Water-proton relaxivities of DNA oligomers carrying TEMPO radicals

Article abstract of DOI:10.1002/mrc.2310

5-Uridine derivative carrying a TEMPO radical (UST) was prepared and its single strand (ssUST) and a double strand (dsUST) with its complementary strand were obtained. Similarly, single strands carrying two and five radicals (ssUST2 and ssUST5, respectively) and the corresponding doubles trands (dsUST2 and dsUST5) were prepared. Their electron paramagnetic resonance (EPR) spectra showed typical anisotropic broadening in the high field line. The rotational correlation times, τ_R, estimated by analyzing the EPR spectra are 1.1 × 10^-10, 5.9 × 10^-10, and 14 × 10^-10 s for UST, ssUSTm, and dsUSTm, respectively. The water-proton relaxivities, r₁ and r₂, at 25 MHz, 0.59 T, and 25°C, also increased in the same order and the r₁ values were 0.26, 0.41, and 0.56 mM^-1 s^-1 for UST, ssUSTm, and dsUSTm, respectively. The r₁ values of 1.00 and 2.06 mM^-1 s ^-1 for dsUST2 and dsUST5, respectively, were obtained. Copyright

Full text of DOI:10.1002/mrc.2310

Note
Received: 19 April 2008
Revised: 10 June 2008
Accepted: 25 July 2008
Published online in Wiley Interscience: 19 September 2008
(www.interscience.com) DOI 10.1002/mrc.2310
Water-proton relaxivities of DNA oligomers
carrying TEMPO radicals
Yuichiro Sato,a Hiroyuki Hayashi,a Manami Okazaki,a Mariko Aso,a
Satoru Karasawa,a Shoji Ueki,b Hiroshi Suemunea and Noboru Kogaa∗
5-Uridine derivative carrying a TEMPO radical (UST) was prepared and its single strand (ssUST) and a double strand (dsUST)
with its complementary strand were obtained. Similarly, single strands carrying two and five radicals (ssUST2 and ssUST5,
respectively)andthecorrespondingdoublestrands(dsUST2anddsUST5)wereprepared.Theirelectronparamagneticresonance
(EPR) spectra showed typical anisotropic broadening in the high field line. The rotational correlation times, τR, estimated by
analyzing the EPR spectra are 1.1 × 10−10, 5.9 × 10−10, and 14 × 10−10 s for UST, ssUSTm, and dsUSTm, respectively. The
water-proton relaxivities, r1 and r2, at 25 MHz, 0.59 T, and 25 ◦C, also increased in the same order and the r1 values were 0.26,
0.41, and 0.56 mM−1 s−1 for UST, ssUSTm, and dsUSTm, respectively. The r1 values of 1.00 and 2.06 mM−1 s−1 for dsUST2 and
c
dsUST5, respectively, were obtained. Copyright ꢀ 2008 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: NMR; 1H; EPR; MRI; nucleic acid; TEMPO radical
Introduction
chelates; 5.5 for GdDTPA [Gd(diethylenetriamine-N, N, Nꢁ, Nꢁꢁ, Nꢁꢁ-
pentaacetate][8] with S = 7/2, and the ease of bioreduction
in vivo, while it may be advantageous for specific targeting, which
is strongly required at the present stage, owing to their chemical
flexibility and feasible preparation.
In this study, the length and rigidity of the linker between
TEMPO and nucleic acid may affect the rotational correlation time
fortherelaxivities. First, aDNAnucleotidecarryingTEMPOthrough
a relatively flexible linker and its oligomers and their relaxivities,
r1 and r2 of longitudinal and transverse relaxivities, respectively,
wereprepared. Therotationalcorrelationtimewasinvestigatedby
pulse NMR spectrometry (25 MHz and 0.59 T) and X-band electron
paramagnetic resonance (EPR) spectrometry, respectively.
The paramagnetic species strongly affect the relaxation time of
the protons of water molecules surrounding them and are used
to enhance the contrast in magnetic resonance imaging (MRI).[1]
Recently, the larger size molecules[2–4] (polymers, dendrimers,
and assemblies) carrying paramagnetic species have been studied
intensively for the improvements of their relaxation time and the
additional function such as the specificity to a tissue or organ.[3] In
thesestudies, thesystemconsistingofthenitroxideradicalandthe
protein[4] showed high water-proton relaxivities and suggested
that although the contribution of the outer sphere translation was
dominant for the relaxation time,[5] the contribution of the inner
sphere including the parameters of the rotational reorientation
time, theelectronrelaxationtime, andtheexchangetimeforwater
molecule could not be ignored and the water molecule bounded
on the surface of protein played an important role in determining
the relaxation time. For the design of a macromolecule with the
high water-proton relaxivity, we considered the introduction of a
paramagneticspeciestoalargeandrigidmoleculecollectingmany
water molecules. Although DNA helix in solution is a semiflexible
macromolecule, this time, a combination of DNA oligonucleotide
and stable radical, 2, 2, 6, 6-tetramethylpiperidine-1-oxyl (TEMPO),
as a main framework and a paramagnetic species, respectively,
was selected as a fundamental model. The DNA derivative can
vary systematically in the molecular size in order of a nucleic
acid monomer, a single strand of oligonucleotide incorporating
monomer, and its double strand with a complementary strand
(Fig. 1). In addition, the structure of DNA double strand binding
the water molecules, and the dynamics of its water molecules
have been investigated in detail by X-ray crystallography[6] and
various NMR methods,[7] respectively. TEMPO radical with spin
quantum number S = 1/2 has some disadvantages such as
the intrinsically low relaxivities (0.2 mM−1 s−1 at 25 MHz and
25 ◦C) compared with that for paramagnetic inorganic metal ion
Experimental
Synthesis
5-(2-Carboxyethyl)-2ꢁ-deoxyuridine[9] and 4-amino TEMPO were
prepared by the procedure reported in the literature. 1-acetoxy-2,
2, 6, 6-tetramethyl-4-amino-piperidine was prepared from 4-oxo-
TEMPO by protection with acetyl moiety and then reductive
amination with NaBH3CN. The synthetic procedures and charac-
terizations for the compounds and oligmers shown in Scheme 1
have been described in Supporting Information (S1).
∗
Correspondenceto:Noboru Koga,GraduateSchoolofPharmaceuticalSciences,
Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582 Japan.
E-mail: koga@fc.phar.kyushu-u.ac.jp
a
Graduate School of Pharmaceutical Sciences, Kyushu University, 3-1-1
Maidashi, Higashi-ku, Fukuoka, 812-8582 Japan
b
Faculty of Pharmaceutical Sciences at Kagawa Campus, Tokushima Bunri
University, 1314-1 Shido, Sanuki, Kagawa, 769-2101 Japan
c
Magn. Reson. Chem. 2008, 46, 1055–1058
Copyright ꢀ 2008 John Wiley & Sons, Ltd.

Y. Sato et al.
strand (dsUST) with its complementary strand is shown in Scheme
1. The UST was prepared by the coupling reaction of 5-(2-
carboxyethyl)-2ꢁ-deoxyuridine[9] and4-aminoTEMPOwith1-ethyl-
3-(3-dimethylaminopropyl)carbodiimidehydrochloride(EDC),and
was obtained as a single crystal containing MeOH. Its molecular
structure was analyzed by X-ray crystallography (Figure S2,
Supporting Information). USTAc, in which the TEMPO radical
is protected with acetyl moiety, was incorporated into the
oligonucleotide (15 mer) by a standard amidite DNA synthesis.[10]
Taking the stability of the double strand formation into account,
the sequence with a relatively high melting temperature, Tm (Tm
is 63.2 ◦C for an unmodified duplex) was selected. After cleavage
withaqueousammonia,thedeprotectionreactionofacetylmoiety
with aqueous NaOH solution (ca. 0.5 M), followed by autoxidation,
gave the single strand (ssUST) incorporating UST. The formation
of radicals was followed by HPLC technique and the reaction was
confirmed to be quantitative (Figure S1, Supporting Information).
By using this procedure, two single strands incorporating UST at
the terminal and the middle position of the sequence, ssUSTt
and ssUSTm, respectively, were obtained. Similarly, the single
strands incorporating two and five UST units, ssUST2 and ssUST5,
respectively, were prepared. After purification by a preparative
HPLC, the single strands obtained were characterized by matrix-
assisted laser desorption/ionization time-of-flight (MALDI-TOF)
mass spectroscopy. Those single strands were mixed with their
complementarystrandtoaffordthecorrespondingdoublestrands
(dsUST). The values of Tm for the formation of double strand were
determined from the temperature dependence of the absorption
at 260 nm (Table 1 and Figure S3). Tm values of around 63 ◦C
were obtained for all dsUST except dsUST5, whose Tm value was
53.2 ◦C. The circular dichroism (CD) spectra of dsUST showed the
strong positive signal at ∼280 nm and negative signal at ∼250 nm
typical to the B form (Figure S4, Supporting Information). These
CD spectra were consistent with that for a natural double strand.
In order to estimate the rotational correlation times, τR, for UST,
ssUST, and dsUST, the EPR spectra were measured at 25 ◦C. The
EPR spectra of UST, ssUSTm, and dsUSTm are shown in Fig. 2.
In the spectra, anisotropic broadening in the high field line was
observed and became broader in the order UST, ssUSTm, and
dsUSTm. The τR values estimated by analyzing the EPR spectra
Figure 1. Change of molecular size of nucleic acid monomer and DNA
strands carrying TEMPO.
Scheme 1. Preparation route.
Table 1. The values of MW, τR, r1, r2, and Tm for UST, ssUST, and
dsUST
Electron paramagnetic resonance (EPR)
a
a
τR
r1
r2
Tm
The EPR spectra were recorded on a Bruker Biospin EMX EPR X-
band (9.4 GHz) spectrometer at room temperature. Solutions (ca
0.1 mM) of UST, ssUST, and dsUST in 10 mM phosphate buffer
were used as samples.
MW
486
(×10−10 s)
(mM−1s−1
)
(mM−1s−1
)
(◦C)
UST
1.1
0.26
0.28
–
Single strand DNA oligomer (15 mer)
ssUSTm
ssUSTt
ssUST2
ssUST5
4811
4811
5022
5655
5.9
5.2
–
0.41
0.39
0.48
0.49
–
–
–
–
1H NMR
0.82 (0.41)c 1.10 (0.55)c
1.83 (0.37)b 1.92 (0.38)b
The spin–lattice and spin–spin relaxation times, T1 and T2, re-
spectively, were obtained on a JEOL JNM-MU25RAN spectrometer
(25 MHz, 0.59 T) equipped with a temperature controller. The solu-
tions (0.1–3.0 mM) of UST, ssUST, and dsUST in 10 mM phosphate
buffer were used as samples and were measured at 25 ◦C.
–
Double strand DNA oligomer (15 mer)
dsUSTm
dsUSTt
dsUST2
dsUST5
9358
9358
14 (14)b
10 (10)b
0.56
0.54
0.78
0.87
61.7
63.9
9569
–
–
1.00 (0.50)c 1.39 (0.70)c 62.6
2.06 (0.41)c 2.52 (0.50)c 53.2
10 202
a 25 MHz, 0.59 T, and 25 ◦C.
Results and Discussion
b See text for the value within parenthesis.
c Number within parenthesis is the value per radical.
Preparation route of 5-uridine nucleoside derivative carrying
TEMPO radical (UST), its single strand (ssUST), and a double
c
www.interscience.wiley.com/journal/mrc
Copyright ꢀ 2008 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2008, 46, 1055–1058

Water-proton relaxivities of DNA oligomers carrying TEMPO radicals
Figure 2. EPR spectra of UST (a), ssUSTm (b), and dsUSTm (c). Dash line (red) shows the simulation curve by Brownian diffusion model.
in terms of Kivelson’s equation[11] are 1.1 × 10−10, 5.9 × 10−10
,
Insummary,anucleosidemonomer(UST),singlestrands(ssUST,
ssUST2, and ssUST5), and the corresponding double strands
(dsUST, dsUST2, and dsUST5) were prepared. The τR estimated
by analyzing the EPR spectra became longer in order of UST,
ssUST, and dsUST; and their relaxivities (r1 and r2) also increased
in the same order. It is noted that the r1 value of 0.56 mM−1 s−1
for dsUSTm is two times higher than that (0.28) for the parent
molecule, 4-hydroxy-TEMPO. In the oligomer carrying plural USTs,
on the other hand, the values of the relaxivity, r1, increased
with an increase in the number of radicals. In dsUST5, an r1 of
2.06 mM−1s−1 was obtained.
and 1.4 × 10−9 (1.4 × 10−9) s for UST, ssUSTm, and dsUSTm,
respectively. The value in the parenthesis was obtained from
the simulation of EPR spectrum in terms of Stochastic Liouville
equation (SLE), followed by Brownian diffusion model.[12] The
value for dsUSTm was 13 times longer than the one for UST.
Obtained τR values, which were close to those for analogous
molecules reported previously,[13] indicate that the rotation of the
molecule becomes slower with variance in the order: monomer,
single strand, and double strand. The other sets of oligomers also
showedsimilartendenciesandtheτR valuesobtainedbyanalyzing
the EPR spectra are listed in Table 1. In the EPR spectra for the
oligomer carrying plural USTs, the significant line broadening of
the characteristic three-line signal was observed for UST5 but
not for UST2 (Figure S5, Supporting Information). The observed
line broadening due to the spin-spin dipolar interaction between
the spin centers suggests that the electron relaxation time is
shortened. Furthermore, the degree of line broadening in ssUST5
was larger than that in dsUST5, indicating that the distance
between the spin centers became long by the formation of the
double strand.
To understand the motion of the DNA strand accompanying
water molecules in detail, the effect of a linker and the DNA
sequence dependence of the relaxivities are being investigated.
Supporting information
Supporting information may be found in the online version of this
article.
The relaxation times, T1 and T2, were also measured at 25 ◦C by
pulse NMR spectrometry (25 MHz and 0.59 T). The concentration
dependence of T1 for UST, ssUSTm, dsUSTm, dsUST2, and
dsUST5, is shown in Figure S6 (Supporting Information). The
relaxivitiesri (i = 1and2)obtainedfromtheTi versusconcentration
plotsincreasedinorderofUST,ssUST,anddsUSTandther1 values
are 0.26, 0.41 (0.39), and 0.56 (0.54) mM−1s−1 for UST, ssUSTm
(ssUSTt), and dsUSTm (dsUSTt), respectively. Interestingly, the r1
valuefordsUSTistwotimeshigherthanthatforthecorresponding
monomer, UST, whose value was close to that for a parent radical,
4-hydroxy-TEMPO. The values for USTm and USTt were similar,
indicating that the relaxivities were not affected by the difference
in the position of radical centers. The observed increase in the
relaxivities might be due to the restricted local motion of TEMPO.
IntheLipari–Szabotheory[14] forasystemwithisotropicmolecular
tumbling, the effective rotational correlation time, τe, is given
by 1/τe = 1/τr + 1/τi, where τr is the correlation time for the
molecular tumbling and τi is the correlation time for the internal
motion within the molecular frame. The observed EPR spectra
might suggest that the τi is still dominant in this system. In the
oligomer carrying plural USTs, on the other hand, the values of
the relaxivity, r1, increased with increasing the number of radicals
and the r1 values were 0.82 (0.41), 1.83 (0.36), 1.00 (0.50), and
2.06 (0.41) mM−1s−1 for ssUST2, ssUST5, dsUST2, and dsUST5,
respectively. However, the values of the relaxivities per TEMPO for
ssUST5 and dsUST5 are lower than those for the corresponding
dsUST, which might be mainly attributed to the decrease in the
electron relaxation time observed in EPR spectra. The r2 values
also showed similar tendency that the values increased in the
same order. The physical data for UST, ssUST, and dsUST are
summarized in Table 1.
References
[1] (a) B. Hamm, T. Staks, A. Mu¨hler, M. Bollow, M. Taupitz, T. Frenzel,
K.-J. Wolf, H.-J. Weinmann and L. Lange, Radiology 1995, 195,
785; (b) P. Pavone G. Patrizio C. Buoni E. Tettamanti R. Passariello
C. Musu P. Tirone E. Felder, Radiology 1990, 176, 61.
[2] (a) A. J. Maliakal, N. J. Turro, A. W. Bosman, J. Cornel, E. W. Meijer,
J. Phys. Chem. A 2003, 107, 8467; (b) G. Francese, F. Dunand,
C. Loselli, A. Merbach, S. Decurtins, Magn. Reson. Chem. 2003, 41,
81; (c) Y. Huang, A. Nan, G. M. Rosen, C. S. Winalski, E. Schneider,
P. Tsai, H. Ghandehari, Macromol. Biosci. 2003, 3, 647; (d) A. Datta,
J. M. Hooker, M. Botta, M. B. Francis, S. Aime, K. N. Raymond, J. Am.
Chem. Soc. 2008, 130, 2546.
[3] (a) P. Caravan, Chem. Soc. Rev. 2006, 35, 512; (b) C. S. Winalski,
S. Shortkroff, R. V. Mulkern, E. Schneider, G. M. Rosen, Magn. Reson.
Med. 2002, 48, 965.
[4] (a) H. F. Bennett, R. D. Brown III, J. F. W. Keana, S. H. Koenig,
H. M. Swartz, Magn. Reason. Med. 1990, 14, 40; (b) P. Vallet,
Y. V. Haverbeke, P. A. Bonnet, G. Subra, J.-P. Chapat, R. N. Muller,
Magn. Reason. Med. 1994, 32, 11.
[5] (a) C. F. Polnaszek, R. G. Bryant, J. Chem. Phys. 1984, 81, 4038;
(b) J.-P. Korb, G. Diakova, R. G. Bryant, J. Chem. Phys. 2006, 124,
134910.
[6] (a) E. C. M. Juan, J. Kondo, T. Kurihara, T. Ito, Y. Ueno, A. Matsuda,
A. Takenaka, Nucleic Acids Res. 2007, 35, 1967; (b) M. Egli,
V. Tereshoko, M. Teplova, G. Minasov, A. Joachimiak, R. Sanishvili,
C. M. Weeks, R. Miller, M. A. Maier, H. An, P. D. Cook, M. Manoharan,
Biopolymers 1998, 48, 234.
[7] (a) B. Halle, V. P. Denisov, Biopolymers 1998, 48, 210; (b) B. Halle,
V. P. Denisov, Methods Enzymol. 2001, 338, 178.
[8] (a) P. Caravan, J. J. Ellison, T. J. McMurry, R. B. Lauffer, Chem. Rev.
1999, 99, 2293; (b) J. R. Reichenbach, T. Hackla¨nder, T. Harth,
M. Hofer, M. Rassek, U. Mo¨dder, Eur. Radiol. 1997, 7, 264.
[9] (a) F.-G. Rong,A. H. Soloway,NucleosidesNucleotides1994,13,2021;
(b) J. Telser, K. A. Cruickshank, L. E. Morrison, T. L. Netzel, J. Am.
Chem. Soc. 1989, 111, 6966.
[10] A. P. Guzaev, M. Manoharan, Tetrahedron Lett. 2001, 42, 4769.
c
Magn. Reson. Chem. 2008, 46, 1055–1058
Copyright ꢀ 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/mrc

Y. Sato et al.
[11] (a) A. Okamoto, T. Inasaki, I. Saito, Bioorg. Med. Chem. Lett. 2004, 14,
3415; (b) P. G. Fajar, in Encyclopedia of Analytical Chemistry, vol. 7,
(Ed.: R. A. Meyers), John Wiley and Sons: Chichester, 2000,, pp 5725;
(c) D. Kivelson, S. Lee, J. Chem. Phys. 1982, 76, 5746.
[13] Z. Liang, J. H. Freed, R. S. Keyes, A. M. Bobst, J. Phys. Chem. B 2000,
104, 5372.
[14] (a) G. Lipari, A. Szabo, J.Am.Chem.Soc. 1982, 104, 4546; (b) G. Lipari,
A. Szabo, J. Am. Chem. Soc. 1982, 104, 4559.
[12] D. E. Budil, S. Lee, S. Saxena, J. H. Freed, J. Magn. Reson., Ser. A 1996,
120, 155.
c
www.interscience.wiley.com/journal/mrc
Copyright ꢀ 2008 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2008, 46, 1055–1058