Purine-Derived Nitroxides for Noncovalent Spin-Labeling of Abasic Sites in Duplex Nucleic Acids

Article abstract of DOI:10.1002/chem.201705410

A series of purine-based spin labels was prepared for noncovalent spin-labeling of abasic sites of duplex nucleic acids through hydrogen bonding to an orphan base on the opposing strand and π-stacking interactions with the flanking bases. Both 1,1,3,3-tet

Full text of DOI:10.1002/chem.201705410

DOI: 10.1002/chem.201705410
Full Paper
&
Spin Labels
Purine-Derived Nitroxides for Noncovalent Spin-Labeling of
Abasic Sites in Duplex Nucleic Acids
Nilesh R. Kamble and Snorri Th. Sigurdsson*[a]
Abstract: A series of purine-based spin labels was prepared
for noncovalent spin-labeling of abasic sites of duplex nucle-
ic acids through hydrogen bonding to an orphan base on
the opposing strand and p-stacking interactions with the
flanking bases. Both 1,1,3,3-tetramethylisoindolin-2-yloxyl
and 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) were con-
jugated to either the C2- or C6-position of the purines, yield-
ing nitroxide derivatives of guanine, adenine, or 2,6-diamino-
purine. The isoindoline-derived spin labels showed extensive
or full binding to abasic sites in RNA duplexes, whereas the
TEMPO-derived spin labels showed limited binding. An ade-
nine-derived spin label (5) bound fully at low temperature
to abasic sites in both DNA and RNA duplexes when paired
with thymine and uracil, respectively, complementing the
´
previously described guanine-derived spin label G, which
´
binds efficiently opposite cytosine. Compound G was also
shown to bind to abasic sites in DNA–RNA hybrids, either in
´
the DNA- or the RNA-strand. G showed only a minor flank-
ing-sequence effect upon binding to abasic sites in RNA.
When the abasic site was placed close to the end of the
´
RNA duplex, the affinity of the spin label G was reduced; full
binding was observed at the fourth position from the
´
duplex end. In summary, spin labels 5 and G showed full
binding to abasic sites in both DNA and RNA duplexes and
are promising spin labels for structural studies of nucleic
acids by pulsed EPR methods.
Introduction
spin-labeling, two main approaches have been used. The first
method utilizes spin-labeled phosphoramidites as building
blocks for automated chemical synthesis of the spin-labeled
oligonucleotide either directly[8] or by using iodo-modified nu-
cleobases for one-column coupling with spin labels.[9] However,
the phosphoramidite approach usually requires a significant
synthetic effort and the reagents used for oligonucleotide syn-
thesis can partially reduce the nitroxide spin labels.[10] Post-syn-
thetic labeling is another approach for covalent labeling,
wherein a spin-labeling reagent is incubated with oligonucleo-
tides that contain a uniquely reactive site.[11] Post-synthetic
modification usually requires less synthetic effort than the
phosphoramidite approach and can often be performed with
commercially available reagents. However, post-synthetic label-
ing can result in side reactions and incomplete labeling.
Electron paramagnetic resonance (EPR) spectroscopy is a bio-
physical technique that is used for the investigation of struc-
ture and dynamics of biomolecules, such as nucleic acids[1] and
proteins.[1a,e,2] EPR studies require a small amount of sample
(nmoles) and can be carried out under biologically relevant
conditions. Continuous-wave (CW) EPR spectroscopy provides
information about dynamics of specific sites through line-
shape analysis of EPR spectra.[1d,3] CW-EPR has also been used
for distance measurements between two spin centers in the
range of 5–20 ꢀ,[4] whereas pulsed EPR methods, such as
pulsed electron–electron double resonance (PELDOR), also
called double electron–electron resonance (DEER), and double
quantum coherence (DQC), have been used for long-range dis-
tance measurements (20–100 ꢀ).[1a,5] When used in conjunction
with rigid spin labels, PELDOR can also provide valuable infor-
mation about the conformational dynamics of nucleic acids.[6]
Most EPR studies of nucleic acids require attachment of
paramagnetic groups at specific sites, referred to as site-direct-
ed spin labeling (SDSL).[1b,g,7] Stable aminoxyl radicals, com-
monly called nitroxides, are usually attached to the desired
sites in the nucleic acid with covalent bonds.[7] For covalent
Noncovalent spin-labeling of nucleic acids utilizes binding
through van der Waals interactions, hydrogen bonding, and p-
stacking interactions. Spin-labeled intercalators and groove
binders can bind to nucleic acids noncovalently, but lack the
sequence specificity that is required for most EPR studies.[12]
There are a few examples of small molecules that bind nonco-
valently to specific sites of nucleic acids. Guanine–guanine mis-
match-binding ligands carrying spin labels have been used to
bind to predetermined sites of nucleic acids.[13] Abasic sites in
duplex nucleic acids have also been used as binding sites for
spectroscopic labels. For example, fluorescent compounds that
bind to abasic sites have been developed by Teramae and co-
workers for detection of single-nucleotide polymorphisms in
DNAs.[14] Lhomme et al. used adenine–acridine conjugates for
noncovalent spin-labeling of a DNA duplex containing an
[a] N. R. Kamble, Prof. S. Th. Sigurdsson
University of Iceland, Department of Chemistry
Science Institute, Dunhaga 3, 107 Reykjavik (Iceland)
E-mail: snorrisi@hi.is
Supporting information and the ORCID identification number for the
author of this article can be found under https://doi.org/10.1002/
chem.201705410.
Chem. Eur. J. 2018, 24, 1 – 9
1
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
abasic site in which the adenine bound to an abasic site and
the acridine carrying the nitroxide spin label intercalated into
the DNA duplex.[15]
We have previously used abasic sites in nucleic acids for
noncovalent and site-directed spin labeling using pyrimidine-
derived spin labels.[16] The spin label Å (Figure 1), which is an
Figure 2. Structures of purine-derived spin labels 2–6.
less affinity than to abasic sites in RNA. We also found that
´
there is a minimal flanking-sequence dependence of G binding
Figure 1. Structures of the spin label Å (left) and the triazole linked-nitroxide
to abasic sites in RNA and that abasic sites need to be at least
three base pairs away from the duplex end for complete bind-
ing.
´
spin label 1 (middle). Proposed base-pairing of spin label G with C (right) at
an abasic site in duplex nucleic acids.
analogue of cytosine (C), was found to bind specifically to
abasic sites in duplex DNA opposite guanine (G).[16a] This label
has been used for determination of distance as well as relative
orientation between two spin labels in duplex DNA and in
DNA–protein complexes.[17] However, the binding of Å was
highly flanking-sequence dependent, showing full binding to
only a few sequences. Moreover, Å showed only partial binding
to an abasic site of an RNA duplex.[16c] A series of pyrimidine-
derived nitroxides was subsequently prepared and screened
for binding to both DNA and RNA duplexes.[16c] Only triazole-
linked nitroxide 1 (Figure 1) showed complete binding to
abasic sites in RNA, but it was not a useful spin label due to
extensive nonspecific binding.[16c]
Given the limitations of the pyrimidine-based spin labels for
noncovalent binding, we turned our attention to purine-de-
rived nitroxides, because purines have a larger area for stack-
ing interactions than pyrimidines. Recently, we reported the
[18]
´
semi-flexible G (G-spin, Figure 1), which is a conjugate of
guanine and an isoindoline-derived nitroxide radical. The spin
´
label G was found to bind specifically to abasic sites in DNA
duplexes at low temperatures when paired with C as an
´
orphan base on the opposite strand. More importantly, G was
found to bind with much higher affinity to RNA, providing for
the first time, a spin label that bound effectively to specific
sites in RNA through noncovalent interactions.[18]
To further explore the use of purine-derived nitroxides for
noncovalent spin-labeling of abasic sites in duplex DNA and
RNA, we have prepared five new spin labels (2–6, Figure 2).
The isoindoline-derived spin labels showed good binding affin-
ity and specificity to both DNA and RNA duplexes containing
an abasic site but the 2,2,6,6-tetramethylpiperidine-1-oxyl
(TEMPO) derived spin labels bound poorly or not at all. Ade-
nine (A)-derivative 5 bound fully to an abasic site in duplex
RNA, providing the first nitroxide for noncovalent labeling of
´
abasic sites opposite uridine (U). G can be used as a spin label
for abasic sites in DNA–RNA hybrids, although it binds with
Scheme 1. Synthetic schemes for spin labels 2–6.
&
&
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
2
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Results and Discussion
The new purine-derived spin labels (Figure 2) contain
nitroxides at either the C2- or the C6-position, afford-
ing either adenine-, guanine-, or 2,6-diaminopurine-
derived spin labels. Two different nitroxides, 2,2,6,6-
tetramethylpiperidine-1-oxyl (TEMPO) and 1,1,3,3-tet-
ramethylisoindolin-2-yloxyl, were conjugated to the
purines. Assuming Watson–Crick pairing to either U
or C on the opposite strand, guanine-derived com-
´
pounds 6 and G were expected to direct the nitr-
oxide into the minor groove, whereas the 6-amino
modified adenine and 2,6-diamino purine spin labels
2–5 would project the label into the major groove.
Syntheses of purine-derived spin labels
The spin labels were synthesized from readily avail-
able halogen-derived purines through direct nucleo-
philic displacement reactions. Commercially available
2-amino-6-chloropurine (7) was heated with either 4-
amino-TEMPO (8) or isoindoline nitroxide 9[19] to
obtain 2,6-diamino purine derivatives 2 and 3, re-
spectively (Scheme 1A and B). 6-Chloropurine (10)
was reacted with 8 or 9 under similar conditions to
give the corresponding spin-labeled adenine deriva-
tives 4 and 5 (Scheme 1C and D). For the synthesis
of the guanine-derived TEMPO derivative 6, we first
attempted a reaction between 8 and 2-bromohypo-
xanthine in N,N-dimethylformamide (DMF) in the
presence of triethylamine, but this resulted in the
formation of a complex mixture of products (data
not shown). Hence, the following sequence of reac-
Figure 3. Binding of spin labels to abasic sites in DNA duplexes. EPR spectra of the spin
labels are shown on the far left and EPR spectra of the labels in the presence of an un-
modified DNA on the far right. The central four columns of EPR spectra show the spin-
label in the presence of DNA duplexes containing an abasic site (denoted by “_”) oppo-
site the orphan bases G, T, A, and C. Only a part of the construct is shown on top; the
complete DNA sequence is 5’-d(GACCTCG_ATCGTG)-3’·5’-d(CACGATXCGAGGTC)-3’, where
X represents the orphan base (G, T, A, or C). The arrows by the spectra in the top left
corner identify a slow-moving component in the EPR spectra, which indicates binding of
a spin label to the nucleic acid. EPR spectra inside the black boxes showed almost fully
´
bound G and 5 to the abasic site of a DNA duplex opposite C and T, respectively. EPR
spectra of the spin labels (200 mm) in the presence of DNA duplexes (400 mm) were re-
corded in phosphate buffer (10 mm NaHPO4, 100 mm NaCl, 0.1 mm Na2EDTA, pH 7.0)
containing 30% ethylene glycol and 2% DMSO at À308C. The spectra were phase-cor-
rected and aligned with respect to the height of the central peak.
tions was used instead. 2-Amino-6-benzyloxy purine[20] (7) was
reacted with sodium nitrite in the presence of tetrafluoroboric
acid[21] to yield 2-fluoro-6-benzyloxypurine[20] (11) (Scheme 1E).
Compound 11 was subsequently reacted with 4-amino-TEMPO
(8) to give 12 in moderate yield, which, upon debenzylation,
gave the target compound 6 in excellent yield.
derivatives 2, 4, and 6 showed very poor binding, whereas the
2,6-diaminopurine derivative 3 showed little to moderate bind-
ing (18–64%) to the abasic sites of the DNA duplexes. Notably,
all the isoindoline-derived spin labels bind most extensively to
an orphan base that can form a Watson–Crick base pair. How-
ever, extensive binding was observed to the other pyrimidine
for these labels. This holds especially true for the 2,6-diamino-
purine derivative 3, because it binds to a somewhat similar
extent to abasic sites with C (49%) and T (64%). For all the
spin labels, there is considerably less binding to an abasic site
containing a purine orphan base. To verify that the spin labels
bind to the abasic site specifically, they were individually
mixed with an unmodified 14-mer DNA duplex (Figure 3, far
Binding of the spin labels to abasic sites in duplex DNA
The extent of binding and binding specificity of spin labels 2–
6 to abasic sites in DNA duplexes were determined by EPR
spectroscopy. Each spin label was incubated with a 14-mer
duplex DNA containing G, T, A, or C as the orphan base at
À308C (Figure 3). The EPR spectrum of each spin label
(Figure 3, far left column) showed comparatively narrow lines,
because of the fast tumbling of the radical in solution. In the
presence of the DNA duplexes containing an abasic site, the
appearance of a slow moving component in the spectrum
(arrows in the top left spectrum) indicated binding of the spin
label to the DNA. All the spin labels showed some binding af-
finity towards at least one DNA duplex. Two spectra indicated
nearly full binding (>98%) to the abasic site (Figure 3, spectra
´
right column). Isoindoline-derived spin labels G, 3, and 5
showed very minor levels of nonspecific binding (<3%).
Binding of the spin labels to abasic sites in duplex RNA
The extent of binding of the spin labels to abasic sites in RNA
duplexes was also investigated (Figure 4). Here, the TEMPO-de-
rived spin labels 4 and 6 also showed very limited binding
(>30%) except for the 2,6-diamino derivative 2, which showed
extensive binding opposite C (87%). The isoindoline-derived
´
in boxes), for previously reported spin label G opposite C and
´
adenine-derived spin label 5 opposite thymine (T). The TEMPO
spin labels G, 3, and 5 had higher affinity to the abasic sites in
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
3
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
an orphan base that can form Watson–Crick base
pairs with the spin label. As was observed for DNA,
the isoindoline-derived spin labels show extensive
binding to abasic sites in RNA that contain either of
the pyrimidine orphan bases. In particular, the ade-
nine-derivative 5 binds almost fully (>98%) when
paired with C. Combined with the fact that these
spin labels have limited or no binding to abasic sites
containing purines as the orphan base, it is clear that
the shape/size of the binding pocket has a large
effect on the specificity of binding.
To obtain information about the relative affinities
´
of the isoindoline-derived spin labels G, 3, and 5, the
temperature dependence of their binding was inves-
´
tigated (Figure 5). Spin label G was by far the best
binder, with extensive binding being observed even
at 208C (>95%, Kd =6.15ꢂ10À6 m)[16a] at which tem-
perature spin labels 3 and 5 showed nearly no bind-
ing.
Binding of the spin labels to abasic sites in DNA–
RNA hybrids
DNA–RNA hybrids are heterogeneous nucleic acids
and key intermediates in many important biological
processes.[22] These hybrids are recognized by RNase
H and have been used for biomedical technologies
such as antisense therapies.[22c,23] Structures of sever-
al hybrids have been characterized using nuclear
magnetic resonance (NMR) spectroscopy[24] and X-ray
crystallography[25] and shown to be in an A-form
duplex. Spin labeling of DNA–RNA hybrid duplexes
have, to our knowledge, not been explored before.
Figure 4. Binding of spin labels to abasic sites in RNA duplexes. EPR spectra of spin
labels in the presence of an unmodified RNA are shown on the far right. The central four
columns of the EPR spectra show the spin labels in the presence of RNAs containing an
abasic site (denoted by “_”) opposite the orphan bases G, U, A, and C. Only a part of con-
struct is shown on top; the complete RNA sequence is 5’-GACCUCG_AUCGUG-3’·5’-CAC-
GAUXCGAGGUC-3’, where X represents the orphan base (G, U, A, or C). EPR spectra
inside the black boxes indicate full or nearly full binding. EPR spectra of the spin labels
(200 mm) in the presence of RNA duplexes (400 mm) were recorded in phosphate buffer
(10 mm NaHPO4, 100 mm NaCl, 0.1 mm Na2EDTA, pH 7.0) containing 30% ethylene glycol
and 2% DMSO at À308C.
´
Given the extensive binding of the spin label G to
RNA duplexes than in DNA duplexes, showing full or nearly full
abasic sites in RNA, we selected it to probe its binding to
abasic sites in DNA–RNA duplexes.
´
binding to at least one sequence (Figure 4, black boxes). G
´
bound to an abasic site opposite C, 2,6-diamino purine 3 op-
posite U, whereas adenine derivative 5 bound opposite both
pyrimidines, although it bound slightly better to U. All these
spin labels show the best binding to abasic sites that contain
To evaluate the binding affinity of G to the abasic site of
DNA–RNA duplexes, two 14-mer duplexes were prepared. The
first duplex contained an abasic site in the DNA strand (DNA-
RNA I) and in the second duplex, the abasic site was placed in
Figure 5. EPR spectra of the spin labels in the presence of an RNA duplex containing an abasic site at different temperatures. Only a part of construct is
shown; the complete RNA sequence is 5’-GACCUCG_AUCGUG-3’·5’-CACGAUXCGAGGUC-3’; the abasic site (denoted by “_”) and X represents orphan base C
or U. EPR spectra of the spin labels (200 mm) in the presence of RNA duplexes (400 mm) were recorded in phosphate buffer (10 mm NaHPO4, 100 mm NaCl,
0.1 mm Na2EDTA, pH 7.0) containing 30% ethylene glycol and 2% DMSO.
&
&
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
4
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
RNA duplexes (III–VI, Figure 7) were prepared, such that the
position of the abasic site was moved one base pair at a time
from the 5’-end towards the center of the duplex. Quite re-
´
markably, G showed extensive binding (77% at À308C) when
´
Figure 6. Binding of G to an abasic site of DNA–RNA hybrid duplexes. The
DNA–RNA I contained an abasic site in the DNA strand (left column), where-
as the DNA–RNA II contained an abasic site in the RNA strand (right
´
column) and the abasic sites are denoted by “_”. EPR spectra of G (200 mm)
´
Figure 7. Binding of G (200 mm) to RNA duplexes III–VI (400 mm) containing
in the presence of DNA–RNA hybrid (400 mm) were recorded in phosphate
buffer (10 mm NaHPO4, 100 mm NaCl, 0.1 mm Na2EDTA, pH 7.0) containing
30% ethylene glycol and 2% DMSO.
abasic sites (denoted by “_”), where the location of the abasic site is moving
from the 5’-end to the center of the duplex. All EPR spectra were recorded
in phosphate buffer (10 mm NaHPO4, 100 mm NaCl, 0.1 mm Na2EDTA,
pH 7.0) containing 30% ethylene glycol and 2% DMSO.
´
the RNA strand (DNA-RNA II). Figure 6 shows that G bound
fully to both hybrid duplexes at À308C, extensive binding (45–
50%) was even observed at 08C. However, increasing the tem-
perature to 208C resulted in almost complete loss of binding.
a single C-overhang was placed at the end of the RNA duplex
(III), similar to the degree of binding to an abasic site placed
one base pair away from the end of the duplex (IV). When the
abasic site was placed at the third position from the duplex
terminus (V), the spin label bound fully at À308C. However,
the data recorded at À208C shows that the abasic site needs
to be at least at the fourth position from the duplex end (VI)
to achieve full binding.
´
We determined the binding affinities of G to all the duplexes
(DNA–DNA, RNA–RNA, and DNA–RNA) at 08C, where the
abasic site was opposite C (see the Supporting Information
´
Figure S1). G had the highest affinity to the RNA duplex (Kd =
1.46ꢂ10À7 m), the second highest to the DNA–RNA (Kd =9.75ꢂ
10À7 m) and the lowest affinity to the DNA duplex (Kd =60.17ꢂ
´
To investigate the binding efficiency of G as a function of
10À7 m). Thus, the binding affinity of G decreased sixfold when
the identity of the base pairs immediately flanking the abasic
site, a series of sixteen 14-mer RNA duplexes, with all possible
combinations of the flanking bases, were prepared and incu-
´
bound to DNA–RNA hybrids and 41-fold when it was bound to
the DNA duplex at 08C, compared with the RNA duplex.
´
bated with G. Visual inspection of the EPR spectra (see the
Supporting Information, Figure S2) revealed relatively minor
variations in the extent of binding between the flanking se-
quences. The binding was quantified by determining the disso-
ciation constant (Kd) at 208C for all the flanking sequences,
which showed a variation of Kd within a factor of two (see the
Supporting Information, Figure S3). For comparison, up to 15-
fold difference in the Kd was observed between flanking se-
quences for binding of the spin label Å to abasic sites in DNA
duplexes.[16b] For a given base at the 5’-side of the abasic site,
Effect of structural changes in RNA duplexes on binding of
´
G
Three main factors contribute to spin-label binding to an
abasic site. First, hydrogen bonding of the spin-label base to
the orphan base on the opposing strand, as described above.
Second, the identity of the bases flanking the abasic site will
affect the stacking interaction with the spin label. Third, place-
ment of the abasic site close to the end of the duplex might
compromise the structural integrity of the abasic binding site,
because terminal base pairs are more dynamic than the central
base pairs.[26] We decided to investigate the latter two factors
´
the highest affinity for G was observed for a U at the 3’-side.
Conclusions
´
in RNA using the spin label with the highest affinity (G).
To probe the effect of the location of the abasic site relative
to the duplex end on spin-label binding, four different 14-mer
We have demonstrated that readily synthesized purine-derived
nitroxide spin labels can be used for noncovalent spin-labeling
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
5
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Compound 4: Yield: 15 mg (16%); Rf =0.28 (CH2Cl2/MeOH 9:1);
HRMS-ESI: m/z calcd for C17H20N7O [M+H]+ 290.1849; found:
290.1850.
of abasic sites in nucleic acid duplexes, in particular RNA–RNA
duplexes. Specifically, isoindoline-derived spin labels 3, 5, and
´
G are superior for noncovalent spin-labeling when compared
Compound 5: Yield: 30 mg (29%); Rf =0.28 (CH2Cl2/MeOH 9:1);
HRMS-ESI: m/z calcd for C17H19N6O [M+H]+ 324.1693; found:
324.1674.
with the TEMPO-derived spin labels 2, 4, and 6. The adenine
´
derivative 5 and the guanine-nitroxide conjugate G bound effi-
ciently to an abasic site of DNA duplexes and showed full
binding to RNA duplexes. For all spin labels, the highest extent
of binding was observed when the orphan base offered the
possibility of Watson–Crick pairing; that is, when the adenine
derivatives bound to T (DNA) or U (RNA) or the guanine deriva-
tives to C. This indicates that hydrogen bonding is a significant
contributor to spin-label binding. The adenine spin label 5
showed full binding to an abasic site of an RNA duplex oppo-
Compound 11: Compound 7 (100 mg, 0.41 mmol) was added to
50% aqueous tetrafluoroboric acid (10 mL) and the solution was
stirred at À208C for 15 min, followed by addition of aqueous
sodium nitrite (2 mL, 1.5m). The reaction mixture was stirred for
1 h at À108C, followed by neutralization with a satd. solution of
sodium carbonate at 108C. The precipitate was filtered off and
washed with cold water (10 mL) and dried in vacuo to give 11
(50 mg, 50%) as a pale-yellow solid. Rf =0.62 (CH2Cl2/MeOH 9:1);
1H NMR (400 MHz, [D6]DMSO): d=13.64 (s, 1H), 8.41 (s, 1H), 7.55
(m, 2H), 7.44 (m, 3H), 5.60 ppm (s, 2H); 19F NMR (400 MHz,
[D6]DMSO): d=À53.08 ppm; HRMS-ESI: m/z calcd for C12H9FN4O
[M+Na]+ 267.0653; found: 267.0650.
´
site U and complements G as a spin label, which pairs with C.
´
It was also demonstrated that G binds efficiently to abasic sites
of DNA–RNA hybrids. Only a minor flanking-sequence effect
´
was observed upon binding of G to an abasic site in RNA–RNA
Compound 12: Compound 11 (10 mg, 0.04 mmol) and 4-amino-
TEMPO (8) (8.5 mg, 0.05 mmol) were added to a solution of anhy-
drous DMF (0.7 mL) and Et3N (18 mL, 0.12 mmol) and heated at
1008C for 12 h. The reaction mixture was cooled to RT and the sol-
vent was evaporated in vacuo. The crude material was purified by
flash column chromatography (silica gel) using a gradient elution
(CH2Cl2: 30% NH3 in MeOH; 100:0 to 90:10) to give 12 (4 mg, 24%)
as a pale-yellow solid. Rf =0.32 (CH2Cl2/MeOH 9:1); HRMS-ESI: m/z
calcd for C12H9FN4O [M+H]+ 396.2268; found: 396.2261.
´
duplexes. Thus, spin labels G and 5 are promising spin labels
for structural studies of RNA and its complexes with macromo-
lecules.
Experimental Section
General materials and methods
Compound 6: To a solution of compound 12 (12 mg, 0.02 mmol)
in MeOH (5 mL) was added 10% Pd/C (1 mg) under argon. The
mixture was stirred under H2 gas (55 psi) at 228C for 16 h, the mix-
ture was filtered through a pad of Celite and the filtrate was con-
centrated in vacuo to give 6 (7 mg, 80%). Rf =0.32 (CH2Cl2/MeOH
9:1); HRMS-ESI: m/z calcd for C14H21N6O2 [M+H]+ 306.1799; found:
306.1792.
All reagents were purchased from Sigma–Aldrich and were used
without further purification. Dichloromethane and acetonitrile
were dried over calcium hydride and freshly distilled before use.
Thin-layer chromatography (TLC) was performed using glass plates
pre-coated with silica gel (0.25 mm, F-25, Silicycle) and compounds
were visualized under UV light. Column chromatography was per-
formed using 230–400 mesh silica gel (Silicycle). 1H NMR spectra
were recorded with a Bruker Avance 400 MHz spectrometer. Chem-
ical shifts are reported in parts per million (ppm) relative to the
partially deuterated NMR solvent [D6]DMSO (2.50 ppm). Nitroxide
radicals show significant broadening in NMR spectra and loss of
NMR signals due to their paramagnetic nature[27] and, therefore,
those spectra are not reported. The EPR spectra of the radicals are
shown in the Supporting Information. Mass spectrometric analyses
were performed with an HRMS-ESI (Bruker, MicroTOF-Q) in positive
ion mode. All EPR data were recorded in a phosphate buffer
(10 mm NaHPO4, 100 mm NaCl, 0.1 mm Na2EDTA; pH 7) containing
30% ethylene glycol and 2% DMSO.
DNA and RNA synthesis and purification
Phosphoramidites, CPG columns, 5-benzylthiotetrazole and aceto-
nitrile for oligomer synthesis were purchased from ChemGenes
Corp., USA. All other required reagents and solvents were pur-
chased from Sigma–Aldrich. Unmodified oligonucleotides and oli-
gonucleotides containing abasic sites were synthesized with an au-
tomated ASM800 DNA synthesizer (Biosset, Russia) using a trityl-off
protocol and phosphoramidites with standard protecting groups
on 1.0 mmole scale (1000 ꢀ CPG columns). The DNAs were depro-
tected from solid support using 33% aqueous ammonia solution
at 558C for 8 h, whereas the general deprotection for RNAs was
done using 1:1 solution (2 mL) of CH3NH2 (8m in EtOH) and NH3
(33% w/w in H2O) at 658C for 45 min. The solvent was removed in
vacuo and the TBDMS-protecting groups were removed by incuba-
tion in NEt3·3HF (600 mL) for 90 min at 558C in DMF (200 mL), fol-
lowed by addition of water (200 mL) and precipitation in 1-butanol.
The oligonucleotides were purified by 20% denaturing polyacryl-
amide gel electrophoresis. The oligonucleotides were visualized
under UV light and the bands were excised from the gel, crushed,
and extracted from the gel matrix with a Tris buffer (250 mm NaCl,
10 mm Tris, 1 mm Na2EDTA, pH 7.5). The extracts were filtered
through 0.45 mm, 25 mm diameter GD/X syringe filters (Whatman,
USA) and desalted using Sep-Pak cartridges (Waters, USA), accord-
ing to the manufacturer’s instructions. After removing the solvent
in vacuo, the oligomers were dissolved in deionized and sterilized
water (200 mL). Oligonucleotides were quantified using Beer’s law
and measurements of absorbance at 260 nm, using extinction coef-
General procedure for the syntheses of spin labels 2–5
6-Chloropurine 7 or 10 (0.29 mmol) and nitroxide radical 8 or 9
(0.29 mmol) were added to a solution of nBuOH (4 mL) and Et3N
(0.88 mmol). The reaction mixture was heated at 1208C for 16 h,
cooled to RT and the solvent was evaporated in vacuo. The crude
product was purified by flash column chromatography (silica gel)
using a gradient elution (CH2Cl2/30% NH3 in MeOH; 100:0 to
90:10) to give compounds 2–5 as pale-yellow solids.
Compound 2: Yield: 20 mg (22%); Rf =0.22 (CH2Cl2/MeOH 9:1);
HRMS-ESI: m/z calcd for C14H22N7O [M+H]+ 305.1959; found:
305.1950.
Compound 3: Yield: 24 mg (25%); Rf =0.25 (CH2Cl2/MeOH 9:1);
HRMS-ESI: m/z calcd for C17H20N7O [M+H]+ 339.1808; found:
339.1783.
&
&
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
6
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
farb, in Structural Information from Spin-Labels and Intrinsic Paramagnet-
ic Centres in the Biosciences, Springer 2012, pp. 163–204.
ficients determined by using the WinLab oligonucleotide calculator
(V2.85.04, PerkinElmer).
[6] a) A. Marko, V. Denysenkov, D. Margraf, P. Cekan, O. Schiemann, S. T. Si-
gurdsson, T. F. Prisner, J. Am. Chem. Soc. 2011, 133, 13375–13379;
b) T. F. Prisner, A. Marko, S. T. Sigurdsson, J. Magn. Reson. 2015, 252,
187–198; c) B. Endeward, A. Marko, V. Denysenkov, S. T. Sigurdsson, T. F.
Prisner, Methods Enzymol. 2015, 564, 403–425; d) C. M. Grytz, A. Marko,
P. Cekan, S. T. Sigurdsson, T. F. Prisner, Phys. Chem. Chem. Phys. 2016, 18,
2993–3002; e) C. M. Grytz, S. Kazemi, A. Marko, P. Cekan, P. Gꢃntert, S. T.
Sigurdsson, T. F. Prisner, Phys. Chem. Chem. Phys. 2017, 19, 29801–
29811.
[7] a) S. A. Shelke, S. T. Sigurdsson, Site-Directed Nitroxide Spin Labeling of
Biopolymers, Springer, 2011, pp. 121–162; b) S. A. Shelke, S. T. Sigurds-
son, Eur. J. Org. Chem. 2012, 2291–2301; c) S. A. Shelke, S. T. Sigurdsson,
in Modified Nucleic Acids, Springer, 2016, pp. 159–187.
[8] a) A. Spaltenstein, B. H. Robinson, P. B. Hopkins, J. Am. Chem. Soc. 1988,
110, 1299–1301; b) N. Barhate, P. Cekan, A. P. Massey, S. T. Sigurdsson,
Angew. Chem. Int. Ed. 2007, 46, 2655–2658; Angew. Chem. 2007, 119,
2709–2712; c) C. Hçbartner, G. Sicoli, F. Wachowius, D. B. Gophane, S. T.
Sigurdsson, J. Org. Chem. 2012, 77, 7749–7754; d) D. B. Gophane, S. T.
Sigurdsson, Chem. Commun. 2013, 49, 999–1001.
EPR measurements
Solutions for CW-EPR experiments were prepared by mixing ali-
quots of stock solutions of a single-stranded oligomer containing
an abasic site, its complementary strand, and the spin label
(1:1.2:0.5). The solvent was evaporated in vacuo and the resulting
residue was dissolved in phosphate buffer (10 mL; 10 mm NaHPO4,
100 mm NaCl, 0.1 mm Na2EDTA, pH 7.0) and annealed: 908C for
2 min, 608C for 5 min, 508C for 5 min, 228C for 15 min and dried
using a SpeedVac. The residue was dissolved in aqueous 30% eth-
ylene glycol (10 mL) containing 2% DMSO and placed in a 50 mL
quartz capillary (BLAUBRAND intraMARK) (final concentration of
nucleic acid duplex 200 mm). The EPR spectra were recorded using
100–200 scans with a MiniScope MS200 (Magnettech Germany)
spectrometer (100 kHz modulation frequency, 1.0 G modulation
amplitude and 2.0 mW microwave power). Magnettech tempera-
ture controller M01(Æ0.58C) was used as a temperature regulator.
[9] O. Schiemann, N. Piton, J. Plackmeyer, B. E. Bode, T. F. Prisner, J. W.
Engels, Nat. Protoc. 2007, 2, 904–923.
[10] a) N. Piton, Y. Mu, G. Stock, T. F. Prisner, O. Schiemann, J. W. Engels, Nu-
cleic Acids Res. 2007, 35, 3128–3143; b) P. Cekan, A. L. Smith, N. Barhate,
B. H. Robinson, S. T. Sigurdsson, Nucleic Acids Res. 2008, 36, 5946–5954.
[11] a) P. Z. Qin, K. Hideg, J. Feigon, W. L. Hubbell, Biochemistry 2003, 42,
6772–6783; b) T. E. Edwards, S. T. Sigurdsson, Nat. Protoc. 2007, 2,
1954–1962; c) U. Jakobsen, S. A. Shelke, S. Vogel, S. T. Sigurdsson, J. Am.
Chem. Soc. 2010, 132, 10424–10428; d) G. Sicoli, F. Wachowius, M. Ben-
nati, C. Hçbartner, Angew. Chem. Int. Ed. 2010, 49, 6443–6447; Angew.
Chem. 2010, 122, 6588–6592; e) S. Saha, A. P. Jagtap, S. T. Sigurdsson,
Chem. Commun. 2015, 51, 13142–13145; f) M. M. Haugland, A. H. El-
Sagheer, R. J. Porter, J. PeÇa, T. Brown, E. A. Anderson, J. E. Lovett, J. Am.
Chem. Soc. 2016, 138, 9069–9072; g) K. Halbmair, J. Seikowski, I. Tkach,
C. Hçbartner, D. Sezer, M. Bennati, Chem. Sci. 2016, 7, 3172–3180.
[12] a) B. K. Sinha, C. F. Chignell, Life Sci. 1975, 17, 1829–1836; b) S.-J. Hong,
L. H. Piette, Cancer Res. 1976, 36, 1159–1171; c) S.-J. Hong, L. Piette,
Arch. Biochem. Biophys. 1978, 185, 307–315; d) I. Hurley, P. Osei-Gyimah,
S. Archer, C. Scholes, L. Lerman, Biochemistry 1982, 21, 4999–5009;
e) M. F. Ottaviani, N. D. Ghatlia, S. H. Bossmann, J. K. Barton, H. Duerr,
N. J. Turro, J. Am. Chem. Soc. 1992, 114, 8946–8952; f) P. B. Dervan,
Bioorg. Med. Chem. 2001, 9, 2215–2235.
[13] a) H. Atsumi, K. Maekawa, S. Nakazawa, D. Shiomi, K. Sato, M. Kitagawa,
T. Takui, K. Nakatani, Chem. Lett. 2010, 39, 556–557; b) H. Atsumi, S. Na-
kazawa, C. Dohno, K. Sato, T. Takui, K. Nakatani, Chem. Commun. 2013,
49, 6370–6372.
[14] a) K. Yoshimoto, S. Nishizawa, M. Minagawa, N. Teramae, J. Am. Chem.
Soc. 2003, 125, 8982–8983; b) B. Rajendar, S. Nishizawa, N. Teramae,
Org. Biomol. Chem. 2008, 6, 670–673.
[15] a) P. Belmont, C. Chapelle, M. Demeunynck, J. Michon, P. Michon, J.
Lhomme, Bioorg. Med. Chem. Lett. 1998, 8, 669–674; b) F. Thomas, J.
Michon, J. Lhomme, Biochemistry 1999, 38, 1930–1937.
[16] a) S. A. Shelke, S. T. Sigurdsson, Angew. Chem. Int. Ed. 2010, 49, 7984–
7986; Angew. Chem. 2010, 122, 8156–8158; b) S. A. Shelke, S. T. Sigurds-
son, Nucleic Acids Res. 2012, 40, 3732–3740; c) S. A. Shelke, G. B. Sand-
holt, S. T. Sigurdsson, Org. Biomol. Chem. 2014, 12, 7366–7374.
[17] G. W. Reginsson, S. A. Shelke, C. Rouillon, M. F. White, S. T. Sigurdsson,
O. Schiemann, Nucleic Acids Res. 2013, 41, e11.
Acknowledgements
This research work was supported by the Icelandic Research
Fund (141062-051). N.R.K. gratefully acknowledges the Univer-
sity of Iceland Research Fund for providing the doctoral re-
search fellowship. We thank Dr. Subham Saha for critically
reading this manuscript and Dr. S. Jonsdottir for assistance in
collecting analytic data for structural characterization of the
compounds.
Conflict of interest
The authors declare no conflict of interest.
Keywords: DNA/RNA
· EPR spectroscopy · noncovalent
interactions · spin labels · structure elucidation
[1] a) O. Schiemann, T. F. Prisner, Q. Rev. Biophys. 2007, 40, 1–53; b) G. Z.
Sowa, P. Z. Qin, Prog. Nucleic Acid Res. Mol. Biol. 2008, 82, 147–197; c) L.
Hunsicker-Wang, M. Vogt, V. J. Derose, Methods Enzymol. 2009, 468,
335–367; d) X. Zhang, P. Cekan, S. T. Sigurdsson, P. Z. Qin, Methods Enzy-
mol. 2009, 469, 303–328; e) G. W. Reginsson, O. Schiemann, Biochem.
´
Soc. Trans. 2011, 39, 128–129; f) I. Krstic, B. Endeward, D. Margraf, A.
Marko, T. F. Prisner, in EPR Spectroscopy, Springer, 2011, pp. 159–198;
g) Y. Ding, P. Nguyen, N. Tangprasertchai, C. Reyes, X. Zhang, P. Qin, Elec-
tron Paramagn. Reson. 2014, 24, 122–147.
[2] a) H.-J. Steinhoff, Front. Biosci. 2002, 7, c97–110; b) G. Jeschke, Annu.
Rev. Phys. Chem. 2012, 63, 419–446; c) J. P. Klare, Biol. chem. 2013, 394,
1281–1300; d) E. R. Georgieva, Nanotechnol. Rev. 2017, 6, 75–92.
[3] a) W. L. Hubbell, D. S. Cafiso, C. Altenbach, Nat. Struct. Mol. Biol. 2000, 7,
735–739; b) P. Z. Qin, J. Iseri, A. Oki, Biochem. Biophys. Res. Commun.
2006, 343, 117–124; c) P. Nguyen, P. Z. Qin, Wiley Interdiscip. Rev.: RNA
2012, 3, 62–72.
[18] N. R. Kamble, M. Grꢄnz, T. F. Prisner, S. T. Sigurdsson, Chem. Commun.
2016, 52, 14442–14445.
[19] D. Reid, S. Bottle, Chem. Commun. 1998, 1907–1908.
[4] a) J. Macosko, M. Pio, I. Tinoco, Y.-K. Shin, RNA 1999, 5, 1158–1166;
b) N.-K. Kim, A. Murali, V. J. DeRose, Chem. Biol. 2004, 11, 939–948;
c) H.-J. Steinhoff, Biol. Chem. 2004, 385, 913–920.
[5] a) P. P. Borbat, J. H. Davis, S. E. Butcher, J. H. Freed, J. Am. Chem. Soc.
2004, 126, 7746–7747; b) G. Jeschke, Y. Polyhach, Phys. Chem. Chem.
Phys. 2007, 9, 1895–1910; c) G. W. Reginsson, O. Schiemann, Biochem. J.
2011, 434, 353–363; d) Z. Yang, Y. Liu, P. Borbat, J. L. Zweier, J. H. Freed,
W. L. Hubbell, J. Am. Chem. Soc. 2012, 134, 9950–9952; e) O. Duss, M.
Yulikov, G. Jeschke, F. H. Allain, Nat. Commun. 2014, 5, 3669; f) D. Gold-
[20] R. C. Moschel, A. E. Pegg, M. E. Dolan, M.-Y. Chae, US005958932A, 1999.
[21] I. R. Hardcastle, C. E. Arris, J. Bentley, F. T. Boyle, Y. Chen, N. J. Curtin, J. A.
Endicott, A. E. Gibson, B. T. Golding, R. J. Griffin, J. Med. Chem. 2004, 47,
3710–3722.
[22] a) F. Wyers, A. Sentenac, P. Fromageot, Eur. J. Biochem. 1973, 35, 270–
281; b) I. Sidorenkov, N. Komissarova, M. Kashlev, Mol. Cell 1998, 2, 55–
64; c) E. Zamaratski, P. Pradeepkumar, J. Chattopadhyaya, J. Biochem.
Biophys. Methods 2001, 48, 189–208; d) M. Nowotny, S. A. Gaidamakov,
R. J. Crouch, W. Yang, Cell 2005, 121, 1005–1016; e) A. Rich, J. Biol.
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
7
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Chem. 2006, 281, 7693–7696; f) C. Ohle, R. Tesorero, G. Schermann, N.
Dobrev, I. Sinning, T. Fischer, Cell 2016, 167, 1001–1013.
[23] T. Aboul-Fadl, Curr. Med. Chem. 2005, 12, 2193–2214.
[24] a) O. Y. Fedoroff, M. Salazar, B. R. Reid, J. Mol. Biol. 1993, 233, 509–523;
b) B. M. Znosko, T. W. Barnes, T. R. Krugh, D. H. Turner, J. Am. Chem. Soc.
2003, 125, 6090–6097.
[26] D. Andreatta, S. Sen, J. L. Pꢅrez Lustres, S. A. Kovalenko, N. P. Ernsting,
C. J. Murphy, R. S. Coleman, M. A. Berg, J. Am. Chem. Soc. 2006, 128,
6885–6892.
[27] a) T. D. Lee, J. F. Keana, J. Org. Chem. 1975, 40, 3145–3147; b) Y. Li, X.
Lei, X. Li, R. G. Lawler, Y. Murata, K. Komatsu, N. J. Turro, Chem.
Commun. 2011, 47, 12527–12529.
[25] a) G. L. Conn, T. Brown, G. A. Leonard, Nucleic Acids Res. 1999, 27, 555–
561; b) Y. Xiong, M. Sundaralingam, Nucleic Acids Res. 2000, 28, 2171–
2176; c) M. L. Kopka, L. Lavelle, G. W. Han, H.-L. Ng, R. E. Dickerson, J.
Mol. Biol. 2003, 334, 653–665; d) G. W. Han, M. L. Kopka, D. Langs, M. R.
Sawaya, R. E. Dickerson, Proc. Natl. Acad. Sci. USA 2003, 100, 9214–
9219.
Manuscript received: November 14, 2017
Version of record online: && &&, 0000
&
&
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
8
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
Binding without bonding: Isoindoline-
derived purine spin-labels bind efficient-
ly to abasic sites of duplex DNA and
RNA when they can form hydrogen
bonds to the orphan base on the op-
posing strand; the adenine-derived spin
label 5 pairs with uracil (or thymine)
&
Spin Labels
N. R. Kamble, S. Th. Sigurdsson*
&& – &&
Purine-Derived Nitroxides for
Noncovalent Spin-Labeling of Abasic
Sites in Duplex Nucleic Acids
´
and the guanine-derivative G pairs with
cytosine. The latter also binds to abasic
sites in DNA–RNA hybrid duplexes and
shows a minimal flanking-sequence
effect upon binding to abasic sites in
RNA.
Chem. Eur. J. 2018, 24, 1 – 9
www.chemeurj.org
9
ꢁ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ