Photochem. Photobiol., 58, 895—900 (1993).
Fiel R. J., J. Biomol. Struc. Dyn., 6, 1259—1274 (1989).
Pasternack R. F., Gibbs E. J., Met. Ions Biol. Syst., 33, 367—397
about rϭ0.2, and then decreases to zero at about rϭ0.4.
These r values are similar to the nCD value (Table 2), indicat-
ing that DNA binding and photonuclease activity are closely
correlated. The nϭ0.17 step was detected only by CD spec-
troscopy. Since the induced CD spectra of C3 at rϭ0.2 and
3
4
)
5
6
7
)
)
)
0
.4 are similar, the binding mode for these steps is the same
(
Fig. 6). Thus, the electronic structure should be quite simi-
(
1996).
lar, and the nϭ0.17 step is not detected by absorption spec-
troscopy.
The face-on binding mode satisfactorily accounts for the
inactivation process for C3. In the rϭ0—0.2 range, C3 lands
softly on the target DNA with self-stacking. At this stage, the
8
)
Marzilli L. G., New. J. Chem., 14, 409—420 (1990).
9) Pasternack R. F., Gibbs E. J., Villafranca J. J., Biochemistry, 22,
406—2414 (1983).
0) Fiel R. J., Howard J. C., Mark E. H., Datta Gupta N., Nucleic Acids
Res., 6, 3093—3118 (1979).
2
1
1
1) Carvlin M. J., Fiel R. J., Nucleic Acids Res., 11, 6121—6139 (1983).
outer C3 is situated quite close to the DNA, it absorbs light, 12) Carvlin M. J., Mark E. H., Fiel R. J., Nucleic Acids Res., 11, 6141—
and generates active oxygen species. In the subsequent
rϭ0.2—0.4 range, C3 piles up and aggregates, further se-
questering the inner C3 species, and ultimately the nuclease
activity is fully inhibited. The distant location of additionally
6154 (1983).
1
1
3) Sari M. A., Battioni J. P., Dupre D., Mansuy D., Le Pecq J. B., Bio-
chemistry, 29, 4205—4215 (1990).
4) Sehlstedt U., Kim S. K., Carter P., Goodisman J., Vollano J. F., Norden
B., Dabrowiak J. C., Biochemistry, 33, 417—426 (1994).
piled C3 in this range is supported by induced CD spectra 15) Marzilli L. G., Banville D. L., Zon G., Wilson W. D., J. Am. Chem.
(
Fig. 6). The signal intensity at rϭ0.2 increases only slightly
Soc., 108, 4188—4192 (1986).
1
1
6) Guliaev A. B., Leontis N. B., Biochemistry, 38, 15425—15437 (1999).
7) Uno T., Aoki K., Shikimi T., Hiranuma Y., Tomisugi Y., Ishikawa Y.,
Biochemistry, 41, 13059—13066 (2002).
at rϭ0.4, although the C3 load is almost doubled. The more
remote location of the additional C3 species from the DNA
axis is expected to induce CD only weakly.
1
8) Bütje K., Nakamoto K., Inorg. Chim. Acta, 167, 97—108 (1990).
The tentacle porphyrins were found to have unique spec- 19) Ishikawa Y., Yamakawa N., Uno T., Bioorg. Med. Chem., 10, 1953—
1
960 (2002).
0) Yamakawa N., Ishikawa Y., Uno T., Chem. Pharm. Bull., 49, 1531—
540 (2001).
1) Ishikawa Y., Yamashita A., Uno T., Chem. Pharm. Bull., 49, 287—293
2001).
tral characteristics, which are well explained by putative
face-on stacking. The C3 linker is effective in enhancing the
binding affinity. The extended hydrocarbon stem of the tenta-
cle, however, promoted piled-up aggregation, which is in-
2
2
1
(
hibitory to the photonuclease activity. Thus, non-aggregating 22) Scatchard G., Ann. N.Y. Acad. Sci., 51, 660—672 (1949).
2
2
3) Correia J. J., Chaires J. B., Methods Enzymol., 240, 593—614 (1994).
4) Uno T., Hamasaki K., Tanigawa M., Shimabayashi S., Inorg. Chem.,
tentacle porphyrins are likely to be more efficient photonu-
cleases and therefore show greater promise in applications in
cancer PDT.
3
6, 1676—1683 (1997).
2
2
5) http://rsb.info.nih.gov/nih-image/
6) Nielsen P. E., Jeppesen C., Egholm M., Buchardt O., Biochemistry, 27,
Acknowledgments The authors thank Dr. Anthony J. Wilkinson, Uni-
versity of York, U.K., for valuable discussion. This work is supported by a
grant for Science Research (13557199 to T. U.) from Japan Society for Pro-
motion of Science.
6
338—6343 (1988).
2
2
7) Pasternack R. F., Garrity P., Ehrlich B., Davis C. B., Gibbs E. J., Orloff
G., Giartosio A., Turano C., Nucleic Acids Res., 14, 5919—5931
(
1986).
8) Kubista M., Åkerman B., Nordén B., J. Phys. Chem., 92, 2352—2356
1988).
References and Notes
(
1
)
Amitage B., Chem. Rev., 98, 1171—1200 (1998).