EPR Spectroscopy-App in Chemistry & Biology (Topics in Current Chemistry v.321) - M.drescher & G.jeschke
EPR Spectroscopy-App in Chemistry & Biology (Topics in Current Chemistry v.321) - M.drescher & G.jeschke
EPR Spectroscopy-App in Chemistry & Biology (Topics in Current Chemistry v.321) - M.drescher & G.jeschke
Editorial Board:
K.N. Houk C.A. Hunter M.J. Krische J.-M. Lehn
l l l
With Contributions by
Editorial Board
Prof. Dr. Kendall N. Houk Prof. Dr. Steven V. Ley
University of California University Chemical Laboratory
Department of Chemistry and Biochemistry Lensfield Road
405 Hilgard Avenue Cambridge CB2 1EW
Los Angeles, CA 90024-1589, USA Great Britain
houk@chem.ucla.edu Svl1000@cus.cam.ac.uk
The series Topics in Current Chemistry presents critical reviews of the present and
future trends in modern chemical research. The scope includes all areas of chemical
science, including the interfaces with related disciplines such as biology, medicine,
and materials science.
The objective of each thematic volume is to give the non-specialist reader, whether
at the university or in industry, a comprehensive overview of an area where new
insights of interest to a larger scientific audience are emerging.
vii
viii Topics in Current Chemistry Also Available Electronically
Thus each review within the volume critically surveys one aspect of that topic
and places it within the context of the volume as a whole. The most significant
developments of the last 510 years are presented, using selected examples to illus-
trate the principles discussed. A description of the laboratory procedures involved
is often useful to the reader. The coverage is not exhaustive in data, but rather
conceptual, concentrating on the methodological thinking that will allow the non-
specialist reader to understand the information presented.
Discussion of possible future research directions in the area is welcome.
Review articles for the individual volumes are invited by the volume editors.
In references Topics in Current Chemistry is abbreviated Top Curr Chem and is
cited as a journal.
Impact Factor 2010: 2.067; Section Chemistry, Multidisciplinary: Rank 44 of 144
Preface
ix
x Preface
M. Drescher
G. Jeschke
References
Radicals in Flavoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Erik Schleicher and Stefan Weber
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
xiii
.
Top Curr Chem (2012) 321: 140
DOI: 10.1007/128_2011_237
# Springer-Verlag Berlin Heidelberg 2011
Published online: 17 September 2011
Abstract The modern chemical industry relies heavily on homogeneous and hetero-
geneous catalysts. Understanding the operational mode, or reactivity, of these catalysts
is crucial for improved developments and enhanced performance. As a result,
various spectroscopic techniques are inevitably used to characterize and interrogate
the mechanistic details of the catalytic cycle. Where paramagnetic centres are involved,
ranging from transition metal ions to defects and radicals, EPR spectroscopy is without
doubt the technique of choice. In this review we will demonstrate the wealth and breadth
of information that can be gleaned from this technique, in the characterization of
homogenous and heterogeneous systems of catalytic importance, whilst illustrating
the advantages that modern high-field and pulsed EPR methodologies can offer.
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1 The Importance of Catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Mechanistic Understanding of Catalysis: The Role
of Electron Paramagnetic Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Scope of the Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Origins of Selectivity in Asymmetric Homogeneous Catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1 Non-Covalent Interactions in Asymmetric Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Chiral Amine Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Chiral Recognition and the Role of the Outer-Sphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 Active Catalytic Oxygen Species: Model Bio-Mimetic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1 Introduction
1
http://nobelprize.org/nobel_prizes/chemistry/laureates/2010/
2
http://nobelprize.org/nobel_prizes/chemistry/laureates/2005/
3
http://nobelprize.org/nobel_prizes/chemistry/laureates/2001/
EPR Spectroscopy in Catalysis 3
examples of this work include olefin methathesis using ruthenium carbene catalysts
(Gubbs catalyst) [2], the Heck reaction using palladium-based catalysts [3] and the
enantioselective preparation of 2,3-epoxyalcohols from primary and secondary
allylic alcohols using a Ti tartrate catalyst [46].
The aim of this review is therefore to provide a selective, rather than exhaustive,
review of the literature over the past 10 years, primarily in the field of EPR applied
to studies in catalysis. From an EPR perspective, the emphasis will be placed on the
role of advanced EPR techniques to study the structure and reactivity of homoge-
neous and heterogeneous catalysts. We will begin with a case study based on our
recent work, demonstrating the role for weak outer-sphere forces in controlling
asymmetric interactions. Next we will review recent developments in the prepara-
tion of model complexes with reactive active oxygen species, used as model
systems for biocatalysts. Recent evidence has also shown how non-innocent
organic ligands play an important role in modulating the redox properties of
organometallic complexes, and how EPR is used to study these ligand-based
radicals. We will then present some recent representative examples of more tradi-
tional areas of homogeneous catalysis where EPR has played an important charac-
terization role, such as polymerization, selective oxidations, CH activation and
DielsAlder reactions. Finally, we will turn our attention to a number of heteroge-
neous systems, specifically focusing on porous catalytic materials.
a
H Ha
Hexo Hendo
N O N N O N
V V
O O O O
[VO(1)] [VO(2)]
Hexo Hendo
N O N
V
O O
[VO(3)]
Hexo Hendo
Hexo Hendo
N N N N
Cu
Cu
O O O O
[Cu(1)] [Cu(4)]
Lewis acid centre ([VO(1)]) simply to focus on the role of outer sphere
interactions in the chiral transfer step (in the absence of unwanted ring opening
reactions). Using CW-ENDOR spectroscopy, we observed the enantiomeric dis-
crimination of chiral epoxides (specifically propylene oxide, C3H6O) by a chiral
vanadyl salen-type complex [VO(1)] [40]. CW-EPR and 1H-ENDOR spectra of
R,R0 -[VO(1)] and S,S0 -[VO(1)] were systematically recorded in R-/S-propylene
oxide. Whilst the EPR spectra of all enantiomeric combinations were virtually
identical, the 1H-ENDOR spectra were characteristically different; the
heterochiral pairwise combinations of R,R-[VO(1)]+R-C3H6O and R,R-[VO(1)]
+S-C3H6O yielded slightly different 1H-ENDOR spectra, which was attributed to
the presence of diastereomeric pairs [40]. This result showed for the first time
how the subtle structural differences between the diastereomeric adducts in frozen
solution could be detected by ENDOR [40]. Importantly, when racemic-[VO(1)]
was dissolved in racemic-propylene oxide, the resulting 1H-ENDOR
spectrum was found to be identical to the spectrum of the homochiral enantio-
meric pair R,R0 -[VO(2)]+R-C3H6O. This result represented clear proof for the
preferential binding of R-C3H6O by R,R0 -[VO(1)] (and likewise of S-C3H6O by
S,S-[VO(1)]).
Although CW-ENDOR revealed the presence of the diastereomeric adducts, it
did not provide any evidence of how the adducts are actually formed and stabilised.
Therefore we prepared two derivatives of ligand (1), [VO(2)] and [VO(3)] (Fig. 1),
and studied their interactions with simple epoxides [4144]. CW ENDOR was
used to identify the role of H-bonds responsible for the stabilisation of the
[VO(1)]+cis-2,3-epoxybutane adduct [41]. By comparison, no evidence for binding
of the trans-2,3-epoxybutane isomer was found. In combination with DFT, a series
of weak H-bonds, formed between the vanadyl complex and the epoxide substrate
were identified. Notably, an H-bond was observed between the epoxide oxygen
atom, Oep, and one of the methine protons (Hexo) of the cyclohexyl group in
[VO(1)]. Two additional H-bonds were also found to exist between the vicinal
epoxide protons and each of the two phenoxide O atoms of the salen ligand.
Crucially these combined H-bonds were proposed to facilitate the overall orienta-
tion of the more symmetrical cis-epoxide between the metal centre and the chiral
salen backbone [41]. The role of these H-bonds in orientating the substrate was
furthermore confirmed using the phenylene derivative [VO(2)] [42]. In the absence
of the key Hexo proton in [VO(2)] (Fig. 1), the H-bonding between the epoxide
and the complex was weakened [42], as evidenced not only by ENDOR but also
by CW EPR.
Other weak outer-sphere forces, such as electrostatic interactions [43] and
steric contributions [44], between the substrate and the VO-complexes [VO(1)]
and [VO(3)] were also shown to contribute to the mode of chiral binding in the
asymmetric adducts. In the specific case of [VO(3)], removal of the bulky inner
tert-butyl groups from the 3,30 positions was not found to moderate the electronic
properties of the VO centre (revealed via EPR) or its interactions with the
surrounding ligand 14N nuclei (revealed via HYSCORE), but was found to
reverse the stereoselectivity of epoxide binding [44]. Whilst homochiral
EPR Spectroscopy in Catalysis 7
achieve these high e.e.s, since the ligand not only stabilises the transition metal ion
and associated transition state, but also modulates the trajectory of the incoming
chiral substrate. Bulky framework substituents (such as tert-butyl groups) are
known to prevent stabilisation of transition states, since these have similar energies
for the two diastereomers. This is particularly true in chiral metal-salen complexes,
whereby the efficiency of the catalyst depends on the nature of the bulky
substituents at the 3,30 and 5,50 positions and regulates the orientation of the
incoming substrates, creating a high diastereofacial preference.
It is clear from the work summarised above [4045] that diastereomeric discrim-
ination of chiral substrates occurs in asymmetric complexes. On the one hand in [Cu
(1)], W-band EPR revealed a strong preference for the heterochiral adducts of [Cu
(1)] compared to the homochiral adducts [45]. On the other hand, X-band ENDOR
revealed an exclusive preference for the homochiral adducts of [VO(1)] with chiral
propylene oxide [40, 43]. The origin of these selectivities was shown to arise from a
combination of weak outer sphere interactions including H-bonding [40, 41],
electrostatic influences of the substrate [43] and the subtle steric perturbations of
key functional groups on the asymmetric ligand [45]. Whilst the bulky tert-butyl
groups may affect the stability of the transition states during the reaction, these
groups undoubtedly also affect the stereo-discrimination of chiral substrates. It is
important to note that, whilst the presence of such diastereomeric adducts are often
presumed as mechanistic intermediates, they are rarely observed directly in cases of
weak complexsubstrate interactions. The results reported here therefore demon-
strate the useful role of EPR techniques in probing such diastereomeric adducts,
which may be of direct relevance to studies in homogeneous asymmetric catalysis.
Nature has evolved iron enzymes, like non-heme iron oxygenases, capable of
carrying out hydrocarbon oxidations with high degrees of selectivity under mild
conditions [4650]. Significant efforts have therefore been made recently to repro-
duce these reactions for fine chemical production by synthesis of low molecular
weight (homogeneous) analogues. A particularly active field of research is the
preparation of artificial metalloenzymes for enantioselective catalysis [36, 51]. In
the specific case of Fe-based systems, one key part of this effort is to determine the
reactivity of non-heme iron-(hydro)peroxo species in oxidation reactions [52] and
thus to understand and mimic the key structural and functional properties of the
natural enzymes. Paramount to these investigations is the availability of synthetic
analogues that can react with molecular oxygen (O2) and its reduced forms (O2
and H2O2). A number of groups have synthesised metal complexes that mimic the
iron site in superoxide reductase (SOR) and thus developed bio-inspired iron-based
oxidation catalysts [53, 54]. The ferrous site in SOR is based on one cysteine and
four histidine ligands bound to the iron centre in 5-coordinate, square pyramidal
geometry. The mechanism of O2 reduction is not well known, but participation of
EPR Spectroscopy in Catalysis 9
+ HOO +
H H H H
N N N N
II O2 (g) III
Fe Fe
N CH3CH2CN or CH2Cl2 N
N N
H H H H
S S
(BF4-) (BF4-)
Cl Cl
(5) (6)
N N
N N N O
II III
(S = 1/2) Fe - H+ Fe (S = 5/2)
N OOH N O
N + H+ N
(7) (8)
N N N
H
OH HO
(9)
2- -
But But
NH NH
O [Cp2Fe]BF4 O
t
But Bu But But
DMF Ar -45C
N O N N O N
N N
O O O O
Mn III Mn IV
N N N N
[Cp2Fe]BF4
N DMSO Ar, rt N
(10) (11)
Fig. 2 Structures of complexes (5)(11)
EPR Spectroscopy in Catalysis 11
Whereas the above cited works of Jiang et al. [58] and Simaan et al. [60] were
heavily focused on the spectroscopic (EPR) characterization of the model FeIII-
hydroperoxo and FeIII-peroxo complexes, Bilis et al. [62] investigated the catalytic
oxidation of hydrocarbons (cyclohexane) by homogeneous and heterogeneous non-
heme FeIII centres using H2O2. The FeIII complex was based on 3-{2-[2-
(3-hydroxy-1,3-diphenyl-allylideneamino)-ethylamino]-ethylimino}-1,3-diphenyl-
propen-1-ol [Fig. 2; complex (9)]. CW EPR initially revealed the presence of a
high-spin FeIII (S 5/2) centre in a rhombic field characterized by E/D ~ 0.33.
This signal was, however, found to be heavily solvent based, since in the presence
of CH3CN a new FeIII signal with g parameters [2.052, 2.005, 1.80] was observed to
form at the expense of the high-spin EPR signal. This low-spin FeIII centre was
proposed to be FeIII-OOH. In situ EPR revealed that this low-spin EPR signal is
progressively lost during the catalytic reaction, whereas the high-spin EPR signal
remains unaffected, indicating the role of FeIII-OOH in the catalysis.
The investigation of reactive metal centres bearing oxygen intermediates has not
been confined to iron. Manganese [63, 64] and copper [65] have also attracted
significant attention. Parsell et al. [64] investigated the properties of a MnIV
complex bearing a terminal oxo ligand, which converted some phosphines to
phosphine oxides. This complex was formed starting from an [MnIIIH3buea(O)]2
([H3buea]3, tris[(N0 -tert-butylureaylato)-N-ethylene]aminato) (10) (Fig. 2), a
monomeric MnIII-O complex in which the oxo ligand derived from dioxygen
cleavage or deprotonation of water. This [H3buea]3 ligand is important since it
regulates the secondary coordination sphere by providing a sterically constrained
H-bond network around the MnIII-O unit. The MnIV-oxo species (11) (Fig. 2) was
then formed using a mild oxidant at low temperatures. The low-temperature (4 K)
X-band EPR spectrum of (11) revealed principal g values of [5.15, 2.44, 1.63],
corresponding to a system having an S 3/2 state with an E/D 0.26 [64]. The
temperature dependence of the EPR spectra and simulation of the signals indicated
a value of D 3.0 cm1 and a 55Mn hyperfine coupling of 190 MHz, comparable
to other MnIV complexes. Complex (11) did not react with PPh3 or PCy3 in DMSO,
but did react with PMePh2 via O-atom transfer to produce OPMePh2 in 5070%
yields. Oxygen-atom transfer is normally a two-electron process and would yield
phosphine oxide and the corresponding MnII complex. Evidence for the formation
of this reduced complex came from the X-band EPR spectra.
For the past 10 years, chemists have been interested in how to define the charge of a
metal ion in complexes bearing non-innocent ligands, most notably through
the works of Bill and Wieghardt [66]. As J orgensen [67] suggested many years
ago, an oxidation number that is derived from a known dn electron configuration
should be specified as the physical (or spectroscopic) oxidation number. However,
12 S. Van Doorslaer and D.M. Murphy
this is not always straightforward. When organic radicals with open-shell electron
configurations are coordinated to a transition metal ion, the oxidation state of the
metal is less well defined because the oxidation may be ligand- or metal-centred:
e
Mn O R ! Mn O R or Mn1 O R (1)
+ +
N N N N
NiIII Temp NiII
O O O O
[NiIII(1)]+ [NiII(1)]+
OAc- OAc-
N N N N
CoIII CoIII
O O O O
OAc- OAc-
t-butyl
N
N N
CoIII t-butyl
N O
O O
Co
Cl
N O t-butyl
N
[Co(13)] t-butyl
[Co(14)]
t-butyl
t-butyl
N N
Cu N
N N
Rh OTf-
OH I(16)]
N
[Rh N
t-butyl
t-butyl
(15)
temperature and the solvent. At low temperature (123 K), a NiIII-phenolate complex
was easily identified by EPR, based on the characteristic rhombic g tensor
(g [2.30, 2.23, 2.02]) typical of a low-spin |z2, 2A1 > ground state. However,
at elevated temperatures (158173 K), this rhombic signal evolves into an isotropic
EPR signal with giso 2.04, which was assigned to the NiII-phenoxyl radical,
indicating the tautomerism that can exist between the two redox states ([NiIII(1)]+
or [NiII(1)]+) (Fig. 3).
Most recently, Rotthaus et al. [80, 81] extended this study to a range of closely
related Schiff base NiII complexes, proving the formation of the NiII-phenoxyl
radical with partial delocalisation of the SOMO onto the metal orbitals. Pratt and
Stack [83, 84] have also generated and characterised the coordinated phenoxyl
radical in [CuII(1)] (Fig. 1) (labelled [CuII(1)]+) as a bio-mimetic model for
galactose-oxidase complexes. The EPR data reported in their study was consistent
with the formation of an anti-ferromagnetically coupled CuII-phenoxyl complex,
whereby oxidation of the CuII complex to CuII-phenoxyl simply resulted in an
attenuation of the original CuII EPR signal by ~15%.
In a more unusual case of a coordinated phenoxyl radical bearing a Schiff-base-
type ligand, we recently reported the identification of [CoII(1)(OAc)n](OAc)m
(n m 1 or n 2, m 0), simply by treatment of [Co(1)] with acetic acid
under aerobic conditions [85]. These conditions are analogous to those employed in
the activation of [Co(1)] for the widely used hydrolytic kinetic resolution (HKR) of
epoxides [8, 9]. Initially we investigated the electronic properties of the parent pre-
catalyst [Co(1)] in the absence of acetic acid and subsequently followed the changes
to this catalyst after the addition of acetic acid under anaerobic conditions [86]. The
pre-catalyst complex produced a CW-EPR spectrum typical for a species
possessing an |yz, 2A2 > ground state. Upon acetic acid addition under anaerobic
conditions, new high-spin [87] and low-spin [86] centres were generated. The latter
low-spin centre was characterized by the parameters g [2.41, 2.27, 2.024];
A [100, 70, 310] MHz indicative of a |z2, 2A1 > ground state, induced
by acetate ligation to the CoII complex. When molecular oxygen was introduced
into this system (or alternatively when the acetic-acid addition was conducted
directly under aerobic conditions) a new signal assigned to the phenoxyl radical
was observed. This signal was characterized by the parameters of g [2.0060,
2.0031, 1.9943]; A [17, 55, 14] MHz, readily identifiable at X- and W-band
microwave frequencies (Fig. 4) [85]. A combination of HYSCORE, Resonance
Raman and DFT results proved conclusively the presence of a coordinated phenoxyl
radical, as opposed to a bound (ligating) substrate based radical [85].
The formation of the phenoxyl radical was proposed to occur in the presence of
acetic acid by coupling the two-electron, two-proton reduction of molecular oxygen
to H2O2 [85]. In some way this process is reminiscent of the half reaction observed
in GAO. The unusual aspect, however, was its identification in the activated HKR
reaction system, although there was no evidence for its involvement in the hydro-
lysis of epoxides [85]. As Wieghardt noted in earlier works, bulky substituents are
required for stabilization of metal-coordinated phenolate radicals [66]. We con-
firmed this by activation of a Co-salen derivative, [Co(12)] (Fig. 3). In the absence
EPR Spectroscopy in Catalysis 15
oxidation. The behaviour was also in good agreement with the S 1 ground state
recently reported for the di-radical ZnII analogue of (14) [89]. As anticipated, the
oxidized [Cu(15)] complex was EPR silent (a residual, 30%, CuII EPR signal
remained in the electrochemically oxidized solution) due to the magnetic interac-
tion between the phenoxyl radicals (S 1/2) and the CuII spin.
The study of coordinated ligand radicals in metal complexes has not been
confined to phenoxyl radicals. Although much rarer, the analogous coordinated
aminyl radicals have also being investigated. B uttner et al. [90] revealed how a Rh-
based coordination complex could also support an aminyl radical, [RhI(trop2N)
(bipy)]+OTf (trop 5-H-dibenzo[a,d]cycloheptene-5-yl) labelled [Rn(16)]+
(Fig. 3). The complex essentially consisted of a trop2N radical coordinated to the
cationic Rh centre. Just like phenoxyl radicals, the unpaired electron in these
coordinated radicals may either be at the N centre or at the metal, but advanced
EPR techniques and DFT demonstrated the existence of the ligand-based paramag-
netic centre. The experimental S-band CW-EPR spectrum was simulated using a
rhombic g matrix [2.0822, 2.0467, 2.0247] and a large 14N principal hyperfine
coupling of 98 MHz [90]. Davies-ENDOR spectra were measured in order to
determine the extent of spin delocalization onto the ligand. HYSCORE was also
used to determine the principal values and orientations of the hyperfine and
quadrupole tensors for the strongly and weakly coupled nitrogens. A large Aiso
for the apical nitrogen atom of 45.1 MHz was identified. This nitrogen has pro-
nounced anisotropy, indicating the unpaired electron resides in an orbital with high
p-character. These experimental results were fully supported by the DFT
calculations and proved the paramagnetic centre was best described as the aminyl
RhI complex, rather than the RhII-amide complex (Fig. 3).
The chemical reactivity of [RhI(16)]+ was also examined in reactions with H-
atom donors, where the complex behaves as a nucleophilic radical [90]. The reaction
rates were dependent on the XH bond dissociation energies, reacting rapidly with
stannane (Bu3SnH) and thiophenol, more slowly with tert-butyl thiol and
thioglycolic acid methyl ester and not at all with phenol and triphenylsilane [90].
Linear alpha olefins (LAOs) are useful intermediates for a range of important
commodity chemicals (including surfactants, lubricants, plasticizers, etc.). They
are produced via ethylene oligomerisation, using transition metal catalysts. A major
problem associated with these catalysts is the formation of a broad chain length
distribution of a-olefins. One approach to solving this problem, operating via a
uniquely different mechanism, is ethylene trimerization and tetramerization to
1-hexene and 1-octene, respectively [91]. Recent developments have focussed on
designing highly selective catalysts and, to date, Cr catalysts account for >90% of
the literature of ethylene oligomerization, with Ti, Co, Ta and Fe catalysts also
available. The mechanism is thought to follow a metallacyclic route, involving
EPR Spectroscopy in Catalysis 17
CO +
Ph2
OC P
N
Cr (R)
OC P S
Al(OC(CF3)3)4]- Co N
CO Ph2 Cl
Cl
[Cr(17)]+
(18)
O O
O
(ClO4)2 O O
V O
N R
N N H2O
NCCH3
Fe (20a)
N NCCH3
N
O O
OH2
O O
(19) Cu O
N R
(20b)
formation of any CrIII centres. Instead a new series of CrI complexes was formed,
some of which retained the coordinated PNP or P(R)P backbone (i.e. [Cr(CO)xL]+)
whilst in other cases new CrI complexes having undergone a ligand slippage
process were identified [100].
A significant number of oligomerisation studies have also been devoted to
cobalt, as investigated by CW EPR [27, 101, 102]. Bianchini et al. [27] showed
how the position of the sulphur atom in the thienyl groups of 6-(thienyl)-2-(imino)
pyridine ligands ((18) in Fig. 5) strongly affects the catalytic activity of the
corresponding tetrahedral high-spin dihalide CoII complex following activation
with methylaluminoxane (MAO). In catalytic experiments, CoII complexes bearing
a sulphur atom in the 3-position of the thienyl ring were found to catalyse the
selective conversion of ethylene to 1-butene. From an EPR perspective, in situ
experiments revealed the occurrence of a spin-state changeover from high-spin
tetrahedral CoII to low-spin CoII following activation with MAO. The tetrahedral
(imino)pyridine cobalt complexes were EPR-silent at room temperature. However,
at low temperatures (10 K), a broad signal was detected for these high-spin states.
After MAO addition the signal changed dramatically, reverting from high-spin to
EPR Spectroscopy in Catalysis 19
low-spin (S 1/2). The most likely coordination geometry of the CoII centres in
the activated species was proposed to be square planar with two nitrogen atoms
from the (imino)pyridine ligand, a carbon atom from a methyl group released
by MAO, and a fourth ligand that might be provided by the organyl group in
the 6-position of the pyridine ring. Indeed, in the absence of either ring, as in
CoCl2N2Br, no EPR signal appeared upon treatment of the complex in toluene with
an excess of MAO in the temperature range from 293 to 20 K. It is therefore very
likely that the organyl group in the 6-position interacted with the cobalt centre.
The nature of the active sites responsible for ethylene polymerisation was also
examined in a series of related Fe and Co bis(imino)pyridine complexes [102, 103].
In these studies EPR was used as a complementary characterisation technique, in
conjunction with M ossbauer and NMR, revealing a change in oxidation state upon
activation with MAO and triethylenealuminium.
and geff 2. Treatment of the [FeIIICl(1)] complexes with PhIO and m-CPBA did
not lead to formation of S 3/2 type spectra; instead a sharp peak at g 4.2
belonging to an unidentified S 5/2 species was found. The species associated
with this signal did not contribute to the catalytic cycle, and the intensity of its EPR
signal accounted for only 10% of the total Fe concentration. From this data, the
authors proposed a new catalytic system for the asymmetric oxidation of sulphides
where the active species was shown to be an [FeIIICl(2)]+PhIO complex [121].
Iron complexes with aminopyridine ligands are also known to catalyze selective
olefin oxidation efficiently using H2O2 or CH3CO3H as terminal oxidants. Duban
et al. [123] used EPR spectroscopy to identify the intermediates formed during
the reaction cycle of [FeII(19)(CH3CN)2](ClO4)2 (Fig. 5). EPR spectra recorded
after the onset of the reaction of [FeII(19)(CH3CN)2](ClO4)2 with CH3CO3H
showed a signal at g 4.23, an axially anisotropic signal with g [2.42, 2.42,
2.67] and a broad signal at g 2. On warming the sample to room temperature,
the signal at g 4.23 decayed and was replaced by a weaker, sharper signal with
the same g factor that remained stable over several hours. The species causing
the axial signal was tentatively assigned to a mixed-valence FeIIIFeIV complex,
[(19)FeIII-O-FeIVO(19)(S)]3+.
In contrast, the EPR spectrum recorded after onset of the reaction between
[FeII(19)(CH3CN)2](ClO4)2 and H2O2 showed several signals. Low-spin ferric
hydroperoxo intermediates [FeIII(19)(OOH)(CH3CN)]2+ (g [2.218, 2.178,
1.967]) and [FeIII(19)(OOH)(H2O)]2+ (g [2.195, 2.128, 1.970]) were observed,
in addition to the dinuclear mixed valence FeIIIFeIV complex described above
[124]. Different reaction intermediates were therefore observed under the different
catalytic conditions and coincided with the differing reactivities and selectivities of
the epoxidation of olefins. Whilst it was not clear if the dinuclear FeIIIFeIV complex
could possibly act as an active species in the corresponding catalytic systems, it was
certainly clear that in the [FeII(19)(CH3CN)2](ClO4)2/H2O2/CH3COOH systems
the mono-nuclear FeIV species [FeIVO(19)(S)]2+ did play an important role.
More recently this group has followed up this work by investigating the oxidation
reactions of a series of iron complexes with aminopyridine ligands [123, 125].
Although much less common for selective oxidation, catalytic vanadium- and
copper-based complexes have also been investigated by EPR [126, 127], including
vanadium complexes based on salen derivatives [128]. These catalysts are particu-
larly active using H2O2 as a readily available oxidant. Maurya et al. [126] examined
the oxidation of p-chlorotoluene and cyclohexene catalysed by the polymer
anchored oxovandiumIV and copperII complexes of amino derived tridentate
ligands (20a, 20b) (Fig. 5). CW-EPR was primarily used to characterise the VO
and Cu EPR signals before and after the reaction. Whilst the VO catalysts displayed
minimal changes, significant changes were detected in the CuII system after the
catalytic reaction. The structural nature of the recovered CuII catalysts was not
however assigned [126]. Di-nuclear and tri-nuclear copper clusters, derived from
the enantiomeric octadentate ligand S-21 (a 1,10 -binaphthyl-2,20 diamine ligand),
were also successfully used in the oxidation of L-/D-Dopa derivatives to quinones
[127]. High enantioselectivities were observed in the oxidation of L-/D-Dopa methyl
22 S. Van Doorslaer and D.M. Murphy
Pi Pr2
CH2tBu
N V
Pi Pr2 CH2tBu
N N
(21) (22)
Cu
N N N
N Cu Cu
N Pi Pr2
N N CHtBu
N N N N V
Pi Pr2
(22a)
ester catalysed by the dinuclear Cu complex, which exhibited strong preference for
the D enantiomer. The enantioselectivity was largely lost for the trinuclear Cu
complex. A detailed X-band CW-EPR study was undertaken on these copper
complexes. The EPR analysis of the trinuclear complex, [Cu3(S)-21]6+ (Fig. 6),
had some remarkable similarities to the copper cluster found in multicopper
oxidases such as laccase. The dinuclear CuII complex exhibited the most interesting
behaviour [127] because it allowed stronger chiral recognition by the binaphthyl
residue. As the authors demonstrated convincingly, the origin of the enantios-
electivity in their complex was indeed ligand induced.
A number of EPR techniques have also been used to investigate the mechanistic
details of the catalytic DielsAlder reaction. Using CW EPR, HYSCORE and
pulsed ENDOR, Bolm et al. [132] examined the changes in the ligand sphere
surrounding their homogeneous CuII catalyst, a chiral bis-sulfoxime CuII complex
bearing labile triflate groups. Introduction of the dienophile [N-(1-oxoprop-2-en-1-yl)
oxazolidin-2-one] resulted in the formation of a new complex with well defined
hyperfine spectra. The geometry of the complex at the different stages of the
catalytic reaction was determined by EPR. In solvent-free conditions, the initial
bis-sulfoxime CuII complex possessed square planar geometry but upon addition of
the dienophile, the EPR parameters were found to be typical of a distorted, non-
symmetric square pyramidal geometry. Mims ENDOR also revealed that at least
one triflate anion directly participates in the first coordination sphere of the CuII by
occupying an axial site [132].
Although the choice of the anion (such as triflate, TfO) in the CuII complex did
not affect the overall conversion in DielsAlder catalysis, the stereoselectivity of
the reaction was considerably influenced by the choice of anion used. The authors
used EPR to investigate a series of CuII catalysts bearing TfO, SbF6, Cl and Br
anions in the presence and absence of the dienophile [133]. The profile of the EPR
spectra were significantly different for the complexes bearing the halide anions
compared to those bearing the bulkier TfO and SbF6 anions, and the spectra were
also notably dependent on the presence of the dienophile.
The authors concluded that using the anions TfO or SbF6, a complex was
formed involving an asymmetric coordination sphere around the CuII bis-sulfoxime
complex (bound via two non-equivalent nitrogens). The dienophile was suggested
to replace the two equatorially bound counterions (bound via two non-equivalent
oxygens) and the weakly bound counterions in an axial position [133]. In contrast
to this, two distinct complexes were established using the anions Cl or Br.
The first revealed CuIICuII electronelectron interactions via the halogen atoms.
Upon addition of the dienophile, this orientation was changed towards a distorted
24 S. Van Doorslaer and D.M. Murphy
7 Nanoporous Catalysts
Zeolites, i.e. microporous aluminosilicate materials with pores smaller than 2 nm,
play key roles in the fields of sorption and catalysis [134, 135]. The global annual
market for zeolites is several million tons. In the past few decades a large variety of
zeolites and related zeotype materials have been produced, whereby transition
metal incorporation is extensively used to modulate the catalytic characteristics
of these materials. Since the catalytic properties depend on the structure and
accessibility of the transition metal sites, a lot of effort is put into probing these
sites. Nevertheless, the exact nature of the transition metal incorporation is often
strongly debated, since most spectroscopic evidence for isomorphous substitution
is indirect.
If the transition metal ion is paramagnetic, EPR techniques offer an unambiguous
way to unravel the nature of the metal incorporation. This is convincingly
demonstrated in the work of D. Goldfarb and co-workers, who used high-field
ENDOR in combination with DFT to probe the isomorphous substitution of MnII
into aluminophosphate zeotypes [136, 137]. They studied a large variety of zeotype
structures, serving as examples for different channel morphologies and framework
densities of the zeolite family. In all cases, the observation of strong 31P hyperfine
interactions proved that MnII can replace the framework Al. Furthermore, high-field
ENDOR allowed these researchers to map out the full process of MnII incorporation
into different aluminophospate zeotypes during synthesis [138].
A careful X-band 27Al HYSCORE and W-band 1H ENDOR analysis showed
that from the three CuII species found in Cu-containing Si:Al zeolite Y (Si:Al 12
and 5), only one CuII was bound to the framework oxygens [139]. The other species
consisted of a copper ion with a complete coordination sphere of water and no direct
bonding with the zeolite framework. In a similar way, combined CW-EPR and
27
Al HYSCORE provided evidence of the interaction of CuII with the framework in
copper-doped nanoporous calcium aluminate (mayenite) [140]. In mayenite,
the positively charged calcium aluminate framework is counter-balanced by
extra-lattice O2 ions. Such free oxide ions are responsible for the ion conductivity
of the materials and are readily replaced by various guest anions, such as O2 and
OH. A native O2 species could indeed be identified with EPR in the Cu-doped
mayenite materials [140].
EPR Spectroscopy in Catalysis 25
Hyperfine techniques also allow for the probing of the accessibility of the
paramagnetic transition metal sites to gases and small molecules. When ammonia
is adsorbed to vanadium (VO2+)-exchanged ZSM-5, 14N HYSCORE features
typical of equatorial ammonia ligation to the vanadyl site are observed [141].
Copper-exchanged zeolites have been known for a long time to be active in NOx
decomposition. The decomposition of NO is shown to occur via formation of a CuI-
(NO)2 dimer with a paramagnetic CuI-NO monomer as a precursor [142]. This has
prompted A. P oppl and co-workers to use multi-frequency pulsed EPR and ENDOR
techniques to investigate CuI-NO adsorption complexes in a range of copper-
exchanged zeolites prepared by both solid- and liquid-state ion exchange
[143145]. 27Al HYSCORE and ENDOR analyses of these complexes in Cu-L
and Cu-ZSM-5 zeolites allowed the estimation of hyperfine parameters of an
aluminium nucleus found near the CuI-NO site. The data showed that the Al
atom is located in the third coordination sphere of the adsorbed NO [144], and
hence supporting the O2AlO2CuINO structure proposed earlier on the basis of
quantum-chemical computations [146].
In contrast, zeolites with poor de-NOx properties may be very promising as
materials to store and deliver NO under controlled conditions in clinical
applications. Indeed, nitric oxide is a crucial biomolecule in the cardiovascular,
nervous and immune systems. The non-toxic zinc-exchanged Linde type A (LTA)
zeolite has a relatively high storage capacity for NO and is hence a promising
material for clinical applications. EPR revealed that the NO monomer is interacting
more strongly with the metal sites in Zn-LTA than in the corresponding
Na-LTA [147].
Despite the current importance of microporous zeotype materials, they have the
drawback that the size of their pores limits their applicability to smaller molecules.
To overcome this, many efforts have led to the production of a wide variety of
mesoporous (2 nm < < 50 nm) siliceous and non-siliceous materials [148]. In
this respect, the discovery of the M41S family by researchers from Mobil Oil
Corporation has played a crucial role [149, 150]. The synthesis process of
mesoporous materials is based on a self-assembly process of organicinorganic
composites where the organic self-organized structures serve as a template for the
inorganic skeleton. One of the most intensively studied M41S materials is MCM-41
that possesses hexagonally packed uni-dimensional cylindrical pores with pore
diameters between 2 and 10 nm. This silica material is synthesized using the
ionic surfactant cetyltrimethylammonium bromide as template. Conversely,
mesoporous materials with larger pores and higher (hydro)thermal stability can
be obtained using non-ionic block co-polymers. In this class of materials, SBA-15
with large tailorable uniform pores (315 nm) is found to be particularly promising
[151153].
26 S. Van Doorslaer and D.M. Murphy
The formation of mesoporous materials can be followed via EPR using nitroxide
spin probes (i.e. nitroxide radicals) [154]. By introducing such a spin probe into the
system, or by labelling a molecule with this nitroxide, EPR can be used to monitor
the direct environment of this radical. The spin probe thus acts as a spy that keeps
track of the changing environment during the formation reaction. Spin-labelled
surfactants, silane-based spin-labels and spin-labelled pluronics are ideal to monitor
the formation of templated mesoporous materials [154]. By varying the type of
probes added in the reaction mixture, different regions in the forming mesostructure
can be studied. CW-EPR experiments give a direct insight in the polarity and
microviscosity of the local environment, while ESEEM experiments reveal the
water content and the presence of additives or ions in the proximity of the label
[155157]. Variations in the size of the micelles can be probed during the initial
stages of the reaction by using DEER spectroscopy, a pulsed-EPR technique
targeted at determining inter-spin distances [158].
Transition metal-based redox centres render the mesoporous silica materials
catalytically active. The transition metals can be introduced during synthesis or
postsynthetic (e.g. by impregnation) and the local structure of the transition metal
site will determine the catalytic properties. As also demonstrated for the zeolite
cases, EPR offers a unique tool to monitor these local sites. The observation
of strong 29Si hyperfine couplings in the HYSCORE spectra of vanadium-doped
MCM-41 unambiguously proved the framework incorporation of vanadium [159].
The material was obtained via a direct synthesis method at room temperature. When
vanadia was post-synthetically deposited on the surface of MCM-41 by the mole-
cular designed dispersion (MDD) method using vanadyl acetylacetonate complexes,
the vanadyl ions were, in contrast, found to be fully hydrated with no binding to the
silica walls. This agrees with our earlier findings that the vanadyl acetylacetonate
complexes have a great tendency to increase their coordination sphere by
coordinating waters when deposited on SBA-15 [160]. Interestingly, this tendency
is highly reduced when a Ti environment (e.g. a TiOx layer) is present [160].
While mesoporous materials have the advantage of larger pores, the crystalline
microporous materials are much more stable. When the silica/surfactant ratio is
increased during the SBA-15 synthesis, microporous amorphous nanoparticles are
formed inside the mesoporous channels [161]. This leads to mechanically more
stable SBA-15 materials that are named plugged hexagonal templated silica
(PHTS). When vanadium-activated zeolitic nanoparticles are deposited inside the
mesoporous channels of SBA-15 via a post-synthetic dry impregnation with the
zeolite nanoparticles, a catalytically active PHTS is formed [162]. The detailed CW
and pulsed EPR analysis of these materials not only revealed valuable information
EPR Spectroscopy in Catalysis 27
Fig. 7 The observation of strong 13C [C(2)] and weak [C(1)] hyperfine couplings in the
HYSCORE spectra after adsorption of 13CO2 to a PHTS formed by deposition of vanadium
silicate-1 nanoparticles in SBA-15 reveals equatorial and axial ligation of CO2 to the vanadyl-
sites. Adapted and reprinted with permission from [164]. Copyright 2011 American Chemical
Society
about the stability, H2O, CO2 and NH3 accessibility of the vanadium sites [163,
164] (Fig. 7) it also gave new insight into the formation process of vanadium
silicalites [163]. In order to obtain the silicalite nanoparticles, the zeolite growth
needs to be stopped by acidification of the solution. Although this has been known
for some time, it was never clear what triggers this process. Our results clearly
showed that acidification affects the stability of the tetrapropylammonium-
hydroxide template molecules such that one propyl ligand is lost by acid hydrolysis.
It was also shown that at this stage of the zeolite synthesis, the vanadyl species are
not fixed completely within the hydrophilic network. The true incorporation of the
vanadyl species in the zeolite only occurs when the solvated vanadyl species get
trapped between the aggregating nanoparticles during the hydrothermal growth of
the zeolite.
With the continuing presence of paramagnetic transition metal ions or defect sites
in homogeneous and heterogeneous catalysis, the important role played by
advanced EPR techniques to uncover the mechanistic details of the catalytic
cycle is clear and justified. The electronic structure and symmetry of the metal
centre are crucial elements of the catalytic active site. These properties can be easily
probed by EPR, whilst the additional resolving power of HF EPR, ENDOR,
ESEEM and HYSCORE offers further insights into the extended (ligand) structure
around the catalytic site. In many ways, when paramagnetic centres are involved,
these combined techniques provide an unparalleled glimpse into the actions of the
working catalyst. The coupling of the traditional CW-EPR techniques with the
more advanced methodologies that have been developed over the past 20 years
facilitates the improved characterisation and indeed the detection of paramagnetic
and radical intermediates of relevance to catalysis. Although this family of EPR
30 S. Van Doorslaer and D.M. Murphy
Acknowledgments DMM would like to thanks EPSRC for funding (EP/H023879/1). SVD
thanks the University of Antwerp for support (via NOI-BOF funding).
References
19. Van Doorslaer S, Caretti I, Fallis IA et al (2009) The power of electron paramagnetic
resonance to study asymmetric homogeneous catalysts based on transition-metal complexes.
Coord Chem Rev 253:21162130
20. Carter E, Murphy DM (2009) Structure function relationships and mechanistic pathways in
homogeneous enantioselective catalysis as probed by ENDOR spectroscopy. In: Douthwaite
R, Duckett S (eds) Spectroscopic properties of inorganic and organometallic compounds, vol
40. RSC, Cambridge
21. Mobius K, Savitsky A (2008) High field EPR spectroscopy on proteins and their model
systems. RSC, Cambridge
22. Goldfarb D (2006) High field ENDOR as a characterization tool for functional sites in
microporous materials. Phys Chem Chem Phys 8:23252343
23. Bennati M, Prisner TF (2005) New developments in high field electron paramagnetic
resonance with applications in structural biology. Rep Prog Phys 68:411448
24. Andersson KK, Schmidt PP, Katterle B et al (2003) Examples of high-frequency EPR studies
in bioinorganic chemistry. J Biol Inorg Chem 8:235247
25. Fortman GC, Kegl T, Li QS et al (2007) Spectroscopic detection and theoretical confirmation
of the role of Cr2(CO)5 (C5R5)2 and.Cr(CO)2(ketene)(C5R5) as intermediates in carbonyl-
ation of NNCHSiMe3 to OCCHSiMe3 by.Cr(CO)3 (C5R5) (R H, CH3). J Am
Chem Soc 129:1438814400
26. Krzystek J, Ozarowsk A, Telser J (2006) Multi-frequency, high-field EPR as a powerful tool
to accurately determine zero-field splitting in high-spin transition metal coordination
complexes. Coord Chem Rev 250:23082324
27. Bianchini C, Gatteschi D, Giambastiani G et al (2007) Electronic influence of the thienyl
sulfur atom on the oligomerization of ethylene by cobalt(II) 6-(thienyl)-2-(imino)pyridine
catalysis. Organometallics 26:726739
28. Schweiger A, Jeschke G (2001) Principles of pulse electron paramagnetic resonance. Oxford
University Press, Oxford
29. Zein S, Kulik LV, Yano J et al (2008) Focusing the view on natures water-splitting catalyst.
Philos Trans R Soc B 363:11671177
30. Hoffman BM (2003) Electron-nuclear double resonance spectroscopy (and electron spin-
echo envelope modulation spectroscopy) in bioinorganic chemistry. Proc Natl Acad Sci USA
100:35753578
31. Peloquin JM, Britt RD (2001) EPR/ENDOR characterization of the physical and electronic
structure of the OEC Mn cluster. Biochim Biophys Acta Bioenerg 1503:96111
32. Lubitz W, Reijerse E, Van Gastel M (2007) [NiFe] and [FeFe] hydrogenases studied by
advanced magnetic resonance techniques. Chem Rev 107:43314365
33. Lubitz W, Lendzian F, Bittl R (2002) Radicals, radical pairs and triplet states in photosyn-
thesis. Acc Chem Res 35:313320
34. Schiemann O, Prisner TF (2007) Long-range distance determinations in biomacromolecules
by EPR spectroscopy. Quart Rev Biophys 40:153
35. Prisner T, Rohrer M, MacMillan F (2001) Pulsed EPR spectroscopy: biological applications.
Annu Rev Phys Chem 52:270313
36. Rosati F, Roelfes G (2010) Artificial metalloenzymes. ChemCatChem 2:916927
37. Palucki M, Finney NS, Pospisil PJ et al (1998) The mechanistic basis for electronic effects on
enantioselectivity in the (salen)Mn(III)-catalyzed epoxidation reaction. J Am Chem Soc
120:948954
38. Katsuki T (1995) Catalytic asymmetric oxidations using optically active (salen)-manganese
(III) complexes as catalysts. Coord Chem Rev 140:189214
39. Schaus SE, Brandes BD, Larrow JF et al (2002) Highly selective hydrolytic kinetic resolution
of terminal epoxides catalyzed by chiral (salen)Co-III complexes. Practical synthesis of
enantioenriched terminal epoxides and 1,2-diols. J Am Chem Soc 124:13071315
40. Fallis IA, Murphy DM, Willock DJ et al (2004) Direct observation of enantiomer discrimina-
tion of epoxides by chiral salen complexes using ENDOR. J Am Chem Soc 126:1566015661
32 S. Van Doorslaer and D.M. Murphy
41. Murphy DM, Fallis IA, Willock DJ et al (2008) Discrimination of geometrical epoxide
isomers by ENDOR & DFT the role of H-bonds. Angew Chem Int Ed 47:14141416
42. Carter E, Murphy DM, Fallis IA et al (2010) Probing the role of weak outer sphere
interactions (H-bond) in VO(3,5-tBu2-salophen) epoxide adducts by EPR, ENDOR and
HYSCORE. Chem Phys Lett 486:7479
43. Murphy DM, Fallis IA, Landon J et al (2009) Enantioselective binding of structural epoxide
isomers by a chiral vanadyl salen complex: a pulsed EPR, CW-ENDOR and DFT investigation.
Phys Chem Chem Phys 11:67576769
44. Carter E, Fallis IA, Kariuki BM et al (2011) Structure and EPR characterization of N,N-bis
(5-tert-butylsalicylidene)-1,2-cyclohexanediamino-vanadium (IV) oxide and its adducts with
propylene oxide. Dalton Trans 40:74547462
45. Murphy DM, Caretti I, Carter E et al (2011) Visualising diastereomeric interactions of chiral
amine-chiral copper Salen adducts by EPR/ENDOR/HYSCORE spectroscopy and DFT.
Inorg Chem (submitted)
46. Merkx M, Kopp DA, Sazinsky MH et al (2001) Dioxygen activation and methane hydroxyl-
ation by soluble methane monooxygenase: a tale of two irons and three proteins. Angew
Chem Int Ed 40:27822807
47. Wallar BJ, Lipscomb JD (1996) Dioxygen activation by enzymes containing binuclear non-
heme iron clusters. Chem Rev 96:26252657
48. Pulver SC, Froland WA, Lipscomb JD et al (1997) Ligand field circular dichroism and
magnetic circular dichroism studies of component B and substrate binding to the hydroxylase
component of methane monooxygenase. J Am Chem Soc 119:387395
49. Wolfe MD, Parales JV, Gibson DT et al (2001) Single turnover chemistry and regulation of
O2 activation by the oxygenase component of naphthalene 1,2-dioxygenase. J Biol Chem
276:19451953
50. Karlsson A, Parales JV, Parales RE et al (2003) Crystal structure of naphthalene dioxygenase:
side-on binding of dioxygen to iron. Science 299:10391042
51. Ward TR (2011) Artificial metalloenzymes based on the biotin-avidin technology:
enantioselective catalysis and beyond. Acc Chem Res 44:4757
52. Park MJ, Lee J, Suh Y et al (2006) Reactivities of mononuclear non-heme iron intermediates
including evidence that iron(III) hydroperoxo species is a sluggish oxidant. J Am Chem Soc
128:26302634
53. Que L, Ho RYN (1996) Dioxygen activation by enzymes with mononuclear non-heme iron
active sites. Chem Rev 96:26072624
54. Costas M, Chen K, Que L (2000) Biomimetic nonheme iron catalysts for alkane hydroxyl-
ation. Coord Chem Rev 200202:517544
55. Katona G, Carpentier P, Niviere V et al (2007) Raman-assisted crystallography reveals end-
on peroxide intermediates in a nonheme iron enzyme. Science 316:449453
56. Dey A, Jenney FE, Adams MK et al (2007) Sulfur K-edge X-ray absorption spectroscopy and
density functional theory calculations on superoxide reductase: role of the axial thiolate in
reactivity. J Am Chem Soc 129:1241812431
57. Clay MD, Yang TC, Jenney FE et al (2006) Geometries and electronic structures of cyanide
adducts of the non-heme iron active site of superoxide reductases: vibrational and ENDOR
studies. Biochemistry 45:427438
58. Jiang YB, Telser J, Goldberg DP (2009) Evidence for the formation of a mononuclear ferric-
hydroperoxo complex via the reaction of dioxygen with an (N4S(thiolate))iron(II) complex.
Chem Commun 68286830
59. Girerd JJ, Banse F, Simaan AJ (2000) Characterization and properties of non-heme iron
peroxo complexes. In: Metal-oxo and metal peroxo species in catalytic oxidations. Structure
and Bonding, vol 97. Springer, Hidelberg, pp 145177
60. Simaan AJ, Banse F, Girerd JJ (2001) The electronic structure of non-heme iron(III)-
hydroperoxo and iron(III)-peroxo model complexes studied by Mossbauer and electron
paramagnetic resonance spectroscopies. Inorg Chem 40:65386540
EPR Spectroscopy in Catalysis 33
61. Berry JF, Bill E, Bothe E et al (2006) Octahedral non-heme oxo and non-oxo Fe(IV)
complexes: an experimental/theoretical comparison. J Am Chem Soc 128:1351513528
62. Bilis G, Christoforidis KC, Deligiannakis Y et al (2010) Hydrocarbon oxidation by homoge-
neous and heterogeneous non-heme iron(III) catalysts with H2O2. Catal Today 157:101106
63. Shi W, Liu Y, Liu B et al (2006) Synthesis and characterization of a six-coordinate
monomeric Mn(III) complex with SOD-like activity. J Coord Chem 59:119130
64. Parsell TH, Behan RK, Green MT et al (2006) Preparation and properties of a monomeric
MnIV-oxo complex. J Am Chem Soc 128:87288729
65. Donoghue PJ, Gupta AK, Boyce DW et al (2010) An anionic, tetragobal copper(II) superox-
ide complex. J Am Chem Soc 132:1586915871
66. Chaudhuri P, Wieghardt K (2001) Phenoxyl radical complexes. Prog Inorg Chem
50:151216
67. Jorgensen CK (1969) Oxidation numbers and oxidation states. Springer, Heidelberg
68. Jazdzewski BA, Tolma WB (2000) Understanding the copper-phenoxyl radical array in
galactose oxidase: contributions from synthetic modeling studies. Coord Chem Rev
200202:633685
69. Kruger H-J (1999) What can we learn from nature about the reactivity of coordinated
phenoxyl radicals? A bioinorganic success story. Angew Chem Int Ed 38:627631
70. Bereman RD, Kosman DJ (1977) Stereo-electronic properties of metalloenzymes 5. Identifi-
cation and assignment of ligand hyperfine splittings in electron spin resonance spectrum of
galactose oxidase. J Am Chem Soc 99:73227325
71. Muller J, Kikuchi A, Bill E et al (2000) Phenoxyl radical complexes of chromium(III),
manganese(III), cobalt(III) and nickel(II). Inorg Chimica Acta 297:265277
72. dos Anjos A, Bortoluzzi AJ, Osorio R et al (2005) New mononuclear CuII and ZnII complexes
capable of stabilizing phenoxyl radicals as models for the active form of galactose oxidase.
Inorg Chem Comm 8:249253
73. Mukherjee A, Lioret R, Mukherjee R (2008) Synthesis and properties of diphenoxo-bridged
CoII, NiII, CuII and ZnII complexes of a new tripodal ligand: generation and properties of MII-
coordinated phenoxyl radical species. Inorg Chem 47:44714480
74. Thomas F, Arora H, Philouze C et al (2010) CoIII and CuII complexes of reduced Schiff bases:
generation of phenoxyl radical species. Inorg Chimica Acta 363:31223130
75. Chaudhuri P, Verani CN, Bill E et al (2001) Electronic structure of bis(o-iminobenzo-
semiquinonato)metal complexes (Cu, Ni, Pd). The art of establishing physical oxidation
states in transition metal complexes containing radical ligands. J Am Chem Soc
123:22132223
76. Weyhermuller T, Paine TK, Bothe E et al (2002) Complexes of an aminebis(phenolate) [O,
N, O] donor ligand and EPR studies of isoelectronic, isostructural Cr(III) and Mn(IV)
complexes. Inorg Chimica Acta 337:344356
77. Telser J (2010) Overview of ligand versus metal centred redox reactions in tetraaza macro-
cyclic complexes of nickel with a focus on electron paramagnetic resonance. J Braz Chem
Soc 21:11391157
78. Shimazaki Y, Tani F, Fukui K et al (2003) One-electron oxidized nickel(II)-(disalicylidene)
diamine complex: temperature-dependent tautomerism between Ni(III)-phenolate and Ni(II)-
phenoxyl radical states. J Am Chem Soc 125:1051210513
79. Shimazaki Y, Yajima T, Tani F et al (2007) Synthesis and electronic structures of one-
electron-oxidized group 10 metal(II) (disalicylidene)diamine complexes (metal Ni, Pd,
Pt). J Am Chem Soc 129:25592568
80. Rotthaus O, Jarjayes O, Del Valle CP et al (2007) A versatile electronic hole in one-electron
oxidized Ni-II bis-salicylidene phenylenediamine complexes. Chem Commun 44624464
81. Rotthaus O, Labet V, Philouze C et al (2008) Pseudo-octahedral schiff base nickel(II)
complexes: does single oxidation always lead to the nickel(III) valence tautomer? Eur J
Inorg Chem 42154224
34 S. Van Doorslaer and D.M. Murphy
82. Storr T, Wasinger EC, Pratt RC et al (2007) The geometric and electronic structure of a one-
electron-oxidized nickel(II) bis(salicylidene)diamine complex. Angew Chem Int Ed
46:51985201
83. Storr T, Verma P, Pratt RC et al (2008) Defining the electronic and geometric structure of
one-electron oxidized copper-bis-phenoxide complexes. J Am Chem Soc 46:1544815459
84. Pratt RC, Stack TDP (2005) Mechanistic insights from reactions between copper(II)-
phenoxyl complexes and substrates with activated C-H bonds. Inorg Chem 44:23672375
85. Vinck E, Murphy DM, Fallis IA et al (2010) Formation of a cobalt(III)-phenoxyl radical
complex by acetic acid promoted aerobic oxidation of a Co(II)salen complex. Inorg Chem
49:20832092
86. Vinck E, Van Doorslaer S, Murphy DM et al (2008) The electronic structure of N,N0 -bis(3,5-
di-tert-butylsalicylidene)-1,2-cyclohexane-diamino cobalt(II). Chem Phys Lett 464:3137
87. Vinck E, Van Doorslaer S, Murphy DM et al (2011) (in prep)
88. Vinck E, Murphy DM, Fallis IA et al (2010) A pulsed EPR and DFT investigation of the
stabilization of coordinated phenoxyl radicals in a series of cobalt Schiff-base complexes.
Appl Magn Reson 37:289303
89. Orio M, Jarjayes O, Philouze C et al (2010) Spin interaction in octahedral zinc complexes of
mono- and diradical Schiff and Mannich bases. Inorg Chem 49:646658
90. Buttner T, Geier J, Frison G et al (2005) A stable aminyl radical metal complex. Science
307:235238
91. Dixon JT, Green MJ, Hess FM et al (2004) Advances in selective ethylene trimerisation a
critical overview. J Organomet Chem 689:36413668
92. Emrich R, Heinemann O, Jolly PW et al (1997) The role of metallacycles in the chromium-
catalyzed trimerization of ethylene. Organometallics 16:15111513
93. Manyik RM, Walker WE, Wilson TP (1997) Soluble chromium based catalysts for ethylene
trimerization and polymerization. J Catal 47:197209
94. Kohn RD, Haufe M, Mihan S et al (2000) Triazacyclohexane complexes of chromium as
highly active homogeneous model systems for the Phillips catalyst. Chem Commun
19271928
95. Kohn R (2008) Reactivity of chromium complexes under spin control. Angew Chem Int Ed
47:245247
96. Moulin JO, Evans J, McGuinness DS et al (2008) Probing the effects of ligand structure on
activity and selectivity of Cr(III) complexes for ethylene oligomerisation and polymerization.
Dalton Trans 11771185
97. Bruckner A, Jabor JK, McConnell AEC et al (2008) Monitoring structure and valence state of
chromium sites during catalyst formation and ethylene oligomerization by in situ EPR
spectroscopy. Organometallics 27:38493856
98. Skobelev IY, Panchenko VN, Lyakin OY et al (2010) In situ EPR monitoring of chromium
species formed during Cr-pyrrolyl ethylene trimerization catalyst formation. Organometallics
29:29432950
99. McDyre LE, Hamilton T, Murphy DM et al (2010) A CW EPR and ENDOR investigation on a
series of Cr(I) carbonyl complexes with relevance to alkene oligomerization catalysis: [Cr
(CO)4L]+ (L Ph2PN(R)PPh2, Ph2P(R)PPh2, Ph2PN(R)NPPh2). Dalton Trans 39:77927799
100. McDyre LE, Carter E, Murphy DM et al (2011) (in prep)
101. Bianchini G, Giambastiani IG, Rios A et al (2007) Synthesis of a new polydentate ligand
obtained by coupling 2,6-bis(amino)pyridine and (imino)pyridine moieties and its use in
ethylene oligomerisation in conjunction with iron(II) and cobalt(II) bis-halides.
Organometallics 26:50665078
102. Soshnikov IE, Semikolenova NV, Bushmelev AN et al (2009) Investigating the nature
of the active species in bis(imino)pyridine cobalt ethylene polymerisation catalysts.
Organometallics 28:60036013
103. Britovsek GJP, Clentsmith GKB, Gibson VC et al (2002) The nature of the active site in bis
(imino)pyridine iron ethylene polymerisation catalysts. Catal Commun 3:207211
EPR Spectroscopy in Catalysis 35
104. Dzik WI, Smits JMM, Reek JNH et al (2009) Activation of carbon monoxide by (Me(3)tpa)
Rh and (Me(3)tpa)Ir. Organometallics 28:16311643
105. Goldberg H, Kaminker I, Goldfarb D et al (2009) Oxidation of carbon monoxide cocatalyzed
by palladium(0) and the H5PV2Mo10O40 polyoxometalate probed by electron paramagnetic
resonance and aerobic catalysis. Inorg Chem 48:79477952
106. Khenkin AM, Efremenko I, Weiner L et al (2010) Photochemical reduction of carbon dioxide
catalyzed by a ruthenium substituted polyoxometalate. Chem A Eur J 16:13561364
107. Che C-M, Huang J-S (2003) Metal complexes of chiral binaphthyl Schiff-base ligands and
their application in stereoselective organic transformations. Coord Chem Rev 242:97113
108. Shitama H, Katsuki T (2007) Synthesis of metal-(pentadentate-salen) complexes: asymmetric
epoxidation with aqueous hydrogen peroxide and asymmetric cyclopropanation (salenH(2):
N,N-bis(salicylidene)ethylene-1,2-diamine). Chem Eur J 13:48494858
109. Groves JT, Stern MK (1988) Synthesis, characterization, and reactivity of oxomanganese(IV)
porphyrin complexes. J Am Chem Soc 110:86288638
110. Ostovic D, Bruice TC (1992) Mechanism of alkene epoxidation by iron, chromium and
manganese higher valent oxo-metalloporphyrins. Acc Chem Res 25:314320
111. Collman JP, Zhang X, Lee VJ et al (1993) Regioselective and enantioselective epoxidation
catalyzed by metalloporphyrins. Science 261:14041411
112. Murphy A, Dubois G, Stack TDP (2003) Efficient epoxidation of electron-deficient olefins
with a cationic manganese complex. J Am Chem Soc 125:52505251
113. Murphy A, Pace A, Stack TDP (2004) Ligand and pH influence on manganese-mediated
peracetic acid epoxidation of terminal olefins. Org Lett 6:31193122
114. de Boer JW, Browne WR, Brinksma J et al (2007) Mechanism of cis-dihydroxylation and
epoxidation of alkenes by highly H2O2 efficient dinuclear manganese catalysts. Inorg Chem
46:63536372
115. Bryliakov KP, Lobanova MV, Talsi EP (2002) EPR and 1H NMR spectroscopic study of the
CrIII(salen)Cl catalysts. J Chem Soc Dalton Trans 22632265
116. Bryliakov KP, Talsi EP (2003) CrIII(salen)Cl catalyzed asymmetric epoxidations: insight into
the catalytic cycle. Inorg Chem 42:72587265
117. Campbell KA, Lashley MR, Wyatt JK (2001) Dual-mode EPR study of Mn(III) salen
and the Mn(III) salen-catalyzed epoxidation of cis-beta-methylstyrene. J Am Chem Soc
123:57105719
118. Scarpellini M, Casellato A, Bortoluzzi AJ et al (2006) EPR and semi-empirical studies as
tools to assign the geometric structures of Fe-III isomer models for transferrins. J Braz Chem
Soc 17:16171626
119. Dyers L, Que SY, Van Derveer D et al (2006) Synthesis and structures of new salen
complexes with bulky groups. Inorg Chimica Acta 359:197203
120. Tani F, Matsu-ura M, Nakayama S et al (2001) Synthesis and characterization of
alkanethiolate-coordinated iron porphyrins and their dioxygen adducts as models for the
active center of cytochrome P450: direct evidence for hydrogen bonding to bound dioxygen.
J Am Chem Soc 123:11331142
121. Bryliakov KP, Talsi EP (2004) Evidence for the formation of an iodosylbenzene(salen)iron
active intermediate in a (salen)iron (III)-catalyzed asymmetric sulfide oxidation. Angew
Chem Int Ed 43:52285230
122. Sivasubramanian VK, Ganesan M, Rajagopal S et al (2002) Iron(III)-salen complexes as
enzyme models: mechanistic study of oxo(salen)iron complexes oxygenation of organic
sulfides. J Org Chem 67:15061514
123. Ottenbacher RV, Bryliakov KP, Talsi EP (2010) Nonheme manganese-catalyzed asymmetric
oxidation. A Lewis acid activation versus oxygen rebound mechanism: evidence for the third
oxidant. Inorg Chem 49:86208628
124. Duban EA, Bryliakov KP, Talsi EP (2007) The active intermediates of non-heme-iron-based
systems for catalytic alkene epoxidation with H2O2/CH3COOH. Eur J Inorg Chem 852857
36 S. Van Doorslaer and D.M. Murphy
125. Lyakin OY, Bryliakov KP, Britovsek GJP et al (2009) EPR spectroscopic trapping of the
active species of nonheme iron-catalyzed oxidation. J Am Chem Soc 131:1079810799
126. Maurya MR, Kumar M, Pessoa JC (2008) Oxidation of p-chlorotoluene and cyclohexene
catalyzed by polymer-anchored oxovanadium(IV) and copper(II) complexes of amino acid
derived tridentate ligands. Dalton Trans 42204232
127. Mimmi MC, Gullotti M, Santagostini L (2004) Models for biological trinuclear copper
clusters. Characterization and enantioselective catalytic oxidation of catechols by the cop-
per(I) complexes of a chiral ligand derived from (S)-()-1,10 binaphthyl-2,20 -diamine. Dalton
Trans 21922201
128. Adao P, Maurya MR, Kumar U et al (2009) Vanadium-salen and -salan complexes: charac-
terization and application in oxygen-transfer reactions. Pure Appl Chem 81:12791296
129. Andino JG, Kilgore UJ, Pink M et al (2010) Intermolecular C-H bond activation of benzene
and pyridines by a vanadium(III) alkylidene including a stepwise conversion of benzene
to a vanadium-benzyne complex. Chem Sci 1:351356
130. Ye S, Neese F, Ozarowski A et al (2010) Family of V(III)-tristhiolato complexes relevant to
functional models of vanadium nitrogenase: synthesis and electronic structure investigations
by means of high-frequency and -field electron paramagnetic resonance coupled to quantum
chemical computations. Inorg Chem 49:977988
131. Krzystek J, Fiedler AT, Sokol JJ et al (2004) Pseudooctahedral complexes of vanadium(III):
electronic structure investigation by magnetic and electronic spectroscopy. Inorg Chem
43:56455658
132. Bolm C, Martin M, Gescheidt G et al (2003) Spectroscopic investigation of bis(sulfoximine)
copper(II) complexes and their relevance in asymmetric catalysis. J Am Chem Soc
125:62226227
133. Bolm C, Martin M, Gescheidt G et al (2007) Mechanistic insights into stereoselective
catalysis the effects of counterions in a CuII-bissulfoximine-catalyzed Diels-Alder reaction.
Chem Eur J 13:18421850
134. Tatsumi T (2004) Zeolites: catalysis. In: Atwood JL, Steed JW (eds) Encyclopedia of
supramolecular chemistry, vol 2. Marcel Dekker, New York, pp 16101616
135. Cejka J (2004) Zeolites: structures and inclusion properites. In: Atwood JL, Steed JW (eds)
Encyclopedia of supramolecular chemistry, vol 2. Marcel Dekker, New York, pp 16231630
136. Arieli D, Delabie A, Vaughan DEW et al (2002) Isomorphous substitution of Mn(II) into
aluminophosphate zeotypes: a combined high-field ENDOR and DFT study. J Phys Chem B
106:75097519
137. Arieli D, Prisner TF, Hertel M et al (2004) Resolving Mn framework sites in large cage
aluminophosphate zeotypes by high field EPR and ENDOR spectroscopy. Phys Chem Chem
Phys 6:172181
138. Arieli D, Delabie A, Groothaert M et al (2002) The process of Mn(II) incorporation into
aluminophosphate zeotypes through high-field ENDOR spectroscopy and DFT calculations.
J Phys Chem B 106:90869097
139. Carl PJ, Vaughan DEW, Goldfarb D (2002) Interactions of Cu(II) ions with framework Al in
high Si:Al zeolite Y as determined from X- and W-band pulsed EPR/ENDOR spectroscopies.
J Phys Chem B 106:54285437
140. Maurelli S, Ruszak M, Witkowski S et al (2010) Spectroscopic CW-EPR and HYSCORE
investigations of Cu2+ and O2 species in copper doped nanoporous calcium aluminate
(12CaO.7Al2O3). Phys Chem Chem Phys 12:1093310941
141. Woodworth J, Bowman MK, Larsen SC (2004) Two-dimensional pulsed EPR studies of
vanadium-exchanged ZSM-5. J Phys Chem B 108:1612816134
142. Giamello E, Murphy D, Magnacca G et al (1992) Interaction of NO with copper ions in
ZSM5. An EPR and IR investigation. J Catal 136:510520
143. Umamaheswari V, Hartmann M, P oppl A (2005) EPR spectroscopy of Cu(I)-NO adsorption
complexes formed over Cu-ZSM-5 and Cu-MCM-22 zeolites. J Phys Chem B
109:15371546
EPR Spectroscopy in Catalysis 37
144. Umamaheswari V, Hartmann M, P oppl A (2005) Pulsed ENDOR study of Cu(I)-NO adsorption
complexes in Cu-L zeolite. J Phys Chem B 109:1084210848
145. Umamaheswari V, Hartmann M, P oppl A (2005) Critical assessment of electron spin
resonance studies on Cu(I)-NO complexes in Cu-ZSM-5 zeolites prepared by solid- and
liquid-state ion exchange. J Phys Chem B 109:1972319731
146. Pietrzyk P, Piskorz W, Sojka Z et al (2003) Molecular structure, spin density and hyperfine
coupling constants of the Z1{CuNO}11 adduct in the ZSM-5 zeolite:DFT calculations and
comparison with EPR data. J Phys Chem B 107:61056113
147. Pal C, Wheatley PS, El Mkami H et al (2010) EPR on medically relevant NO adsorbed to
Zn-LTA. Appl Magn Reson 37:619627
148. Soler-Illia J, Sanchez C, Lebeau B et al (2002) Chemical strategies to design textured
materials: from microporous and mesoporous oxides to nanonetworks and hierarchical
structures. Chem Rev 102:40934138
149. Kresge CT, Leonowicz ME, Roth WJ et al (1992) Ordered mesoporous molecular sieves
synthesized by a liquid-crystal template mechanism. Nature 359:710712
150. Beck JS, Vartuli JC, Roth WJ et al (1992) A new family of mesoporous molecular sieves
prepared with liquid crystal templates. J Am Chem Soc 114:1083410843
151. Zhao D, Feng J, Huo Q et al (1998) Triblock copolymer syntheses of mesoporous silica
with periodic 50 to 300 Angstrom pores. Science 279:548552
152. Kruk M, Jaroniec M, Ko CH et al (2000) Characterization of the porous structure of SBA-15.
Chem Mater 12:19611968
153. Bennadja Y, Beaunier P, Margolese D et al (2001) Fine tuning of the interaction between
pluronic surfactants and silica walls in SBA-15 nanostructured materials. Microporous
Mesoporous Mater 4445:147152
154. Ruthstein S, Goldfarb D (2008) An EPR tool box for exploring the formation and properties
of ordered template mesoporous materials. Electron Paramagn Reson 21:184215
155. Zhang J, Carl PJ, Zimmermann H et al (2002) Investigation of the formation of MCM-41 by
electron spin-echo envelope modulation spectroscopy. J Phys Chem B 106:53825389
156. Ruthstein S, Frydman V, Kababya S et al (2003) Study of the formation of the mesoporous
material SBA-15 by EPR spectroscopy. J Phys Chem B 107:17391748
157. Ruthstein S, Frydman V, Goldfarb D (2004) Study of the initial formation stages of the
mesoporous material SBA-15 using spin-labeled block co-polymer templates. J Phys Chem B
108:90169022
158. Ruthstein S, Goldfarb D (2008) Evolution of solution structures during the formation of the
cubic mesoporous material, KIT-6, determined by double electron-electron resonance. J Phys
Chem C 112:71027109
159. Zamani S, Meynen V, Hanu AM et al (2009) Direct spectroscopic detection of framework-
incorporated vanadium in mesoporous silica materials. Phys Chem Chem Phys
11:58235832
160. Van Doorslaer S, Segura Y, Cool P (2004) Structural investigation of vanadyl-
acetylacetonate-containing precursors of TiOx-VOx mixed oxides on SBA-15. J Phys Chem
B 108:1940419412
161. Van Der Voort P, Ravikovitch PI, De Jong KP et al (2002) Plugged hexagonal templated
silica: a unique micro- and mesoporous composite material with internal silica nanocapsules.
Chem Commun 10101011
162. Meynen V, Beyers E, Cool P et al (2004) Post-synthesis deposition of V-zeolitic
nanoparticles in SBA-15. Chem Commun 898899
163. Chiesa M, Meynen V, Van Doorslaer S et al (2006) Vanadium silicalite-1 nanoparticles
deposition onto mesoporous walls of SBA-15. Mechanistic insights from a combined EPR
and Raman study. J Am Chem Soc 128:89558963
164. Zamani S, Chiesa M, Meynen V et al (2010) Accessibility and dispersion of vanadyl sites
of vanadium silicate-1. J Phys Chem C 114:1296612975
38 S. Van Doorslaer and D.M. Murphy
165. Carp O, Huisman CL, Reller A (2004) Photoinduced reactivity of titanium dioxide. Prog
Solid State Chem 32:33177
166. Ayoub K, van Hullebusch ED, Cassir M et al (2010) Application of advanced oxidation
processes for TNT removal: a review. J Hazard Mater 178:1028
167. Pascual J, Camassel J, Mathieu H (1977) Resolved quadrupole transition in TiO2. Phys Rev
Lett 39:14901493
168. Asahi R, Morikawa T, Ohwak T et al (2001) Visible-light photocatalysis in nitrogen-doped
titanium oxides. Science 293:269271
169. Du GH, Chen Q, Peng LM (2001) Preparation and structure analysis of titanium oxide
nanotubes. Appl Phys Lett 79:37023704
170. Naccache C, Meriaude P, Che M et al (1971) Identification of oxygen species adsorbed on
reduced titanium dioxide. Trans Faraday Soc 67:506512
171. Hurum DC, Agrios AG, Crist SE et al (2006) Probing reaction mechanisms in mixed phase
TiO2 by EPR. J Electron Spectrosc Relat Phenomena 150:155163
172. Micic OI, Zhang YN, Cromack KR et al (1993) Trapped holes on TiO2 colloids studied by
electron paramagnetic resonance. J Phys Chem 97:72777283
173. Berger T, Sterrer M, Diwals O et al (2005) Charge trapping and photoadsorption of O2
on dehydroxylated TiO2 nanocrystals an electron paramagnetic resonance study.
ChemPhysChem 6:21042112
174. Serpone N (2006) Is the band gap of pristine TiO2 narrowed by anion- and cation-doping of
titanium dioxide in second generation photocatalysts? J Phys Chem B 110:2428724293
175. Livraghi S, Paganini MC, Giamello E et al (2006) Origin of photoactivity of nitrogen-doped
titanium dioxide under visible light. J Am Chem Soc 128:1566615671
176. Fittipaldi M, Gombac V, Montini T et al (2008) A high-frequency (95 GHz) electron
paramagnetic resonance study of B-doped TiO2 photocatalysts. Inorganica Chim Acta
361:39803987
177. Riss A, Berger T, Stankic S et al (2008) Charge separation in layered titanate nanostructures:
effect of ion exchange induced morphology transformation. Angew Chem Int Ed
47:14961499
178. Berger T, Sterrer M, Diwald O et al (2005) Light-induced charge separation in anatase TiO2
particles. J Phys Chem B 109:60616068
179. Kumar CP, Gopal NO, Wang TC et al (2006) EPR investigation of TiO2 nanoparticles with
temperature-dependent properties. J Phys Chem B 110:52235229
180. Ribbens S, Caretti I, Beyers E et al (2011) Unraveling the photocatalytic activity of
multiwalled hydrogen trititanate and mixed-phase anatase/trititanate nanotubes: a combined
catalytic and EPR study. J Phys Chem C 115:23022313
181. Green J, Carter E, Murphy DM (2009) Interaction of molecular oxygen with oxygen
vacancies on reduced TiO2: site specific blocking by probe molecules. Chem Phys Lett
477:340344
182. Riss A, Elser MJ, Bernardi J et al (2009) Stability and photoelectronic properties of layered
titanate nanostructures. J Am Chem Soc 131:61986206
183. Pattier B, Henderson M, P oppl A et al (2010) Multi-approach electron paramagnetic reso-
nance investigations of UV-photoinduced Ti3+ in titanium oxide-based gels. J Phys Chem B
114:44244431
184. Maurelli S, Livraghi S, Chiesa M et al (2011) Hydration structure of the Ti(III) cation as
revealed by pulse EPR and DFT studies: new insights into a textbook case. Inorg Chem
50:23852394
185. Hertzsch T, Hulliger J, Weber E et al (2004) Organic zeolites. In: Atwood JL, Steed JW (eds)
Encyclopedia of supramolecular chemistry, vol 2. Marcel Dekker, New York, pp 9961004
186. Meilikhov M, Yusenko K, Torrisi A et al (2010) Reduction of a metal-organic framework by
an organometallic complex: magnetic properties and structure of the inclusion compound
[(Z5-C5H5)2Co]0.5@MIL-47(V). Angew Chem Int Ed 49:62126215
EPR Spectroscopy in Catalysis 39
187. Mendt M, Jee B, Stock N et al (2010) Structural phase transitions and thermal hysteresis in
the metal-organic framework compound MIL-53 as studied by electron spin resonance
spectroscopy. J Phys Chem C 114:1944319451
188. Poppl A, Kunz S, Himsl D et al (2008) CW and pulsed ESR spectroscopy of cupric ions in the
metal-organic framework compound Cu3(BTC)2. J Phys Chem C 112:26782684
189. Jee B, Eisinger K, Gul-E-Noor F et al (2010) Continuous wave and pulsed electron spin
resonance spectroscopy of paramagnetic framework cupric ion in the Zn(II) doped porous
coordination polymer Cu3-xZnx(btc)2. J Phys Chem C 114:1663016639
Top Curr Chem (2012) 321: 4166
DOI: 10.1007/128_2011_301
# Springer-Verlag Berlin Heidelberg 2011
Published online: 19 November 2011
Radicals in Flavoproteins
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2 EPR Investigations of Flavoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3 ENDOR Investigations of Flavoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.1 ENDOR Investigations of (64) Photolyase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2 ENDOR Investigations of LOV Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4 Transient EPR Investigations of Flavoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.1 TREPR Studies of Reactive Paramagnetic Intermediates in Cryptochrome . . . . . . . . . 55
4.2 TREPR Studies of BLUF Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1 Introduction
During the past two decades an increasing number of enzymatic reactions were
found to proceed via mechanisms involving radicals as intermediate states (see,
e.g., [15]). Hence, a paramagnetic transient state is formed for a diamagnetic
resting state of an enzyme. Radical reactions in enzymology are most often
associated with redox-active cofactors, which can be divided into three subgroups.
(1) Enzymes carrying a transition metal, such as iron or copper, are often involved
in one-electron transfer processes. However, it has to be noted that, depending on
the type and redox state of the metal ion, the resting state of the metalloenzyme
could be intrinsically paramagnetic. (2) A few enzymatic reactions that involve
molecular oxygen as one substrate proceed without cofactor involvement if
the other substrate possesses appropriate chemical properties [6]. (3) Most radical
reactions in enzymology utilize redox-active organic cofactors, such as the hetero-
aromatic heme group, flavins, or quinones. To comprehend the reactivity of these
enzymes on a molecular level, it is necessary to localize and characterize the
paramagnetic center and its close vicinity, and to collect chemical and kinetic
information on reaction intermediates. This also includes electronic and geometric
information of traits that control the reactivity of enzymes. Protein systems are
usually quite complex; hence, the complementary use of various molecular and
structure determining methods is mandatory to obtain such details. However, for
the characterization of the radical state itself, EPR spectroscopy is usually the
method of choice that provides the most reliable and detailed information. To
illustrate how various EPR techniques can be used for this purpose, we focus in
this review on proteins carrying the flavin cofactor, which is the most ubiquitous
redox-active organic coenzyme.
Flavins (Fl) catalyze many different bioreactions of physiological importance
[79]. Riboflavin, flavin mononucleotide (FMN), and flavin adenine dinucleotide
(FAD) have the 7,8-dimethyl isoalloxazine ring in common but differ in the side
chain attached to N10. With their five redox states, fully oxidized, one-electron
reduced semiquinoid (FlH and Fl), and fully reduced hydroquinone (FlH2 and
FlH), flavins are involved in one-electron and two-electron transfer reactions [10].
Radicals in Flavoproteins 43
Flavin semiquinones were one of the first enzymatic radicals identified in electron-
transfer reactions of flavoproteins and reactions of dihydroflavins with molecular
oxygen [11]. Flavin semiquinones function in terminal electron transport complexes
by facilitating electron transfer between obligatory two-electron-reducing agents
(e.g., NADH), and one-electron acceptors (e.g., iron sulfide centers). These pro-
cesses are often further mediated by quinone cofactors such as the coenzyme Q [1].
Moreover, flavin semiquinones play important roles as intermediates in many light-
activated processes ranging from blue-light photoreception [1216] and magnetor-
eception [17, 18] to DNA photorepair [1921].
This review is divided into the following sections. First, EPR measurements of
g-tensors of flavoproteins and the modulation of the principal values of g by the
protein surroundings of the cofactor are discussed. Then, two recent examples of
application of pulsed ENDOR spectroscopy will be reviewed, and, finally, time-
resolved EPR spectroscopy, that is most favorably used to study photo-excited
triplet states and radical pairs, will be introduced.
different flavin radical species, namely neutral, anionic, and 5-methionine flavin
radicals, are depicted and the ranges of typical g-principal values are shown.
For protein-bound flavin radicals it turned out that the g-tensor reflects the
overall electronic structure on the redox-active isoalloxazine ring, and thus is
potentially a valuable probe by which chemically different flavin radicals
(e.g., non-covalently vs covalently bound at specific isoalloxazine ring positions,
and neutral radical vs anion radical) may be distinguished [32]. This is because the
g-principal values of non-covalently bound flavins seem rather unaffected towards
changes in the isoalloxazines local surroundings, unlike other organic cofactors
(such as quinones, tyrosines, or tryptophans) or the nitroxides frequently used in
spin-label EPR studies [37], which show broader distributions of gX-, gY-, and gZ-
principal values, depending, e.g., on the polarity and/or the hydrogen-bonding
situation of the cofactor binding pocket. Flavin neutral radicals in quite dissimilar
protein surroundings, however, render g-principal values that are quite robust,
and hence the subtle differences of gi, i {X, Y, Z}, in different flavoproteins, can
only be observed by EPR performed in very strong magnetic fields.
So far, no single-crystal EPR studies of flavin radicals have been reported.
However, the orientations of the g-principal axes relative to the molecular frame
of the flavins isoalloxazine moiety have been derived from orientation-selection
effects of the quite anisotropic hyperfine coupling of H5 (or D5 in an isotope-
exchange experiment) of the neutral flavin radical, both with EPR [28] and ENDOR
Radicals in Flavoproteins 45
around 12 MHz and 17 MHz. (5) The broad, rhombic (Ax 6 Ay 6 Az) feature
extending from 21 to 34 MHz in the pulsed ENDOR spectrum is assigned to the
proton bound to N5 [24, 25, 38]. Its contribution to the overall spectrum is easily
discriminated from that of other protons in the isoalloxazine moiety due to the
exchangeability of H5 with a deuteron upon buffer deuteration. Observation of this
very anisotropic hyperfine coupling beautifully demonstrates the advantages of
pulsed ENDOR over the conventional continuous-wave methodology. In the latter
the first derivative of the signal intensity (with respect to the radio frequency) is
recorded, which becomes very small when broad spectral features are to be measured.
Hence, such couplings often escape direct detection in continuous-wave
ENDOR [25].
A flavin anion radical shows a markedly different proton ENDOR spectrum as
compared to that of a neutral radical. The most pronounced differences are, of
course, the absence of the signal from H5 and the larger splittings of the signal pairs
arising from H8a and H6 in the anion radical case. Hence, in addition to the
g-tensor, the hyperfine pattern of a flavin radical allows for an unambiguous
discrimination of the radicals protonation state [33, 34].
With the commercial availability of pulsed EPR instrumentation, other pulsed
methods such as electron-spin echo envelope modulation (ESEEM) or hyperfine
sublevel correlation spectroscopy (HYSCORE), which are quite useful to study
specific hyperfine and quadrupolar couplings, have also been applied to
flavoproteins [59, 61, 62]. These studies have been reviewed recently, e.g., in [46].
Pulsed ENDOR has been favorably applied to characterize the electronic structure
of the FADH cofactor and its surroundings in (64) photolyase. Photolyases are
DNA repair enzymes that restore damaged and potentially lethal DNA by splitting
UV-generated cyclobutane pyrimidine dimers (CPDs) or (64) photoproducts in a
light-activated electron-transfer reaction, thereby preventing erratic DNA replica-
tion [19, 20]. For CPD-repairing CPD photolyase, the proposed repair mechanism
includes a photo-induced single electron-transfer step from the fully reduced FAD
cofactor (FADH) to the CPD, resulting in the formation of a CPD anion radical and
a neutral FADH radical [63]. The cyclobutane ring of the putative CPD radical then
splits, and subsequently the electron is likely to be transferred back to the FADH
radical, thus restoring the initial redox states [19]. Hence, the entire process
represents a true catalytic cycle with net-zero exchanged electrons. In contrast,
(64) photolyases are unable to restore the original bases from DNA containing
(64) photoproduct lesions in a single reaction step; rather, following binding of the
DNA lesion, the overall repair reaction consists of at least two different steps, one
of which could be light-independent while the other must be light-dependent
[6466]. Hitomi and coworkers first proposed a detailed reaction mechanism
based on a mutational study, model geometries calculated on the basis of previously
48 E. Schleicher and S. Weber
published CPD photolyase coordinates [67], and the important finding that the
repair rate of (64) photolyases strongly depends on the pH [68, 69]. In the initial
light-independent step, a 6-iminium ion intermediate is generated from the (64)
photoproduct aided by two highly conserved histidines (His354 and His358 in
Xenopus laevis (64) photolyase). The 6-iminium ion then spontaneously
rearranges to an oxetane intermediate by intramolecular nucleophilic attack [66].
The oxetane species was proposed earlier in analogy to the repair mechanism of
CPD photolyases, and because it was identified as an intermediate in the formation
of (64) photoproducts [64, 70]. This putative repair mechanism of (64)
photolyase requires one histidine to act as a proton acceptor and the other as a
proton donor, which implies that the two histidines should have markedly different
pKa values. The subsequent blue-light-driven (350 < l < 500 nm) reaction splits
the oxetane intermediate presumably via an electron-transfer mechanism similar to
that of CPD photolyases.
The unanswered question regarding the detailed repair mechanism of (64)
photolyases and the involvement of functionally relevant amino acids led to the
design of an ENDOR study [56] which will be briefly reviewed here. In general, as
the function of a histidine is markedly influenced by its protonation state, it seems
likely that the histidines at the solvent-exposed active site cause the unusual pH
dependence in the (64) photolyase repair activity in vitro [68]. The principal idea
was that the protonation of a histidine alters its polarity, which may be indirectly
probed by proton-ENDOR spectroscopy using the neutral radical state of the FAD
cofactor as the observer molecule. Figure 3 depicts sections of the complete
ENDOR spectra (in the radio-frequency region between 17.8 and 21.2 MHz
corresponding to hyperfine couplings between about 6 and 13 MHz) of X. laevis
(64) photolyase, where the three H8a and one of the two H1 protons resonate;
compare Fig. 2.
It is apparent that the intensity of the H8a ENDOR signal changes significantly
as a function of pH. In contrast, the resonances of the other protons are pH
independent (data not shown) [56]. For a detailed data analysis it was taken into
account that the signal of H8a overlaps with that arising from H1; by spectral
simulation, the individual signal contributions of these protons could be
deconvoluted. The overall shapes of the ENDOR spectra of the H358A mutant
protein largely resemble those of the wild type at the respective pH conditions. In
contrast to the wild type (or H358A), the ENDOR spectra of the H354A protein are
markedly different. Thus, replacement of His354 with alanine leads to significant
modifications of the cofactor-binding site at the 8a-methyl group and at the linkage
of the ribityl side chain. Hence, as a first result, structural information regarding the
distance and the location of the two histidines with respect to the flavin observer
was obtained: the strong shift of the isotropic hyperfine coupling of H1 in H354A,
as compared to the wild type or the H358A protein, observed at all measured pH
values, suggested that His354 is close to H1 (it has to be mentioned that experi-
mental data on the structure of a (64)-photolyase enzyme was unavailable at the
time when the ENDOR experiments were performed). A slight geometrical reori-
entation due to the histidine-to-alanine replacement results in an altered direction of
Radicals in Flavoproteins 49
Fig. 3 ENDOR experiments on wild type, H354A and H358A X. laevis (64) photolyase
measured at pH 6.0 and 9.5, respectively. Upper part: experimental (dots) and simulated (dashed
line) pulsed ENDOR spectra of different X. laevis (64) photolyase samples. The red and the blue
curves show the contributions of the H8a and the H1 hyperfine couplings to the overall ENDOR
spectrum [56]. Lower part: three-dimensional structure of the active site of (64) photolyase
including selected amino acids [71]. Proposed changes in the protonation state of His354 and
His358 upon pH variations are displayed
the C1H1 bond with respect to the p-plane of the isoalloxazine ring, thus
changing the dihedral angle, and, hence, the strength of the H1 hyperfine coupling.
On the other hand, the shift of the isotropic hyperfine coupling of H8a, with respect
to the wild type, was larger in the H358A mutant than in H354A. From this finding
it was concluded, that His358 is closer than His354 to the H8a protons.
As a second result of the ENDOR study, the H8a ENDOR signal in the H354A
mutant was also demonstrated to be strongly pH-dependent, an effect that must
originate from a protonation change of His358 when going from pH 9.5 to pH 6
[56]. For steric reasons, it was concluded that at pH 9.5 the (deprotonated) His358
residue should turn towards the smaller Ala354 in the H354A mutant, which affects
50 E. Schleicher and S. Weber
the axial symmetry of the hyperfine tensor. This implies that His354 does not
change its protonation state when going from pH 9.5 to pH 6. Hence, the protonated
histidine that is proposed to catalyze intermediate formation must be His354,
because His358 is deprotonated at pH 9.5; see Fig. 3. In summary, these findings
beautifully demonstrate the potential of ENDOR to study protonation states of
amino acids by probing them with a nearby radical, such as the intrinsic flavin
radical in the working site of photolyases.
Very recently, the long-awaited crystal structures of Drosophila melanogaster
(64) photolyase in complex with DNA containing a (64) photoproduct lesion, and
in complex with DNA after in situ repair were presented (see Fig. 3, lower panel)
[71]. The overall structure of the (64)-photolyase looks surprisingly similar
compared to the previously published structures from CPD-photolyases [67, 72],
although the binding pocket for the DNA lesion is smaller but deeper than those
from the CPD-repairing enzymes, and governed by less hydrophobic amino acids.
This change in amino acid composition reflects the altered geometry of the enzyme-
bound (64) photoproduct and could be an argument for an alternative repair
mechanism. The previously discussed two conserved histidines are indeed located
in the binding pocket of the substrate, even though only His354 was found to be in
direct contact with the (64) photoproduct lesion via hydrogen bonding. Based on
their structure data, the authors proposed a new mechanism for the repair of the
(64) photoproduct [71]. However, in contrast to previously suggested reaction
schemes, this mechanism does not involve an oxetane intermediate, but electron
transfer from the flavin directly to the (64) photoproduct. Protonation of the one-
electron reduced (64) photoproducts 5-OH group by the nearby histidine then
facilitates elimination of a water molecule, which subsequently attacks the
acylimine molecule. This intermediate is proposed to split into the two thymines
in the DNA lesion and, after back-electron transfer to the flavin, the intact bases are
restored. During the last year, a number of studies, from both theoretical and
experimental groups, have been published [7375]. Surprisingly, all these recent
studies favor yet other, substantially different, reaction mechanisms for the enzy-
matic repair of (64) photoproducts. Clearly, from structure data alone, the pending
question on the reaction mechanism of (64) photolyase cannot be solved. Cer-
tainly, further EPR/ENDOR spectroscopic studies might be useful to solve this
interesting problem.
Fig. 4 ENDOR spectra of various A. sativa LOV2 domains. Left panel: pulsed X-Band Davies-
type proton ENDOR spectra of various AsLOV2 single - and double mutants. Spectra were
recorded at 120 K (dashed lines), 80 K (black lines), and 10 K (gray shaded) for all samples.
Right panel: sections of AsLOV2 spectra (dashed lines) recorded at 10 K with accompanying
spectral simulations (spectral contributions of individual hyperfine couplings are shown in color
shades of blue, green, and red; the superpositions of all deconvoluted hyperfine contributions are
shown as thick black lines). Two protein samples, AsLOV2 C450A/F509A and AsLOV2 C450A/
N425C, require another hyperfine component of axial symmetry for accurate spectral fitting. This
feature represents hyperfine coupling from fast rotating 8a-methyl group protons (H8aisotropic) and
is shown in orange. Adapted from [52]
at 120, 80, and 10 K are highly comparable to those of AsLOV2 C450A samples.
On the other hand, mutations of Phe509 and particularly Asn425 lead to a drasti-
cally altered temperature behavior, which is illustrated in the left panel of Fig. 4.
Conservative mutations such as F509V cause only subtle changes, whereas F509A,
N425S, and especially N425C show an increase in rotational freedom of the methyl
group even down to 10 K, which makes spectra of the latter mutation highly
Radicals in Flavoproteins 53
Fig. 6 H8a hyperfine couplings as a function of the rotation angle of the methyl group. Hyperfine
coupling values of the three H8a protons as a function of the rotation angle were obtained from
density functional theory. In gray, the isotropic part of the calculated hyperfine tensor for each of
the three hydrogens is shown. The dots mark the isotropic hyperfine values of two 8a protons
obtained by spectral simulation of the experimental data set. The inset shows a graphical view of
the angles of the three 8a-methyl-group protons with respect to the isoalloxazine ring plane for
AsLOV2 C450A, and for C450A/N425C. Adopted from [52]
comparable to those obtained from other LOV samples, such as LOV1 from
Chlamydomonas reinhardtii (data not shown) [52]. (2) Point mutations of specific
amino acid residues that change the angle of the arrested 8a-methyl protons with
respect to the isoalloxazine ring plane based on results from density functional
theory (DFT), dihedral angles could be determined, in which methyl rotation comes
to a halt (see Fig. 6). Compared to the C450A samples, mutations of Asn425 change
this situation dramatically: whereas the angle in N425S samples is only moderately
increased, N425C mutations shifted the dihedral angle by about 10 [52].
54 E. Schleicher and S. Weber
It turned out that some of the above-mentioned mutations with strong impact on
methyl group rotation also exhibit significantly altered dark-state recovery, i.e., the
breakage of the FMN C4aS bond, when introduced into wild type AsLOV2.
For example, mono-exponential fitting of the back reaction of the N425C mutant
results in an adduct-state lifetime of only 7.5 s, which is sixfold decreased as
compared to that of wild type AsLOV2 (45.8 s) [52]. Hence, interaction between
Asn425 and the 8a methyl group seems to coerce the FMNs isoalloxazine
moiety into a specific conformation, thus stabilizing the intrinsically weak FMN
C4aS bond of the cysteinyl-FMN 4a adduct. Replacement of the amino acid leads
to an increase in conformational flexibility of the FMN cofactor and, thus, to a
destabilization of the SC4a bond. Hence, careful examinations of hyperfine
couplings of a paramagnetic cofactor, and their modifications as a function of the
cofactor surroundings, revealed protein sites that have previously escaped attention
because they are rather remote from the location of FMN adduct formation.
Redox reactions have been proposed to play a key role in light-responsive activities
of cryptochromes [98, 99], blue-light photoreceptors in plants, animals, and bacte-
ria with widespread functions ranging from the regulation of circadian rhythms of
plants and animals [13] to the sensing of magnetic fields in a number of species
56 E. Schleicher and S. Weber
[17]. Both in vitro and in vivo experiments suggest that the FAD redox state is
changed from fully oxidized (FADox) to the radical form when it adopts the
signaling state [50, 51, 98]. The results agree with the redox activity of photolyases,
which share (to a high degree) amino acid sequence and three-dimensional structure
with cryptochromes [20]. Despite this similarity, cryptochromes generally do not
show DNA repair activity. In photolyases, when starting from FADox, photoin-
duced intraprotein electron transfer produces a radical pair, comprising an FAD
radical and either a tyrosine or tryptophan radical, which is directly observable by
time-resolved EPR [19, 100]. The specific amino acid involved in photoreduction
of FAD in Escherichia coli CPD photolyase was first identified by a comprehensive
point-mutational study in which each individual tryptophan of the enzyme was
replaced by phenylalanine [101]. Only the W306F mutation abolished photoreduc-
tion of FADox or FADH. Trp306 is situated at the enzyme surface at a distance of
approximately 20 A to FAD [67]. However, this distance is too large for a rapid
direct electron transfer between Trp306 and the FAD, which is completed within
30 ps, as has been determined recently by time-resolved optical spectroscopy [102].
Hence, a chain of tryptophans comprising Trp359 and Trp382 was postulated early
on upon elucidation of the three-dimensional structure [67] of the enzyme to
provide an efficient multistep electron-transfer pathway through well-defined
intermediates between Trp306 and the FAD [103]. This chain of tryptophans has
been conserved throughout all structurally characterized photolyases to date and is
also found in cryptochromes. While the relevance of this intraprotein electron
transfer for photolyase function is still under debate [104], the cascade is believed
to be critical for cryptochrome signaling. For example, it has been shown recently
that substitutions of the surface-exposed tryptophan or the tryptophan proximal to
FAD reduce in-vivo photoreceptor function of Arabidopsis cryptochrome-1 [105].
Radical pairs generated along the tryptophan chain by light-induced electron
transfer to FADox in cryptochromes have been proposed to function as essential
ingredients of a biological compass to collect information on the geomagnetic
orientation in a large and taxonomically diverse group of organisms [106]. In
principle, a compass based on radical-pair photochemistry requires (1) generation
of a spin-correlated radical pair with coherent interconversion of its singlet and
triplet states in combination with a spin-selective reaction, such as further for-
ward reactions that compete with charge recombination (which regenerates the
ground-state reactants) the latter is only allowed for the singlet radical-pair but not
the triplet radical-pair configuration, (2) modulation of the singlet-to-triplet inter-
conversion by Zeeman magnetic interactions of the unpaired electron spins with the
magnetic field, and (3) sufficiently small inter-radical exchange and dipolar
interactions such that they do not attenuate the radical pairs singlet-to-triplet
interconversion [17]. Hence, understanding the suitability and potential of
cryptochromes for magnetoreception requires identification of radical-pair states
and examination of their origin, and the detailed characterization of magnetic
interaction parameters and kinetics. TREPR with a time resolution down to 10 ns
allows real-time observation of such spin states generated by pulsed laser excita-
tion. Cryptochromes of the DASH type are ideal paradigm systems for such
Radicals in Flavoproteins 57
studies, because these proteins can be expressed from diverse species, their three-
dimensional structure is known [107, 108], and they are stable and available in the
amounts required for spectroscopic studies.
Recently, a TREPR study of light-induced paramagnetic intermediates from
wild-type cryptochrome DASH of the frog X. laevis was presented and the results
were compared with those from a mutant protein (W324F) lacking the terminal
tryptophan residue of the conserved putative electron-transfer chain [109]; see
Fig. 8.
The TREPR signal of wild-type cryptochrome recorded at physiologically
relevant temperature is depicted in three dimensions as a function of the magnetic
field and the time after pulsed laser excitation. Positive and negative signals
indicate enhanced absorptive and emissive electron-spin polarization of the EPR
transitions, respectively. The signal is assigned to a radical pair based on its spectral
shape and narrow width. A spin-polarized flavin triplet state detected under com-
parable experimental conditions would span more than 150 mT due to the large
spinspin interactions between the two unpaired electrons (see above). The time
evolution reveals that the radical-pair state exists for at least 6 ms; a more precise
determination is not possible because the exponential signal decay is affected by
58 E. Schleicher and S. Weber
populated. Similar to the singlet state, |S, the |T0-state has an equal number of
electrons with spin-down and spin-up projection. The |T+- and |T-states, on
the other hand, may contain different populations because of second-order effects
created by the size of the zero-field splitting parameters (D and E) relative to the
strength of the external magnetic field B0. If |T is in excess, then net enhanced
absorption occurs because |T contains two electrons with spin-down projec-
tion. Accordingly, if |T+ is in excess, net emission is observed (because of the two
electrons with spin-up projection). The stronger B0 is relative to the size of zero-
field splitting of the triplet zero-field sublevels |TX, |TY, and |TZ (X, Y and Z are
the principal axes of the dipolar coupling tensor in the triplet), the smaller the
population difference between the high-field triplet levels |T and |T+ despite the
non-Boltzmann population of the zero-field triplet levels [113]. In this so-called
high-magnetic-field limit, where the energies of |T and |T+ decrease and
increase, respectively, in a linear fashion with the external magnetic field B0, one
expects increasingly symmetric TREPR spectra [112]. In other words, the higher
the magnetic field, the more efficiently the enhanced absorptive and emissive
spectral contributions will cancel out, and the more the integral of the TREPR
signal intensity over the magnetic field will approach zero. Given typical zero-field
splitting parameters (|D| 70 mT, |E| 20 mT) and zero-field populations
(rX 2/3, rY 1/3, and rZ 0) of a flavin triplet state at physiological pH
[93], a net emissive TREPR spectrum is expected from a potentially triplet-born
cryptochrome radical pair at X-band, whereas at Q-band the spectrum will become
more symmetric with vanishing net electron-spin polarization. This is, however, not
observed [112]. Similar considerations can be applied to the potential case of a
triplet precursor state that is not spin-polarized but thermally relaxed prior to
radical-pair generation. In this situation TREPR spectra become more asymmetric,
i.e., net electron-spin polarization shifts towards enhanced absorption, with increas-
ing magnetic field B0. At X-band EPR, fairly symmetric spectra are expected,
whereas at Q-band net-absorption will be observed [112]. This nicely demonstrates
that a two-microwave-frequency TREPR approach can be used to draw conclusions
on the nature of the precursor electronic state in light-induced spin-correlated
radical-pair formation.
The TREPR results clearly show that cryptochromes (exemplified for the
DASH-type) readily form radical-pair species upon photoexcitation. Spin correla-
tion of such radical-pair states (singlet vs triplet), which is a necessary condition
for magnetoselectivity of radical-pair reactions, manifests itself as electron-spin
polarization of EPR transitions, which can be directly detected by TREPR in real
time. Such observations support the conservation of photo-induced radical-pair
reactions and their relevance among proteins of the photolyase/cryptochrome
family. The results are of high relevance for studies of magnetosensors based on
radical-pair (photo-)chemistry in general [114], and for the assessment of the
suitability of cryptochrome radical pairs in animal magnetoreception in particular
[17, 115].
60 E. Schleicher and S. Weber
Blue-light sensitive photoreceptor BLUF (blue light using FAD) domains are
also flavoproteins, which regulate various, mostly stress-related processes in bacte-
ria and eukaryotes. BLUF domains were found in various proteins, e.g., in the N-
terminus of the AppA protein of the purple bacterium Rhodobacter sphaeroides. In
this organism, AppA acts as a transcriptional antirepressor and interacts with the
photosynthesis repressor protein PpsR to form a stable AppA-(PpsR)2 complex in
the dark and under low-light conditions [116]. The BLUF photoreceptor class was
also discovered in a series of gene products derived from proteobacteria,
cyanobacteria, and some eukaryotes [117]. Three-dimensional single-crystal and
NMR structures for a number of BLUF proteins reveal a novel FAD binding fold
(see, e.g., [118122]). Photo-excitation of BLUF domains induces a small but
characteristic red-shift of the FAD absorption of about 1015 nm in the UV/vis
region. This was rationalized with a reaction mechanism, in which a very short-
lived biradical species comprising an FAD radical and a tyrosine radical are
generated upon optical excitation of the FAD [123]. The proposed radical-pair
species are, however, too short-lived (<1 ns) to be directly detected by TREPR at
physiological temperatures, first because the temporal resolution of TREPR rarely
reaches values below 10 ns, and second because the initial electron-spin polariza-
tion is quickly abolished by overall protein motion in liquid solution. In frozen
aqueous solution, however, paramagnetic species could be detected by TREPR on a
time scale of up to 100 ms [124]; see Fig. 9.
Triplet states were observed for various dark-adapted BLUF domains from
different organisms. Based on the zero-field splitting parameters, the spin-polarized
TREPR signals could be assigned to flavin triplet states. Radical-pair signatures
were also detected, although, despite high conservation of amino acid sequence and
three-dimensional structure, only from BLUF domains of Synechocystis sp. PCC
6803 (Slr-BLUF). The radical-pair signal of the wild type exhibits strong net-
enhanced absorptive spin polarization and is superimposed by a similarly strongly
spin-polarized flavin triplet spectrum. The narrow feature near g 2 was assigned
to a strongly coupled radical pair comprising the flavin chromophore and an
electron-donating amino acid, both in the radical form. Two aromatic amino acid
residues are reasonable close to the isoalloxazine ring of the FAD to transfer
efficiently an electron upon photoexcitation: Tyr8 and Trp91 in Slr-BLUF are
located within 0.50.6 nm to C4a of FAD, respectively (measured from the points
of highest unpaired electron-spin density of the respective residue). From the
magnetic-field position of extreme and inflection points, the zero-field splitting
parameters of the dipoledipole coupling tensor can be extracted. Under the
assumption of perfect axial symmetry (i.e., E 0) of the strongly-coupled radical
pair, the inflection points of the outer wings are separated by 2|D|, and
the inner features by |D|; see Fig. 9. Examination of the radical-pair feature yielded
|D| (10.3 0.3) mT. Using the point-dipole approximation, a distance of about
0.7 nm was calculated as the average distance between the two unpaired electron
spins of the radical pair, which is close to the distances found between Tyr8 and
FAD, or Trp91 and FAD. Hence, based on the dipolar coupling alone, the precise
origin of the radical pair TREPR signal cannot be unambiguously determined. To
assign the amino acid involved in electron transfer to the FAD, the radical-pair
signal of the wild type Slr-BLUF domain was compared to that of a mutant, where
Tyr8 was replaced by a tryptophan, Y8W. Removal of Tyr8 completely changed the
radical-pair feature, which now showed emissively and enhanced absorptively spin-
polarized signal contributions. These results show that Tyr8 is directly involved in
light-induced electron transfer to the flavin. However, the precise origin of the
radical pair signal from Y8W Slr-BLUF remains unclear. The electron-donating
residue could be Trp8, Trp91, or another as yet unidentified aromatic amino acid.
This issue needs to be addressed by a more comprehensive inspection of aromatic
amino acids and their involvement in the photocycle using TREPR preferably
performed at higher magnetic fields than in X-band EPR to resolve better the
radical pair signals by means of the different g-values of the coupled radicals.
5 Concluding Remarks
substrate binding and DNA repair? Why are the cryptochromes, despite their high
protein-sequence homology to (64) photolyases, incapable of repairing UV-
induced DNA lesions? Are cryptochromes capable of sensing and transducing
magnetic-field information, and if so, how is this task achieved in detail? Solving
these questions will be a challenge for the next decade(s). We are confident that
application of modern EPR will make an important contribution to this.
References
24. Schleicher E, Wenzel R, Okafuji A, Ahmad M, Batschauer A, Essen L-O, Hitomi K, Getzoff
ED, Bittl R, Weber S (2010) Appl Magn Reson 37:339
25. Weber S, Kay CWM, Bacher A, Richter G, Bittl R (2005) Chemphyschem 6:292
26. Garca JI, Medina M, Sancho J, Alonso PJ, Gomez-Moreno C, Mayoral JA, Martnez JI
(2002) J Phys Chem A 106:4729
27. Acocella A, Jones GA, Zerbetto F (2010) J Phys Chem B 114:4101
28. Fuchs M, Schleicher E, Schnegg A, Kay CWM, T orring JT, Bittl R, Bacher A, Richter G,
Mobius K, Weber S (2002) J Phys Chem B 106:8885
29. Barquera B, Morgan JE, Lukoyanov D, Scholes CP, Gennis RB, Nilges MJ (2003) J Am
Chem Soc 125:265
30. Kay CWM, Bittl R, Bacher A, Richter G, Weber S (2005) J Am Chem Soc 127:10780
31. Schnegg A, Kay CWM, Schleicher E, Hitomi K, Todo T, M obius K, Weber S (2006) Mol
Phys 104:1627
32. Schnegg A, Okafuji A, Bacher A, Bittl R, Fischer M, Fuchs MR, Hegemann P, Joshi M, Kay
CWM, Richter G, Schleicher E, Weber S (2006) Appl Magn Reson 30:345
33. Okafuji A, Schnegg A, Schleicher E, M obius K, Weber S (2008) J Phys Chem B 112:3568
34. Kay CWM, El Mkami H, Molla G, Pollegioni L, Ramsay RR (2007) J Am Chem Soc
129:16091
35. Pauwels E, Declerck R, Verstraelen T, De Sterck B, Kay CWM, Van Speybroeck V,
Waroquier M (2010) J Phys Chem B 114:16655
36. Mobius K (2000) Chem Soc Rev 29:129
37. Steinhoff H-J, Savitsky A, Wegener C, Pfeiffer M, Plato M, M obius K (2000) Biochim
Biophys Acta 1457:253
38. Kay CWM, Schleicher E, Hitomi K, Todo T, Bittl R, Weber S (2005) Magn Reson Chem 43:
S96
39. Gulla AF, Budil DE (2001) J Phys Chem B 105:8056
40. Sinnecker S, Reijerse E, Neese F, Lubitz W (2004) J Am Chem Soc 126:3280
41. Teutloff C, Hofbauer W, Zech SG, Stein M, Bittl R, Lubitz W (2001) Appl Magn Reson
21:363
42. Murphy DM, Farley RD (2006) Chem Soc Rev 35:249
43. van Doorslaer S, Vinck E (2007) Phys Chem Chem Phys 9:4620
44. Kulik L, Lubitz W (2009) Photosynth Res 102:391
45. Kay CWM, Feicht R, Schulz K, Sadewater P, Sancar A, Bacher A, M obius K, Richter G,
Weber S (1999) Biochemistry 38:16740
46. Medina M, Cammack R (2007) Appl Magn Reson 31:457
47. Bittl R, Kay CWM, Weber S, Hegemann P (2003) Biochemistry 42:8506
48. Barquera B, Ramirez-Silva L, Morgan JE, Nilges MJ (2006) J Biol Chem 281:36482
49. Nagai H, Fukushima Y, Okajima K, Ikeuchi M, Mino H (2008) Biochemistry 47:12574
50. Banerjee R, Schleicher E, Meier S, Munoz Viana R, Pokorny R, Ahmad M, Bittl R,
Batschauer A (2007) J Biol Chem 282:14916
51. Hoang N, Schleicher E, Kacprzak S, Bouly JP, Picot M, Wu W, Berndt A, Wolf E, Bittl R,
Ahmad M (2008) PLoS Biol 6:e160.1559
52. Brosi R, Illarionov B, Mathes T, Fischer M, Joshi M, Bacher A, Hegemann P, Bittl R, Weber
S, Schleicher E (2010) J Am Chem Soc 132:8935
53. Hamdane D, Guerineau V, Un S, Golinelli-Pimpaneau B (2011) Biochemistry 50:5208
54. Goldfarb D, Arieli D (2004) Annu Rev Biophys Biomol Struct 33:441
55. Goldfarb D (2006) Phys Chem Chem Phys 8:2325
56. Schleicher E, Hitomi K, Kay CWM, Getzoff ED, Todo T, Weber S (2007) J Biol Chem
282:4738
57. Kurreck H, Bock M, Bretz N, Elsner M, Kraus H, Lubitz W, M uller F, Geissler J, Kroneck
PMH (1984) J Am Chem Soc 106:737
58. Cinkaya I, Buckel W, Medina M, Gomez-Moreno C, Cammack R (1997) Biol Chem 378:843
64 E. Schleicher and S. Weber
Dariush Hinderberger
Abstract Synthetic polymers belong to the vast realm of soft matter and are one
of the key types of materials to address societal needs at the beginning of the
twenty-first century. Polymer science progressively addresses questions that deal
with tuning mesoscopic and macroscopic structures and functions of polymers by
understanding the effects that govern these systems on the nanoscopic level. EPR
spectroscopy as a local, sensitive, and extremely specific magnetic resonance
technique in many cases shows sensitivity on well-suited length- (010 nm) and
time scales (msps) and can deliver unique information on structure, dynamics, and
in particular function of polymeric systems. A short review of recent literature is
given and the power of simple EPR methods, especially CW EPR performed on
a low-cost benchtop spectrometer, to elucidate complex polymeric materials is
shown with specific examples from thermoresponsive polymer systems. These
bear great potential in molecular transport and biomedical applications (e.g., drug
delivery) and insights into interactions between carrier and small molecule are
fundamental for designing and tuning these materials.
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
1.1 Non-covalent Interactions Shape the Molecular Scale of Complex
Polymeric Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2 Radicals and CW EPR for the Study of Polymeric Systems on the Molecular Scale . . . . . 71
2.1 Radicals, EPR Spectra, and Their Information Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
D. Hinderberger
Max-Planck-Institut f
ur Polymerforschung, Ackermannweg 10, 55128 Mainz, Germany
e-mail: dariush.hinderberger@mpip-mainz.mpg.de
68 D. Hinderberger
1 Introduction
Synthetic polymers were one of the major driving forces of industrial and scientific
development in the second half of the last century. A large portion of research on
polymers initially dealt with understanding macroscopic properties, in particular
their mechanical features. Nowadays, research in the polymer or macromolecular
sciences has evolved far beyond these questions and specifically aims at under-
standing, manipulating, and designing the molecular level of polymers. Further-
more, not only classical structural polymers like polyethylene or polystyrene are
studied but also the focus has shifted towards functional and smart polymeric
materials.
In a recent article, leading researchers in the field have summarized challenges
and opportunities for research in macromolecular science [1]. Five societal fields
were identified, in which polymers can play a decisive role in the coming years and
that researchers in polymer science should be aware of. These fields are energy,
sustainability, health, and security, informatics/defense/protection. This extremely
broad collection of fields that certainly shape our near future already implies
that polymers (or even more generally, synthetic soft matter systems) are extremely
versatile in applications. This huge range of possible applications is a direct
consequence of the extreme variability of physical properties that can be achieved
by chemical fine tuning on the molecular level. Modern polymer science is
highly interdisciplinary, covering fields from chemical synthesis via processing
technology and physical characterization all the way into biology and theoretical
chemistry/physics. In the above-mentioned summary of current challenges and
opportunities in polymer science, key research fields are named, in which advances
are urgently needed to achieve progress for beneficial application of polymers
in the mentioned societal issues: (1) new synthetic routes, (2) complex polymer
systems formed through weak, non-covalent interactions, (3) characterization and
properties, (4) theory and simulations, and (5) macromolecular processing and
assembly.
Of these five key fields, electron paramagnetic resonance (EPR) spectroscopy
as a local, sensitive, and highly selective technique can naturally add important
information to fields (2) and (3), i.e., formation of complex polymer systems through
non-covalent interactions and characterization and properties thereof, respectively.
Polymeric systems have been under investigation with EPR spectroscopy for more
EPR Spectroscopy in Polymer Science 69
Polymers are one major part of the very broad field of soft matter. This term is
a classification that summarizes a wide variety of systems, from liquids via liquid
crystals, micelles, synthetic macromolecules, membranes, colloidal suspensions,
foams, and gels, right into the realm of biology with biomacromolecules (proteins,
enzymes, DNA, RNA), viruses, and even whole cells [310]. All these systems,
man-made or biological, have in common that their structure and function are
governed by physical processes and interactions that are comparable to the thermal
energy at room temperature (kBT 4.1 1021 J). Some of these interactions
are schematically depicted in Fig. 1. This differentiates them from hard matter, in
which typical energies are several orders of magnitude larger.
One of the most fascinating albeit complicating features of soft matter in
general and complex polymers in particular is that usually many (weak), often
counteracting, interactions on the molecular scale control their unique mesoscopic
and/or macroscopic structure and function. In addition, the contribution of the
chemical composition of the individual interacting species and entropic contributions
play a substantial role. One aspect that is of utmost importance in particular for
function in biological polymers is the existence of non-equilibrium processes on the
molecular scale [1113]. To achieve such molecularly controlled non-equilibrium
states it is in fact advantageous to utilize weak interactions. With interactions that are
too strong, living systems would be much better structured but their dynamics and
hence their function would simply disappear.
70 D. Hinderberger
+
electrostatic
interactions
+
O H O
(induced) dipole
interactions
H-bonding
hydrophobic - interactions
interactions
Fig. 1 A selection of weak interactions shaping polymers and soft matter in general on the
molecular scale. These interactions, often in combination with each other, govern polymer
structure and function
Ever since De Gennes and contemporaries shaped this vast field of condensed
matter research, many experimental and theoretical studies focused on their meso-
and macroscopic scale properties. Due to the weakness of the individual forces, the
complexity of the many contributing interactions and the varying degrees of order,
fewer studies have dealt with the molecular scale origin of soft matter and polymer
properties [6, 9, 1416].
The relevant length scales in polymers span a range of at least four orders
of magnitude, from Angstroms (see the forces in Fig. 1) up to at least a few micro-
meters. In soft matter and polymer theory the multiscale modeling approach, which
attempts to simulate such systems in total including all relevant interactions, has
recently attracted a lot of attention [17]. In experimental polymer research an
analogous approach would be the combination of results from many techniques
[18].
The central point of this chapter is to show that methods of EPR spectroscopy,
in particular simple CW EPR, can unravel structural, dynamic, and functional
properties of complex, synthetic polymers on the molecular scale. It will be
shown that although EPR spectroscopy is a local method that has greatest versatility
in the range of several nanometers, it is possible to obtain insight into fundamental
interactions on the short length scales and into macroscopically observable
functions on much longer length scales.
EPR Spectroscopy in Polymer Science 71
The work presented in Sect. 3 shows that with simple CW EPR on spin probes
that are directed to the site of interest solely through non-covalent interactions it is
possible to unravel and characterize structural and functional inhomogeneities on
the nanoscale in thermoresponsive polymeric systems. As no chemical attachment
was needed, this could be called a site-directed spin probing approach (see also,
e.g., [19] or Chap. 7 of [2]). The molecular-scale characterization in these cases
delivered evidence that the macroscopically observable thermal response has its
origin in these inhomogeneities on the nanoscale.
Section 4 then gives a very short summary of other polymer-oriented studies of
the last 5 years, again with no claim of completeness.
It should be noted that EPR spectroscopy in fields like bioinorganic chemistry or
protein research is nowadays often combined with other experimental and simula-
tion methods. In contrast, for rather randomly structured materials such as
thermoresponsive polymer systems presented here, one can obtain meaningful
insight with simple CW EPR methods. Before the specific recent examples are
presented, the necessary basics of CW EPR spectroscopy are explained in the next
section.
EPR today is a spectroscopic method for determining the structure, dynamics, and
spatial distribution of paramagnetic species [20]. For the first 25 years after its
discovery in Kazan in the winter of 1944, EPR was a playground for scientists
studying the fundamental physical properties of condensed matter. There are
several reasons why this has changed markedly today and why EPR nowadays
has found its firm albeit specialized place in materials science and biophysics: (1)
the seminal work by Freed and others [2124] starting in the late 1960s on the
detailed analysis of molecular dynamics from EPR spectra, (2) the advent of several
pulse EPR techniques [25, 26] and the first commercially available pulse EPR
spectrometer in the 1980s, and (3) the breakthrough of making possible distance
measurements in the range between 1 and ~8 nm in the late 1990s [2729].
EPR applications are nonetheless limited due to lack of naturally occurring
paramagnetic systems. In many cases one has to introduce stable free radicals
artificially. These are called spin labels if they are covalently linked to the
investigated system and spin probes if their interaction is of non-covalent nature
[19, 3032]. This apparent disadvantage, though, can often be of great value, as
these spin probes can be detected very specifically and with higher sensitivity as
compared to NMR on, e.g., 1H-nuclei [33].
This section gives a concise introduction to CW EPR spectroscopy and only
highlights the points needed for understanding the examples from polymer science
in the following section. For a more detailed introduction to EPR spectroscopic
techniques the reader is referred to [19, 34].
72 D. Hinderberger
The starting point for every kind of EPR study is the choice of the radical that is best
suited to deliver valuable information on the systems that are investigated. In this
chapter only three types of radicals are used for this purpose. Nitroxide radicals
such as TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl; Fig. 2), its derivatives, and
its five-ring analogs (Fig. 2) are the by far most widely used spin probes and spin
labels and they are the spin probes of choice for the majority of EPR spectroscopic
applications in naturally diamagnetic systems [2, 19].
For completeness one should note that the second very important type of
paramagnetic species that are studied intensively are paramagnetic transition
metals, mainly in the form of complexes. Whether transition metal ions are para-
magnetic or not depends on the oxidation state and, without giving a complete
list, examples often found in the literature are Cu2+, Cr3+, Mn2+, Fe3+, Co2+, and
Ni1/3+ [35].
Fig. 2 Chemical structures of common nitroxide spin probes: Fremys salt (potassium nitrosodi-
sulfonate); TEMPO and derivatives (2,2,6,6-tetramethylpiperidine-1-oxyl), DOXYL (4,4-dimethyl-
oxazolidine-1-oxyl); PROXYL (2,2,5,5-tetramethylpyrrolidine-1-oxyl); Dehydro-PROXYL
(2,2,5,5-tetramethylpyrroline-1-oxyl); 5-DSA (5-DOXYL stearic acid); 16-DSA (16-DOXYL stearic
acid)
EPR Spectroscopy in Polymer Science 73
Immediate environment:
Coupled remote spin
g-value / tensor (fingerprint)
14
N hyperfine coupling: aiso / A-tensor
dipole-dipole
coupling
CW EPR
10 ps-1 ms
0-2 nm
N
dipolar hyperfine O
coupling
O
P
O O
ENDOR
Observer spin HYSCORE
0-1 nm
O
Fig. 3 Graphical summary of the information content of EPR spectroscopic methods on nitroxide
radicals and a summary of the length and time scales of interest
74 D. Hinderberger
(DEER), allows distance measurements through the odd / r3 dependence (odd
dipolar coupling frequency; r interspin distance). Due to the frequency range
of EPR interactions, the accessible distances are in the extremely interesting range
of 1.58 nm. Particularly for biopolymers and soft matter in general, this closes
a gap between distance information from NMR (<1 nm, [33]) and from fluores-
cence-based methods, which on the other hand can address single molecule
(>3 nm, [4245]).
For nitroxides in dilute liquid solution, the generally anisotropic spin Hamilton
operator is simplified tremendously and, if unresolved proton hyperfine couplings
are treated as line broadening, only the electron-Zeeman interaction and the hyper-
fine coupling to the magnetic 14N nucleus (I 1) remain [20]. The g- and hyperfine
(A-) tensors are averaged to isotropic values due to fast motion of the spin probe and
the resonance condition for the irradiated microwave becomes
DEnit
hoS gbe B0 aiso mI (1)
which explains the three-line pattern of the nitroxide CW EPR spectra shown in
Fig. 4. Due to the interaction with the nitrogen nucleus (I 1) there are three
transitions allowed that lead to the three-line nitroxide spectrum.
Equation (1) also nicely shows that the major advantage of performing high-
field/high-frequency EPR, e.g., going to W-band (~94 GHz) frequencies, is the
improved g-resolution, while the hyperfine resolution remains unaltered. It thus
becomes possible to separate contributions to the spectra from electron-Zeeman and
hyperfine anisotropies.
The effect of rotational motion and chemical environment on X-band CW EPR
spectra is also summarized in Fig. 4 [19]. Rotational diffusion of spin probe
molecules in CW EPR (Fig. 4a) can be roughly characterized by the rotational
correlation time tc as belonging to one of the following regimes: (1) fast motion
(tc 100 ps), (2) intermediate motion (100 ns > tc > 1 ns), and finally (3) rigid
limit (tc 1 ms).
Rotational motion can be isotropic or anisotropic (e.g., when spin labels are
attached to larger polymer molecules) and analysis of CW EPR spectra most often
is quantified by spectral simulations assuming a rotational model of some sort
(e.g., isotropic Brownian or uni-axial motion or more complicated models like
microscopic order, macroscopic disorder, or MOMD; see [19, 21]).
The effect of the local environment on the nitroxide spectra is illustrated in
Fig. 4b. The electronic structure of a nitroxide is slightly altered depending on the
EPR Spectroscopy in Polymer Science 75
Fig. 4 Information from nitroxide CW EPR spectra. (a) Right: principal axis system of electron-
Zeeman and hyperfine tensors (collinear). Left: the effect of rotational dynamics on the CW EPR
spectra. Fast rotation (i.e., faster than a typical rotational correlation time of tc ~ 10 ps) leads to the
averaged spectrum. The isotropic g-value giso determines the center of the central line and spacing
between the lines that is dominated by aiso. In the intermediate motion regime 100 ns > tc > 1 ns
and the rigid limit is reached at tc ~ 1 ms [19]. (b) Influence of the chemical environment on CW
EPR spectra. As both, hydrophilic and polar environments lead to an increased electron spin
density at the nitroxide 14N nucleus (see gray inset), aiso and hence the line splitting in the spectra
in hydrophilic and polar surroundings is larger than in non-polar and hydrophobic environments
The changes in the line splitting are small but add up at the high-field line.
Hence, spin probes in different nanoscopic environments in inhomogeneous
samples can be distinguished and analyzed separately. In fact, this ability to
distinguish regions of different environments is used to reveal nano-inhomogeneities
in the thermoresponsive polymeric systems, which are described in the next section.
76 D. Hinderberger
When the much higher resolution of high-field EPR is available, one can even
obtain a more detailed picture and distinguish polarity- and proticity-based effects
[46, 47].
3 Nano-inhomogeneities in Thermoresponsive
Polymeric Systems
can directly analyze the nanoscopic inhomogeneities from their CW EPR spectra.
In three-dimensional hydrogels based on poly(N-isopropylacrylamide), PNIPAAM,
as the thermoresponsive polymer we find static inhomogeneities not only of struc-
ture but most remarkably also of function (Sect. 3.1 [51]). In contrast, in thermore-
sponsive dendronized polymers, dynamic inhomogeneities over a broad temperature
range are accompanied by one of the sharpest macroscopic thermal responses ever
recorded (Sect. 3.2 [5254]).
These findings have led to the conclusion that the polymer network collapse is
a continuous, nano-inhomogeneous process, in which individual polymeric pockets
are in a collapsed state even at temperatures significantly below the LCST and
that the macroscopic collapse takes place only when a certain number and/or volume
of collapsed pockets is reached.
This is schematically depicted in Fig. 5c. These nano-inhomogeneities can
furthermore be considered as static on the EPR-time scale, as they neither change
with time nor with temperature. With temperature, only the number or volume
fraction of collapsed regions increases but there is no dynamic opening/closing of
the collapsed regions. Once they are formed and once they have trapped TEMPO
molecules, the only way to open the collapsed pockets for the aqueous phase is to
reduce temperature again. It is worthwhile to note that, if the cationic and more
hydrophilic radical Cat1 is used instead of TEMPO, no distinct second species can
be observed; Cat1 does not sample hydrophobic regions.
Remarkably, the hydrophilic regions further form nanoreactors, which strongly
accelerate acid-catalyzed disproportionation reactions if acidic protons are present,
EPR Spectroscopy in Polymer Science 79
Nanoshelter (B)
A 0 min
B
30 min
60 min
100 min
Nanoreactor (A)
Fig. 6 The high-field peak region of TEMPO CW EPR spectra in methacrylic-acid containing
PNIPAAM copolymer networks (left). A denotes the peak from TEMPO in hydrophilic envi-
ronment, while B denotes the peak from TEMPO in the collapsed regions. The spectra were
recorded in the course of 120 min and in this timeframe the B type species is only minimally
reduced, while the hydrophilic A species is significantly reduced due to acid-catalyzed dispro-
portionation of TEMPO to diamagnetic compounds. The coexistence of regions, in which this
chemical reaction is facilitated (nanoreactors) and simultaneously of regions, in which TEMPO
molecules are protected from the reaction (nanoreactors) is schematically depicted on the right
e.g., from methacrylic acid co-monomers in the hydrogel, while simultaneously the
hydrophobic regions act as nanoshelters, in which enclosed spin probes are
protected from decay (see Fig. 6). The results show that the system consisting of
a statistical binary or tertiary copolymer displays remarkably complex behavior that
mimics spatial and chemical inhomogeneities observed in functional biopolymers
such as enzymes.
Note that in a subsequent study a similar result of micro-/nanophase separation
has been observed in block copolymers of PNIPAAM and N-isopropylmetha-
crylamide (PNIPMAM). This small-angle neutron scattering (SANS) study used a
scattering analysis with a new form factor model taking into account a nanophase
separated internal morphology [50].
The simple and cheap EPR method presented here not only directly mirrors the
nanophase separation in the CW EPR spectra without the use of any model but even
allows the characterization of the nano-inhomogeneity on the functional level of a
chemical reaction.
Fig. 7 (a) Chemical structure of the thermoresponsive dendronized polymer PG2(ET), denoting a
second generation dendron that is capped with ethoxy groups. (b) The high-field peak region of
TEMPO CW EPR spectra in PG2(ET). In contrast to the spectra in PNIPAAM (Figs. 5 and 6) there
is a temperature dependent shift of the hyperfine splitting. This indicates dynamic inhomogeneities
on the nanoscale, as explained in the text
collapsing and assembling into mesoglobules, does not exist. This severely hampers
rational materials design.
In studies aiming at detection of unusual properties of dendronized polymers
[7478], Schl uter, Zhang, and coworkers recently discovered that such systems
based on oligoethyleneglycol (OEG) units exhibit fast and fully reversible phase
transitions with a sharpness that is amongst the most extreme ever observed (see
Fig. 7a) [7981].
These dendronized polymers with terminal ethoxy groups are soluble in water
and their LCST is found in a physiologically interesting temperature regime
between 30 and 36 C. It is also interesting to note that the LCST of these OEG
dendronized polymers is as low as is known for poly(ethylene oxide) and long chain
ethylene oxide oligomers. For the latter the influence of hydrophobic end groups on
the LCST has been thoroughly investigated both experimentally and theoretically
[82]. Given this extraordinary behavior, these polymers are particularly suited to
gain a deeper fundamental understanding of the processes involved. As described
in the previous section, there are clear indications that thermal responses proceed
via the formation of structural inhomogeneities of variable lifetimes on the nano-
meter scale that are still poorly understood. In this section, the focus is on a
clearer understanding of the formation, structure, and lifetimes of these local
inhomogeneities, the effect of the individual chemical structures on the physical
processes, and the influence of the local heterogeneities on the aspired function
(e.g., drug delivery).
The remarkable macroscopic behavior of such materials results from the systems
being far from classical macroscopic equilibrium. It can be viewed as an example of
EPR Spectroscopy in Polymer Science 81
Hence, slightly above TC the few hydrophobic cavities formed are still small,
i.e., in the range of a few nanometers. Then spin probe movement and/or local
polymer fluctuations lead to an exchange of the probe molecules on the EPR time
scale between the hydrophobic and large hydrophilic regions. The latter are still
overwhelmingly more abundant. The spin probes thus mainly sample the interface
between the two fundamentally different regions. Note that a few local dynamic
heterogeneities on a nanometer scale are sufficient to induce a macroscopically
observable (by turbidity measurements) transition in the sample. This macroscopic
transition is detected at the same temperature as the change in the EPR spectra
occurs. This suggests that the small hydrophobic regions detected by EPR might
be visualized as cross-links affecting the organization of the dendronized
macromolecules on much larger scales. The sharp macroscopic transition can
then be viewed as the onset of a complex de-swelling process that is broad on the
molecular scale rather than a sharp transition.
When increasing the temperature, not only the fraction of the hydrophobic
regions but also their size grows and exchange of probe molecules between
hydrophobic and hydrophilic sites becomes unlikely. The spin probes now sample
the bulk hydrophobic (and remaining hydrophilic) regions rather than their inter-
face. Together with the increase in size, the dynamics of the polymer fluctuations
slow down, as both effects are coupled. In combination, a final state of distinct
hydrophobic and hydrophilic regions is observed that are static on the EPR time
scale (denoted B (final) in Fig. 7b). The complex collapse transition is
schematically illustrated in Fig. 8.
The aggregation and the collapse can further be characterized by analyzing the
effective hyperfine coupling constants of those TEMPO molecules in hydrophobic
environments aB and the fraction of TEMPO in hydrophilic environments yA as
a function of temperature. By plotting these parameters vs the reduced temperature
(T TC)/TC it was possible to check whether the collapse results from a well-
behaved phase transition. Both parameters do not follow one straight line, expected
Fig. 8 Model of the collapse of thermoresponsive dendronized polymers as seen by EPR spin
probes: few individual hydrophobic and dynamic patches of ~5 nm are sufficient to achieve
a macroscopic collapse at the cloud point. Only at temperatures well above Tc a static state
(on EPR time scales) is reached
EPR Spectroscopy in Polymer Science 83
for a simple phase transition, but instead strongly deviate from linearity [85]. Thus,
in this wide temperature range, a complex dehydration takes place which cannot be
described in the picture of a classical phase transition based on a single de-swelling
process.
Furthermore, the effect of the chemical structure in the core of the dendrons on
the thermoresponsive behavior can be tested. It is found that the initial dehydration
and aggregation process at Tc turns out to be most effective when the dehydration
is supported by a hydrophobic core. It deteriorates when the core contains
oxyethylene groups, which can trap more water [53].
Altogether, a collapse transition as sketched in Fig. 8 is found: below Tc, no
aggregation of dendrons is observed. At and above Tc, there is a growing number of
uncorrelated hydrophobic regions up to a concentration and/or a volume fraction
that is similar to that of the remaining hydrophilic regions. This could be an
indication that the growth of hydrophobic regions reaches a threshold that could
be interpreted as a percolation point. When the fractions of species A and B become
equal, the likelihood of two hydrophobic regions (which up to that point can be
largely uncorrelated) becoming neighbors increases immensely and the role of the
interface becomes less important. Hence, at temperatures well above Tc (i.e., at least
30 C), a static situation is achieved where TEMPO molecules are either trapped
in hydrophilic or hydrophobic regions. There is no exchange on relevant EPR time
scales between these two regions any more.
References
1. Ober CK, Cheng SZD, Hammond PT, Muthukumar M, Reichmanis E, Wooley KL, Lodge TP
(2009) Macromolecules 42:465471
2. Schlick S (2006) Advanced ESR methods in polymer research. Wiley-Interscience, Hoboken,
NJ, USA
3. de Gennes PG (1992) Rev Mod Phys 64:645648
4. Kamien RD (2002) Rev Mod Phys 74:953971
5. Jain S, Bates FS (2003) Science 300:460464
6. Whitesides GM, Lipomi DJ (2009) Faraday Trans 143:373384
7. George M, Weiss RG (2006) Acc Chem Res 39:489497
8. Schild HG (1992) Prog Polym Sci 17:163249
9. MacKintosh FC, Schmidt CF (2010) Curr Opin Cell Biol 22:2935
10. Janmey PA, Hvidt S, Kas J, Lerche D, Maggs A, Sackmann E, Schliwa M, Stossel TP (1994)
J Biol Chem 269:3250332513
11. Cady F, Qian H (2009) Phys Biol 6:036011
12. Brutlag D, Kornberg A (1972) J Biol Chem 247:241248
13. Drobny GP, Long JR, Karlsson T, Shaw W, Popham J, Oyler N, Bower P, Stringer J, Gregory
D, Mehta M, Stayton PS (2003) Annu Rev Phys Chem 54:531571
14. Israelachvili JN (1991) Intermolecular and surface forces, 2nd edn. Academic Press, London
15. Monkenbusch M, Richter D (2007) C R Physique 8
16. Messina R (2009) J Phys Condens Matter 21:113102
17. Praprotnik M, Delle Site L, Kremer K (2008) Annu Rev Phys Chem 59:545571
18. Spiess HW (2010) Macromolecules 43:54795491
EPR Spectroscopy in Polymer Science 87
53. Junk MJN, Li W, Schl uter AD, Wegner G, Spiess HW, Zhang A, Hinderberger D (2011)
Macromol Chem Phys 212:12291235
54. Junk MJN, Li W, Schl uter AD, Wegner G, Spiess HW, Zhang A, Hinderberger D (2011)
J Am Chem Soc 133:1083210838
55. Hirotsu S, Hirokawa Y, Tanaka T (1987) J Chem Phys 87:13921395
56. Yu H, Grainger DW (1993) J Appl Polym Sci 49:15531563
57. Ikkai F, Shibayama M (2005) J Polym Sci B Polym Phys 43:617628
58. Kariyo S, Kuppers M, Badiger MV, Prabhakar A, Jagadeesh B, Stapf S, Blumich B (2005)
Magn Reson Imaging 23:249253
59. Hinderberger D, Schmelz O, Rehahn M, Jeschke G (2004) Angew Chem 2004,
116:47164721; Angew Chem Int Ed 2004, 43:46164621
60. Harvey RD, Schlick S (1989) Polymer 30:1116
61. Rex GC, Schlick S (1987) Polymer 28:21342138
62. Beines PW, Klosterkamp I, Menges B, Jonas U, Knoll W (2007) Langmuir 23:22312238
63. Wu C, Zhou SQ (1995) Macromolecules 28:53885390
64. Wu C, Zhou SQ (1995) Macromolecules 28:83818387
65. Wu C, Zhou SQ (1996) Phys Rev Lett 77:30533055
66. Wang X, Qiu X, Wu C (1998) Macromolecules 31:29722976
67. Van Durme K, Verbrugghe S, Du Prez FE, Van Mele B (2004) Macromolecules
37:10541061
68. Luo S, Xu J, Zhu Z, Wu C, Liu S (2006) J Phys Chem B 110:91329138
69. Ono Y, Shikata T (2006) J Am Chem Soc 128:1003010031
70. Cheng H, Shen L, Wu C (2006) Macromolecules 39:23252329
71. Van Durme K, Van Assche G, Aseyev V, Raula J, Tenhu H, Van Mele B (2007)
Macromolecules 40:37653772
72. Ono Y, Shikata T (2007) J Phys Chem B 111:15111513
73. Keerl M, Smirnovas V, Winter R, Richtering W (2008) Angew Chem 120:344347; Angew
Chem Int Ed 2008, 47:338341
74. Schluter AD, Rabe JP (2000) Angew Chem 112:860880; Angew Chem Int Ed 2000,
39:864883
75. Zhang A, Shu L, Bo Z, Schl uter AD (2003) Macromol Chem Phys 204:328339
76. Schluter AD (2005) Top Curr Chem 245:151191
77. Frauenrath H (2005) Prog Polym Sci 30:325384
78. Rosen BM, Wilson CJ, Wilson DA, Peterca M, Imam MR, Percec V (2009) Chem Rev
109:62756540
79. Li W, Zhang A, Feldman K, Walde P, Schl uter AD (2008) Macromolecules 41:36593667
80. Li W, Zhang A, Schl uter AD (2008) Chem Commun 55235525
81. Li W, Wu D, Schl uter AD, Zhang A (2009) J Polym Sci A Polym Chem 47:66306640
82. Dormidontova EE (2004) Macromolecules 37:77477761
83. Knauer BR, Napier JJ (1976) J Am Chem Soc 98:4395
84. Kovarskii AL, Wasserman AM, Buchachenko AL (1972) J Magn Reson 7:225237
85. Chaikin PM, Lubensky TC (1995) Principles of condensed matter physics. Cambridge
University Press, Cambridge
86. Deo P, Deo N, Somasundaran P, Moscatelli A, Jockusch S, Turro NJ, Ananthapadmanabhan
KP, Ottaviani MF (2007) Langmuir 23:59065913
87. Schleidt S, Spiess HW, Jeschke G (2006) Colloid Polym Sci 284:12111219
88. Panek G, Schleidt S, Mao Q, Wolkenhauer M, Spiess HW, Jeschke G (2006) Macromolecules
39:21912200
89. Mao Q, Schleidt S, Zimmermann H, Jeschke G (2007) Macromol Chem Phys 208:21452160
90. Mao Q, Schleidt S, Zimmermann H, Jeschke G (2008) Phys Chem Chem Phys 10:11561167
91. Miwa Y, Drews AR, Schlick S (2006) Macromolecules 39:33043311
92. Miwa Y, Drews AR, Schlick S (2008) Macromolecules 41:47014708
EPR Spectroscopy in Polymer Science 89
Malte Drescher
Contents
1 Site-Directed Spin-Labeling of Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
2 SDSL EPR Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.1 Mobility Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
2.2 Distance Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
2.3 Accessibility Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3 Applications to Intrinsically Disordered Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3.1 a-Synuclein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
M. Drescher
Department of Chemistry, Zukunftskolleg, and Konstanz Research School Chemical Biology,
University of Konstanz, 78457 Konstanz, Germany
e-mail: malte.drescher@uni-konstanz.de
92 M. Drescher
The main purpose of this chapter is to summarize the state of the art in Electron
Paramagnetic Resonance (EPR) spectroscopy combined with spin-labeling
approaches as a tool for studying structure and dynamics of proteins. It should
give the non-specialist reader an overview of the tremendous methodological
developments and applications which have a huge impact on the field of biophysics,
biology, and biochemistry. In particular, contributions of site-directed spin labeling
EPR (SDSL EPR) to the rapidly advancing field of Intrinsically Disordered Proteins
(IDPs) are described.
SDSL EPR as pioneered by W. L. Hubbel and co-workers has become a
powerful tool for studying structure and dynamics of macromolecules, in particular
biological macromolecules as proteins, which do not necessarily contain endoge-
nous paramagnetic centers [14]. While SDSL EPR is applied to many bio-
macromolecules, this chapter provides a rather selective insight into the field of
SDSL EPR of proteins and is organized as follows.
First a basic introduction into approaches of SDSL of proteins is given, followed
by a summary of the important approaches for mobility measurements, accessibility
studies, and distance determination. Finally, we will address IDPs. Since structure
and dynamics of IDPs drastically depend on the environment and corresponding
details are notoriously difficult to unravel by NMR or X-ray structure determina-
tion, SDSL EPR can significantly contribute in the investigation of those systems.
To showcase the use of SDSL EPR in this field, recent results on a-synuclein being
a canonical model among the IDPs are reviewed.
Usually, nitroxides are used as spin-labels in SDSL EPR [5, 6]. However, with
increasing spread of high-field EPR the relevance of other paramagnetic labels, e.g.,
Gd3+, could gain increased relevance [7, 8]. Nitroxides are stable free radicals of the
general form O-NR1R2. The unpaired electron required for EPR detection is (de-)
localized on the NO bond, about 40% of the spin density at the nitrogen atom and
60% at the oxygen atom. Since many biological macromolecules are diamagnetic,
the nitroxide resonance is most often the only signal in the EPR spectrum. Nitroxide
radicals are very stable, e.g., they can be stored for months under ordinary
conditions. This stability is mainly due to steric protection of the NO bond.
Other parts of the molecule can be modified in order to attach the radical covalently
to larger molecules [9].
Nitroxides can be used either as spin probes or as spin labels. Spin probes are
subject to non-covalent interactions with the system under study. In many cases, spin
probes very similar to one component of the system, e.g., spin-labeled lipids, are
introduced. In contrast, spin labels are covalently linked to a complex structure, in
many cases to a specific site, e.g., of a protein (SDSL) [10]. An alternative approach
includes spin labeling of ligands interacting with the protein under study [11, 12].
The most common spin labeling strategy for proteins uses cysteine substitution
mutagenesis followed by modification of the unique sulfhydryl group with specific
EPR in Protein Science 93
From EPR experiments with spin labels, four primary parameters are obtained: (1)
side chain mobility, (2) distances to other paramagnetic centers, e.g., a second
spin label or a metal ion within the very same or another molecule, (3) solvent
94 M. Drescher
and oxygen accessibility, and (4) a measure for polarity of the environment of the
spin-label.
In the following sections methods for obtaining the first three types of informa-
tion are explained. Polarity measurements are particularly useful for membrane
proteins and are reviewed in [148].
The typical CW EPR spectrum of nitroxide labels in X band consists of three lines
due to the hyperfine interaction with the 14N nucleus featuring a nuclear spin I 1.
As a consequence of the common experimental procedure which includes signal
detection using a lock-in technique, EPR spectra are usually shown as first
derivatives of the absorption spectrum (Fig. 2). In a homogeneous magnetic field
the EPR signal does not depend on the spatial position of the label, and therefore
motion and translational diffusion cannot be detected unless applying magnetic
field gradients [27, 28]. However, since the Zeeman interaction and, in particular,
the hyperfine interaction of nitroxides are anisotropic, the EPR signal is sensitive to
the molecular orientation of the label with respect to the external magnetic field.
Thus, rotational diffusion can generally be detected by EPR.
Rotation of the label with correlation times of the order of the inverse width of
the nitroxide spectrum (5.5 ns) partially averages the anisotropy, resulting in
spectral changes (Fig. 2). Slight narrowing of the spectrum is detectable up to
rotational correlation times of 1 ms and anisotropy-related line broadening is
detectable down to rotational correlation times of 10 ps. Therefore, SDSL EPR
is sensitive to dynamics on the picosecond to microsecond timescales, covering a
variety of the important biological molecular mechanisms such as the dynamics of
proteins in solution [2931].
The spectra do not directly report on the dynamics of the labeled macromolecule
as a whole but contain information on three types of motion: (1) internal motion of
the nitroxide about the chemical bonds of the linker (cf. Fig. 1, left), (2) motion of
the site of attachment relative to the rest of the macromolecule (conformational
flexibility), and (3) motion of the macromolecule as a whole. The internal motion of
the label may be restricted by the environment, depending on the extent to which
the molecular environment engulfs the label. These three dynamic components
significantly complicate the spectral analysis. However, a spectrum can often be
approximated by a simple motional model to provide information on the properties
of the macromolecule [32]. Temperature dependent experiments can adjust the
contributions of the different types of motion to the motional properties reported by
the EPR spectra [33].
The accessible timescale depends on the experimental frequency, e.g., the slow
overall and collective motion will show up best at lower operating frequencies and
fast motion will show up best at higher operating frequencies. In particular, high
field EPR can be used to analyze anisotropic motional dynamics.
One EPR spectrum measured at one single frequency does not allow complete
description of the spin label motion. Therefore, multifrequency EPR studies are
preferable to separate various motional modes in a protein according to their
timescales [34, 35].
In order to analyze the rotational mobility of the spin label quantitatively,
spectral simulations are required. Simulation programs for CW EPR spectra are
available [26, 3638]. For the case of fast isotropic motion, approximate values of
the rotational correlation time can be calculated from the line height ratios [39].
A semiquantitative measure for nitroxide mobility is the inverse central line
width [18], another measure is the inverse second moment of the entire spectrum.
Plotting the inverse central line width vs the inverse second moment allows for
distinguishing different topological regions. So, different categories, namely sites
in loops or unfolded regions, sites on the surface of ordered structures, e.g., helices,
or sites that are buried inside the core of a protein can be identified [18, 4043]. The
periodic dependence of mobility along a sequence can be used to identify secondary
structure elements and protein topography [44].
While quantitative labeling is often checked by mass spectrometry, free labels
and labels attached to a macromolecule can be distinguished by EPR mobility
measurements (Fig. 3). For comparison, the correlation time of the unbound, free
label in aqueous solution at room temperature is ~0.05 ns; in the example shown in
Fig. 3 its mobility is reduced to a correlation time of 0.8 ns upon attachment to a
protein fragment. Labels immobilized in well folded proteins feature typical corre-
lation times in the order of several nanoseconds [46].
Fig. 3 Experimental CW
EPR spectra of free label
(black) and the same label
attached to a soluble protein
fragment (TonB) in aqueous
solution at room temperature
in X band. (Experimental data
taken from [45])
By proper data analysis, distance distributions can be obtained [5561] allowing for
analyzing flexible structures or coexisting conformations. Such techniques have
been successfully applied on soluble proteins [6264].
While the complementary technique F orster resonant energy transfer (FRET),
which is widely used for studying distances in proteins requires two different,
relatively large chromophores, which must be chosen according to the expected
distance, EPR distance measurements can be performed using two identical much
smaller nitroxide labels and are precise over a broad range of distances [51, 65, 66].
Depending on the labeling strategy, inter- and intramolecular distances are
accessible. While for measuring intramolecular distance constraints doubly spin
labeled molecules can be used, intermolecular distances can be determined between
singly labeled molecules (Fig. 4).
Distance measurements by EPR rely on the dipoledipole coupling between
spins, which is inversely proportional to the cube of the distance [67]. Additionally,
the dipoledipole coupling also depends on the angle between the spinspin vector
and the magnetic field (Fig. 5a). For oriented samples this angular dependency can
be observed (Fig. 5b). Fast reorientation of the spinspin vector, e.g., fast rotational
diffusion of the doubly labeled protein under investigation, results in averaging
over all possible orientations and cancels the dipoledipole interaction to zero.
Therefore, most often distance measurements are performed in a frozen state upon
shock freezing in glass forming solution, e.g., aqueous buffer solution mixed with
glycerol, resulting in an isotropic orientation distribution. Hence, the dipolar spec-
trum of such a macroscopically isotropic sample consists of a superposition of
dipolar spectra of all possible orientations of the spinspin-vector resulting in a
classic Pake pattern (Fig. 5c).
Folding kinetics of proteins can be determined by a combination of rapid freeze-
quench experiments and SDSL EPR distance measurement or stopped-flow EPR
[68, 69].
In EPR distance measurements, two cases have to be distinguished. For
distances between nitroxide spin labels below 2 nm the dipoledipole coupling
EPR in Protein Science 97
Fig. 4 Cartoon representation of EPR distance measurements. (a) Doubly labeled monomeric
proteins give rise to intra- and intermolecular spinspin interactions. In order to determine
intramolecular distances, experimental data has to be corrected for intermolecular contributions.
(b) Intermolecular distance measurements using singly labeled proteins in protein oligomers or
aggregates. Multiples of the distance are also expected. This may be even more complicated for
different types of aggregation and can be analyzed by studying a series of samples with increasing
content of non-labeled molecules (diamagnetic dilution). (c) To measure intramolecular distances
within oligomers/aggregates, a mixture of doubly labeled and non-labeled proteins can be used
exceeds the inhomogeneous line width of the EPR spectrum caused by unresolved
hyperfine couplings and g-anisotropy. In this case, distances can be derived by CW
EPR. For distances between nitroxide spin labels larger than approximately 2 nm,
the dipoledipole coupling is much smaller than competing interactions of the spin
Hamiltonian. Thus the dipoledipole coupling has to be separated from those larger
interactions, which is usually done by pulsed EPR approaches, among these the
four-pulse double electron electron resonance (DEER) [48, 51, 70, 71]. By using
multiple techniques a more complete picture is obtained than can be found by a
single technique [72]. In the borderline region of applicability of CW EPR and
DEER (1.72.0 nm) both methods have to be combined in order to obtain accurate
inter-nitroxide distances [73]. The lower limit of precise distance measurements is
determined by the exchange coupling between two spins. With respect to the
dipoledipole interaction it can usually be neglected for distances larger than
about 1.2 nm [74, 75]. On the other hand, for much shorter distances characteristic
98 M. Drescher
Fig. 5 (a) The dipoledipole interaction between two spins A and B depends inter alia on the
angle y between the spinspin-vector and the external magnetic field B0. (b) This angular
dependency can be observed for oriented samples as shown with this experimental data for a
biradical in a liquid crystal under different orientations. (c) Simulated dipolar spectrum for a
macroscopically isotropic sample (Pake pattern). Adapted from [66], copyright Wiley-VCH
Verlag GmbH and Co. KGaA. Reproduced with permission
exchange narrowed single line EPR spectra indicate orbital overlap between multi-
ple spin labels in close contact [7678].
Pulsed methods [82] increase the range of distance sensitivity. They can be used to
separate the dipoledipole interaction from other contributions of the spin Hamil-
tonian. At very large available microwave power, distances can be measured well
by double quantum coherence (DQC) that uses a single frequency. With the power
available on commercial spectrometers, double electron electron resonance [DEER,
an acronym which is synonymously used with PELDOR (Pulsed Electron Double
Resonance)] is the more sensitive technique and is thus most widely applied in the
100 M. Drescher
Fig. 7 Pulse sequence of the dead-time free four-pulse DEER experiment. The pulse sequence at
frequency n1 (refocused spin echo) addresses the A spins, only, while the pulse at n2 flips the B
spins. Applying the pump pulse at variable time T results in a modulation of the refocused echo
intensity V (cf. Fig. 8a)
Originally, DEER was introduced as a three-pulse experiment [83, 84]. The dead
time inherent with this pulse sequence prevents one from recording the important
first data points of the DEER curve. Therefore, a dead-time free four-pulse variant
of DEER was introduced [70] and is now used extensively.
The pulse sequence of the four-pulse DEER experiment is shown in Fig. 7. We
consider a system of electron spins A (observer spins) and B (pumped spins)
possessing a resonant microwave frequency n1 and n2, respectively. Spins A and
B are usually chemically identical nitroxide labels. Since the nitroxide spectrum
features a width of approximately 180 MHz, it is possible to apply pulses at two
different frequencies with non-overlapping excitation profiles and subsequently
divide the nitroxides in the sample into A- and B-spins, respectively. Accordingly,
the pulse sequence at n1 addresses the A spins only, while the pulse at n2 flips the
B spins. At frequency n1 a two-pulse Hahn-echo sequence is first applied resulting
in an echo depicted in Fig. 7 with negative phase, which is followed by a third pulse
at this so-called observer frequency which leads to a refocused echo of the observer
spins A. The dipolar interaction can be studied by insertion of an additional p-pulse
at the second frequency n2 between the two p-pulses at n1. This pulse affects the
B spins only. Their inversion leads to a change of the local magnetic field at
the A spins. Applying the pump pulse at variable time T results in a modulation
of the refocused echo intensity V with the frequency of the dipolar coupling
between A and B spins. Plotting V vs T yields the typical DEER curve (Fig. 8a).
EPR in Protein Science 101
a 1.0
b c
1.0
0.8
0.8
V/ V0
V/ V0
0.6
0.6
0.4
0.2 0.4
0.0 2.0 4.0 6.0 8.0 0.0 2.0 4.0 6.0 8.0 2 3 4 5 6
T/s T/s r/nm
Fig. 8 (a) Simulated DEER data for a doubly labeled model system. The intensity V of the
refocused observer echo (cf. Fig. 7) is plotted vs the delay T of the pump pulse (black). The DEER
curve can be corrected for a background signal (red) originating from intermolecular interactions
(cf. Fig. 4a). (b) Dipolar evolution (form factor) derived from DEER data in (a) by correcting for
the intermolecular background. (c) Corresponding distance distribution. Simulations were
performed using DEERAnalysis [61]
Measuring in frozen solution is desired in order to avoid the averaging out of the
dipoledipole interactions and, in particular, the strong decrease in transverse
relaxation time T2 that is induced even by moderate spin label dynamics. Addition-
ally, the proper choice of temperature is important in pulsed EPR to optimize
relaxation rates. T 50 K is ideal for DEER at nitroxides in aqueous solution, so
liquid helium cooling is advantageous [85].
A dramatic increase in sensitivity can be obtained by lengthening the transverse
relaxation time by choosing the right solvent. At low temperatures the transverse
relaxation time is significantly longer in a fully deuterated matrix than in a
protonated one. Deuteration of the underlying protein, as well as the solvent,
extends the transverse relaxation time to a considerable degree and gives enhanced
sensitivity and an extended accessible distance range [86].
Typical sample volumes for X-band measurements are in the order of some
10 mL at minimum concentration of some 10 mM. There is an optimum concentra-
tion depending on the required maximum accessible distance. For distances of up to
2.5 nm, concentrations up to 4 mM can be used; for measuring distances up to 8 nm,
the concentration should not exceed 0.35 mM [66].
While most experiments reported in the literature were performed in X band, Q
band DEER gains increasing attention owing to its superior sensitivity revealing
higher-quality distance data as well as significantly increased sample throughput
[87, 88]. DEER in W band gives access to the relative orientation of spin labels due
to orientation selection at high fields. More precisely, selective excitation by
microwave pulses may unravel if the orientation of the spinspin vector is
correlated to the orientations of the molecular frame of the two nitroxides [89,
90]. However, due to conformational freedom of the labels, such correlation is often
not very strong.
In most cases, EPR distance measurements are performed to determine a dis-
tance within a nanoobject, e.g., the spinspin distance in a doubly labeled protein.
It is desirable to consider an isolated pair of spin labels; therefore dipolar
interactions to spins of neighboring objects, e.g., intermolecular interactions,
102 M. Drescher
This uncertainty can be reduced by molecular modeling of the spin label behav-
ior. Several approaches were made to overcome this problem [64, 102, 103]. For
instance, the program package MMM describes spin labels by a set of alternative
conformations, rotamers, which can be attached without serious clashes with atoms
of other residues or cofactors. The individual rotamers are assigned Boltzmann
populations corresponding to an estimate of the sum of their internal energy and
interaction energy with the protein. All simulations of experiments on spin labels are
then based on the population weighted average over the ensemble of rotamers [104].
Experimental data for DEER experiments can be predicted for a modeled
structure and favorable attachment sites can be predicted by scanning the whole
protein [105].
The most powerful techniques for protein structure determination in general are
X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy. Very
limited structural information is available if these techniques are not applicable.
104 M. Drescher
This holds true for many membrane proteins which are difficult to crystallize or
concentrate; therefore the determination of their structures is one of the most
challenging fields in structural biology. Furthermore, structure determination of
membrane proteins is an important application of SDSL EPR which is reviewed
[33, 111113, 148].
X-ray crystallography and NMR spectroscopy are also less successful in deter-
mination of structure and dynamics of IDPs [114]. IDPs have been recognized as a
unique protein class as well, justified by the clear structural and functional separa-
tion which they have in common, and again, SDSL EPR can significantly contribute
to their characterization as illustrated in the following.
IDPs comprise a large fraction of eukaryotic proteins (>30%). They lack a well-
defined three-dimensional fold and display remarkable conformational flexibility.
This property potentially enables them to be promiscuous in their interactions and
to adapt their structure according to the needed function.
Since structure and dynamics of IDPs drastically depend on the environment,
corresponding details are notoriously difficult to unravel. Because of their inherent
flexibility, IDPs often fail to crystallize in their free form. When crystallization is
successful, it only leads to a snapshot of a single conformation not representing the
whole conformational ensemble [115]. The most common goal of structural studies,
the determination of unique high-resolution structures, is not attainable for IDPs
due to the absence of a well-defined structure.
Upon interaction with other cellular components, IDPs adopt more highly
ordered conformations. These are subject to high-resolution structures in some
cases; however, at least in some cases, the bound states of IDPs remain highly
non-compact and retain substantial mobility [116, 117].
SDSL EPR offers a powerful tool to study IDPs. Nitroxide spin labels introduce
a minimal perturbation of the system, and probe the very local environment of the
label [3, 4, 38, 118]. The higher sensitivity of EPR compared to NMR allows for
much lower concentrations of protein samples.
Several IDPs have already been subjected to SDSL EPR investigations. So EPR
data showed decreased flexibility in a region of residual helical structure in the
disordered C-terminal domain of the measles virus nucleoprotein, and
demonstrated the further ordering of this region upon interaction with a binding
partner [119, 120].
The amyloid b peptide Ab was the subject of several studies [91, 121123];
structural constraints on Ab140 fibrils were obtained from measurements of CW
EPR spectra and determination of spinspin couplings in a series of spin-labeled
cysteine mutant samples. Conclusions about molecular structure and supramolecu-
lar organization were drawn from these data. The observation of co-fibrillization of
Ab140 and Ab142 suggested the absence of large structural differences between
Ab140 and Ab142 fibrils.
Further examples of IDPs studied by SDSL EPR are the prion protein H1 [124],
ubiquitin [125], or serum albumin [12]. In the following, recent results concerning
a-synuclein are reviewed to illustrate the potential of SDSL EPR in the field
of IDPs.
EPR in Protein Science 105
3.1 a-Synuclein
3.1.1 Fibrils
The aggregation of proteins into amyloid fibrils is associated with several neurode-
generative diseases. It is believed that the aggregation of ASYN from monomers
by intermediates into amyloid fibrils is the toxic disease-causative mechanism
of PD.
A large set of singly labeled ASYN derivatives were used in order to investigate
the structural features of ASYN fibrils. Fibrils grown from spin-labeled ASYN
featured a fibril morphology being very similar to fibrils taken from wild type
ASYN as verified by electron microscopy. Additionally, co-mixing of wild type and
labeled ASYN indicated that both species are able to adopt similar structures within
the fibril, confirming that the introduction of a spin label is tolerated remarkably
well in amyloid fibrils [91].
Analyzing the intermolecular spinspin interaction within the fibrils in terms of
dipolar broadening depending on diamagnetic dilution, it was shown that similar
sites from different molecules come into close proximity. While the accuracy of this
analysis was not sufficient to distinguish fully whether parallelism occurred
between strands or sheets (corresponding to distances of 4.7 or 10 A) [127], a
highly ordered and specifically folded central core region of about 70 amino acids
was identified. The N-terminus is structurally more heterogeneous; the C-terminus
consisting of 40 amino acids is completely unfolded [128].
The latter encouraged Chen et al. [127] to employ a C-terminal truncation
mutant of ASYN containing residues 1115. This allowed for optimizing spectral
quality and minimizing components from non-fibrillized protein or other back-
ground labeling possibly due to codon mistranslation [129]. Single-line, exchange
narrowed EPR spectra were observed for the majority of all sites within the core
region of ASYN fibrils. Such exchange narrowing requires the orbital overlap
between multiple spin labels in close contact and therewith confirmed that the
core region of ASYN fibrils is arranged in a parallel, in-register structure wherein
similar residues from different molecules are stacked on top of each other. This core
region extends from residue 36 to residue 98 and is tightly packed. Accessibility
measurements suggested the location of potential b-sheet regions within the
fibril. Furthermore, the data provide structural constraints for generating three-
dimensional models.
Not only misfolding and fibril formation of ASYN but also membrane binding
are of particular interest, especially for unraveling its physiological role. The
N-terminus of ASYN contains 7 repeats, each of which is made up of 11 amino
acids. Sequence analysis suggested that this part is likely to mediate lipid
interactions [130, 131]. NMR studies are limited by the size of the complex under
investigation. Hence, the structural information available concerns NMR studies of
ASYN on micelles [132135].
EPR in Protein Science 107
Fig. 10 Solvent accessibility analysis of singly labeled ASYN derivatives. The ratios of the
accessibilities to O2 and NiEDDA (PO2 and P NiEDDA, respectively) for residues 2590
summarized by the depth parameter F ln(PO2/P NiEDDA), with increasing F values
indicating deeper membrane immersion depth. The blue line indicates the best fit to a cosine
function and the resulting periodicity corresponds to the theoretically predicted periodicity of 3.67
amino acids per turn. Copyright 2005 National Academy of Sciences, USA, reproduced from [136]
a, b, and c are determined by least square fits to the data according to S aS1
bS2 cS3 [126].
Hence, the local binding affinity can be determined. Using this approach a
systematic study varying the charge density of the membrane allowed for a locally
resolved analysis of the proteinmembrane binding affinity. The results showed
that binding of ASYN to artificial phospholipid membranes is initiated by the
N-terminus (Fig. 11) [126].
The NMR structure of ASYN bound to SDS micelles, commonly used for mem-
brane mimicking, revealed a break in the helix, resulting in two antiparallel alpha-
helices [132]. This model was confirmed by distance measurements exploiting
SDSL EPR utilizing 13 different ASYN double mutants each containing 2 spin-
labeled cysteines (horseshoe model, Fig. 12) [137].
In this study, one mutant includes a pair of cysteines placed within a single helix
to provide an internal distance control. Distance distributions were obtained by
DEER measurements and Tikhonov regularization. Studying ASYN bound to
detergent and lysophospholipid micelles, it has been shown that the inter-helical
EPR in Protein Science 109
Fig. 11 Spin-label EPR revealed that ASYN membrane binding is triggered by its N-terminus.
Schematic representation of ASYN at the membranewater interface. Positions of spin labels used
in this study are depicted as red circles, and the number of the labeled residue is given.
Representing electrostatic interactions, the cationic residue K80 is shown as a white circle.
Adapted from [126]
Fig. 12 Cartoon
representation of the two
helices and linker region of
ASYN bound to an ellipsoidal
micelle, illustrating the
different distances measured
using pulsed EPR (taken
from [137])
separation between the two helices formed upon binding to micelles is dependent
on micelle composition, with micelles formed from longer acyl chains leading to an
increased splaying of the two helices. The distance constraints were in accord with
the NMR data. The data suggested that the topology of ASYN is not strongly
constrained by the linker region between the two helices and instead depends on the
geometry of the surface to which the protein is bound.
The geometry of micelles, however, differs significantly from those of biological
membranes. Micelles have typical diameters of 5 nm and therefore may be too
small to accommodate ASYN in the extended conformation (around 15 nm for an
extended helix of 100 residues). Therefore, it had been postulated that the small size
of the micelles may have artificially constrained the protein into a horseshoe
structure.
A subsequent study considered two selected possible conformations for ASYN
bound to SUVs, namely an extended helix and the horseshoe structure. Theoreti-
cally expected spinspin distance distributions for doubly labeled ASYN taking the
possible rotamers of the spin labels into account were calculated. This enabled one
to identify label positions in the crucial location close to the potential linker region
between the two horseshoe helices which would allow distinguishing between these
conformations by CW EPR distance measurements. CW EPR spectra of corre-
spondingly labeled ASYN bound to POPC SUVs were measured and, using the
theoretical distance distributions, calculated. The authors interpreted their results in
110 M. Drescher
Fig. 13 The distance distribution for ASYN bound to LUV and labeled at residue 27 and 56
clearly consists of two contributions. The shorter distance agrees well with the expected distance
of 2.7 nm for the horseshoe conformation derived from the NMR structure (pdb access code
1XQ8) while the longer distance is consistent with an extended alpha-helix. Taken from [140]
such a way that an unbroken helical structure around residue 40 was ruled out and
confirmed the picture of the interhelix region characterized by conformational
disorder [138]. Later, a close inspection of the data resulted in suggesting that the
measured distances may be more consistent with an extended helix form than with
the horseshoe model [139].
ASYN bound to vesicles, bicelles, and rodlike micelles was also studied by
DEER allowing for measuring longer distances. Jao et al. reported results
suggesting an extended helix conformation being significantly different from that
of ASYN in the presence of SDS micelles. Their DEER study showed that for
several double mutants the average distance per residue was 1% of that for an
alpha-helix, which argues strongly for an extended helix [137].
Already in this study, a number of samples have yielded somewhat bimodal
distance distributions, suggesting distinct conformations of the protein. Actually,
this was confirmed by a further DEER study [140], which used the ability of DEER
to measure distance distributions for direct evidence of coexisting horseshoe and
extended helix conformations of membrane bound ASYN (Fig. 13).
A DEER study [141] measuring distances of up to 8.7 nm showed that the PD-
linked ASYN mutations also remain capable of adopting both structures, and that
the protein to lipid ratio determines whether the protein adopts the broken or
extended helix conformation.
This ability of ASYN to adopt different structures can provide an explanation of
the disparate results obtained using similar experimental techniques and often with
only slight variations in experimental conditions reported in the literature
[99, 133136, 138, 139, 142].
A complementary approach studying proteinmembrane interaction by SDSL
EPR is to utilize spin-labeled lipids. In the case of ASYN, different restrictions of
EPR in Protein Science 111
segmental motion in the chains of different lipids were observed upon ASYN
binding. This observation indicates that ASYN associates at the interfacial region
of the bilayer where it may favor a local concentration of certain phospholipids
[143146].
Using singly spin labeled mutants, intermolecular distances are also accessible by
DEER. It was shown that ASYN may influence the membrane structure and even
disrupt membranes. Under those conditions intra- and intermolecular distance
measurements by DEER allowed for the conclusion that ASYN forms aggregates
once in contact with SUVs [99, 147].
The simplest model for such an aggregate was proposed based on the distance
constraints (Fig. 14). In these aggregates, two ASYN molecules are in close contact,
but they could form part of a larger aggregate in which the proteins are arranged in
an ordered fashion.
4 Concluding Remarks
SDSL EPR has developed as a powerful tool in order to study structure and
dynamics of bio-macromolecules. Mobility and distance measurements being sen-
sitive to dynamics on the picosecond to microsecond timescales, covering the
Fig. 14 Left: ASYN forms well-defined aggregates with lipids. In these aggregates, two ASYN
molecules are in close contact, but they could form part of a larger aggregate in which the proteins
are arranged in an ordered fashion. Spin labels are depicted as red circles. Right: Cartoon
representation of EPR distance constraints. Taken from [141]
112 M. Drescher
References
32. Brustolon M (2008) What can be studied with electron paramagnetic resonance? In: Electron
paramagnetic resonance. Wiley, Hoboken, New Jersey
33. Strancar J, Kavalenka A, Urbancic I, Ljubetic A, Hemminga MA (2010) SDSL-ESR-based
protein structure characterization. Eur Biophys J Biophys Lett 39:499511
34. Zhang ZW, Fleissner MR, Tipikin DS, Liang ZC, Moscicki JK, Earle KA, Hubbell WL,
Freed JH (2010) Multifrequency electron spin resonance study of the dynamics of spin
labeled T4 lysozyme. J Phys Chem B 114:55035521
35. Liang ZC, Freed JH (1999) An assessment of the applicability of multifrequency ESR to
study the complex dynamics of biomolecules. J Phys Chem B 103:63846396
36. Hanson GR, Gates KE, Noble CJ, Griffin M, Mitchell A, Benson S (2004) XSophe-Sophe-
XeprView (R). A computer simulation software suite (v. 1.1.3) for the analysis of continuous
wave EPR spectra. J Inorg Biochem 98:903916
37. Bax A (1989) Two-dimensional NMR and protein-structure. Annu Rev Biochem 58:223256
38. Belle V, Fournel A, Woudstra M, Ranaldi S, Prieri F, Thome V, Currault J, Verger R,
Guigliarelli B, Carriere F (2007) Probing the opening of the pancreatic lipase lid using site-
directed spin labeling and EPR spectroscopy. Biochemistry 46:22052214
39. Lund J, Dalton H (1985) Further characterization of the FAD and FE2S2 redox centers of
component-C, the NADH acceptor reductase of the soluble methane monooxygenase
of methylococcus-capsulatus (BATH). Europ J Biochem 147:291296
40. Isas JM, Langen R, Haigler HT, Hubbell WL (2002) Structure and dynamics of a helical hairpin
and loop region in annexin 12: a site-directed spin labeling study. Biochemistry 41:14641473
41. Margittai M, Fasshauer D, Pabst S, Jahn R, Langen R (2001) Homo- and heterooligomeric
SNARE complexes studied by site-directed spin labeling. J Biol Chem 276:1316913177
42. Jao CC, Der-Sarkissian A, Chen J, Langen R (2004) Structure of membrane-bound alpha-
synuclein studied by site-directed spin labeling. Proc Natl Acad Sci USA 101:83318336
43. Alexander N, Bortolus M, Al-Mestarihi A, McHaourab H, Meilerl J (2008) De novo high-
resolution protein structure determination from sparse spin-labeling EPR data. Structure
16:181195
44. Gross A, Columbus L, Hideg K, Altenbach C, Hubbell WL (1999) Structure of the KcsA
potassium channel from Streptomyces lividans: a site-directed spin labeling study of the
second transmembrane segment. Biochemistry 38:1032410335
45. Domingo Kohler S, Weber A, Howard SP, Welte W, Drescher M (2010) The proline-rich
domain of TonB possesses an extended polyproline II-like conformation of sufficient length
to span the periplasm of Gram-negative bacteria. Prot Sci 19:625630
46. Steinhoff HJ (1990) Residual motion of hemoglobin-bound spin labels and protein dynamics
viscosity dependence of the rotational correlation times. Eur Biophys J 18:5762
47. Constantine KL (2001) Evaluation of site-directed spin labeling for characterizing protein-
ligand complexes using simulated restraints. Biophys J 81:12751284
48. Jeschke G, Polyhach Y (2007) Distance measurements on spin-labelled biomacromolecules
by pulsed electron paramagnetic resonance. Phys Chem Chem Phys 9:18951910
49. Steinhoff HJ, Suess B (2003) Molecular mechanisms of gene regulation studied by site-
directed spin labeling. Methods 29:188195
50. Rabenstein MD, Shin YK (1995) Determination of the distance between 2 spin labels
attached to a macromolecule. Proc Natl Acad Sci USA 92:82398243
51. Jeschke G, Bender A, Paulsen H, Zimmermann H, Godt A (2004) Sensitivity enhancement in
pulse EPR distance measurements. J Magn Reson 169:112
52. Godt A, Schulte M, Zimmermann H, Jeschke G (2006) How flexible are poly(para-
phenyleneethynylene)s? Angew Chem Int Ed 45:75607564
53. Borbat PP, Davis JH, Butcher SE, Freed JH (2004) Measurement of large distances in
biomolecules using double-quantum filtered refocused electron spin-echoes. J Am Chem
Soc 126:77467747
54. Berliner LJ, Eaton SS, Eaton GR (2002) Distance measurements in biological systems by
EPR. In: Biological magnetic resonance, vol 19. Academic, New York
EPR in Protein Science 115
55. Jeschke G, Koch A, Jonas U, Godt A (2002) Direct conversion of EPR dipolar time evolution
data to distance distributions. J Magn Reson 155:7282
56. Bowman MK, Maryasov AG, Kim N, DeRose VJ (2004) Visualization of distance distribu-
tion from pulsed double electron-electron resonance data. Appl Magn Reson 26:2339
57. Jeschke G, Panek G, Godt A, Bender A, Paulsen H (2004) Data analysis procedures for pulse
ELDOR measurements of broad distance distributions. Appl Magn Reson 26:223244
58. Milov AD, Tsvetkov YD, Formaggio F, Oancea S, Toniolo C, Raap J (2004) Solvent effect on
the distance distribution between spin labels in aggregated spin labeled trichogin GA IV
dimer peptides as studied by pulsed electron-electron double resonance. Phys Chem Chem
Phys 6:35963603
59. Chiang YW, Borbat PP, Freed JH (2005) The determination of pair distance distributions by
pulsed ESR using Tikhonov regularization. J Magn Reson 172:279295
60. Chiang YW, Borbat PP, Freed JH (2005) Maximum entropy: a complement to Tikhonov
regularization for determination of pair distance distributions by pulsed ESR. J Magn Reson
177:184196
61. Jeschke G, Chechik V, Ionita P, Godt A, Zimmermann H, Banham J, Timmel CR, Hilger D,
Jung H (2006) DeerAnalysis2006 a comprehensive software package for analyzing pulsed
ELDOR data. Appl Magn Reson 30:473498
62. Zhou Z, DeSensi SC, Stein RA, Brandon S, Dixit M, McArdle EJ, Warren EM, Kroh HK,
Song LK, Cobb CE, Hustedt EJ, Beth AH (2005) Solution structure of the cytoplasmic
domain of erythrocyte membrane band 3 determined by site-directed spin labeling. Biochem-
istry 44:1511515128
63. Nakamura M, Ueki S, Hara H, Arata T (2005) Calcium structural transition of human cardiac
troponin C in reconstituted muscle fibres as studied by site-directed spin labelling. J Mol Biol
348:127137
64. Fajer PG (2005) Site directed spin labelling and pulsed dipolar electron paramagnetic
resonance (double electron-electron resonance) of force activation in muscle. J Phys Condens
Mat 17:S1459S1469
65. Jeschke G, Abbott RJM, Lea SM, Timmel CR, Banham JE (2006) The characterization of
weak protein-protein interactions: evidence from DEER for the trimerization of a von
Willebrand factor A domain in solution. Angew Chem Int Ed 45:10581061
66. Jeschke G (2002) Distance measurements in the nanometer range by pulse EPR.
ChemPhysChem 3:927932
67. Jeschke G, Spiess HW (2006) Distance measurements in solid-state NMR and EPR spectros-
copy. In: Dolinsek J, Vilfan M, Zumer S (eds) Novel NMR and EPR techniques, vol 684.
Lecture Notes in Physics, Springer, Berlin/Heidelberg
68. Dockter C, Volkov A, Bauer C, Polyhach Y, Joly-Lopez Z, Jeschke G, Paulsen H (2009)
Refolding of the integral membrane protein light-harvesting complex II monitored by pulse
EPR. Proc Natl Acad Sci USA 106:1848518490
69. Qu KB, Vaughn JL, Sienkiewicz A, Scholes CP, Fetrow JS (1997) Kinetics and motional
dynamics of spin-labeled yeast iso-1-cytochrome c.1. Stopped-flow electron paramagnetic
resonance as a probe for protein folding/unfolding of the C-terminal helix spin-labeled at
cysteine 102. Biochemistry 36:28842897
70. Pannier M, Veit S, Godt A, Jeschke G, Spiess HW (2000) Dead-time free measurement of
dipole-dipole interactions between electron spins. J Magn Reson 142:331340
71. Martin RE, Pannier M, Diederich F, Gramlich V, Hubrich M, Spiess HW (1998) Determina-
tion of end-to-end distances in a series of TEMPO diradicals of up to 2.8 nm length with a
new four-pulse double electron electron resonance experiment. Angew Chem Int Ed
37:28342837
72. Persson M, Harbridge JR, Hammarstrom P, Mitri R, Martensson LG, Carlsson U, Eaton GR,
Eaton SS (2001) Comparison of electron paramagnetic resonance methods to determine
distances between spin labels on human carbonic anhydrase II. Biophys J 80:28862897
116 M. Drescher
73. Banham JE, Baker CM, Ceola S, Day IJ, Grant GH, Groenen EJJ, Rodgers CT, Jeschke G,
Timmel CR (2008) Distance measurements in the borderline region of applicability of CW
EPR and DEER: a model study on a homologous series of spin-labelled peptides. J Magn
Reson 191:202218
74. Jeschke G (2002) Determination of the nanostructure of polymer materials by electron
paramagnetic resonance spectroscopy. Macromol Rapid Commun 23(4):227246
75. Riplinger C, Kao JPY, Rosen GM, Kathirvelu V, Eaton GR, Eaton SS, Kutateladze A, Neese
F (2009) Interaction of radical pairs through-bond and through-space: scope and limitations
of the point-dipole approximation in electron paramagnetic resonance spectroscopy. J Am
Chem Soc 131(29):1009210106
76. Margittai M, Langen R (2004) Template-assisted filament growth by parallel stacking of tau.
Proc Natl Acad Sci USA 101:1027810283
77. Margittai M, Langen R (2006) Side chain-dependent stacking modulates tau filament struc-
ture. J Biol Chem 281:3782037827
78. Molin YN, Salikhov KM, Zamaraev KI (1980) Spin exchange. Springer, Berlin
79. Altenbach C, Oh KJ, Trabanino RJ, Hideg K, Hubbell WL (2001) Estimation of inter-residue
distances in spin labeled proteins at physiological temperatures: experimental strategies and
practical limitations. Biochemistry 40:1547115482
80. Scarpelli F, Drescher M, Rutters-Meijneke T, Holt A, Rijkers DTS, Killian JA, Huber M
(2009) Aggregation of transmembrane peptides studied by spin-label EPR. J Phys Chem B
113:1225712264
81. Steinhoff HJ, Radzwill N, Thevis W, Lenz V, Brandenburg D, Antson A, Dodson G, Wollmer
A (1997) Determination of interspin distances between spin labels attached to insulin:
comparison of electron paramagnetic resonance data with the X-ray structure. Biophys J 73
(6):32873298
82. Schweiger A, Jeschke G (2005) Principles of pulse electron paramagnetic resonance. Oxford
University Press, Oxford, reprinted 2005 edn
83. Milov AD, Ponomarev AB, Tsvetkov YD (1984) Electron electron double-resonance in
electron-spin echo model biradical systems and the sensitized photolysis of decalin.
Chem Phys Lett 110(1):6772
84. Milov AD, Salikohov KM, Shirov MD (1981) Application of endor in electron-spin echo for
paramagnetic center space distribution in solids. Fiz Tverd Tela 23(4):975982
85. Jeschke G (2011) DeerAnalysis2011 user manual. http://wwweprethzch/software/
DeerAnalysis_2011_manual.pdf. Accessed 7 Apr 2011
86. Richard W, Bowman A, Sozudogru E, El-Mkami H, Owen-Hughes T, Norman DG (2010)
EPR distance measurements in deuterated proteins. J Magn Reson 207:164167
87. Zou P, Mchaourab HS (2010) Increased sensitivity and extended range of distance
measurements in spin-labeled membrane proteins: Q-band double electron-electron
resonance and nanoscale bilayers. Biophys J 98(6):L18L20. doi:DOI 10.1016/j.
bpj.2009.12.4193
88. Hofer P, Heilig R, Schmalbein D (2003) The superQ-FT accessory for pulsed EPR, ENDOR
and ELDOR at 34 GHz. Bruker SpinReport 152(153):3743
89. Larsen RG, Singel DJ (1993) Double electron-electron resonance spin-echo modulation
spectroscopic measurement of electron-spin pair separations in orientationally disordered
solids. J Chem Phys 98:51345146
90. Lovett JE, Bowen AM, Timmel CR, Jones MW, Dilworth JR, Caprotti D, Bell SG, Wong LL,
Harmer J (2009) Structural information from orientationally selective DEER spectroscopy.
Phys Chem Chem Phys 11:68406848
91. Torok M, Milton S, Kayed R, Wu P, McIntire T, Glabe CG, Langen R (2002) Structural and
dynamic features of Alzheimers A beta peptide in amyloid fibrils studied by site-directed
spin labeling. J Biol Chem 277:4081040815
92. Jeschke G, Sajid M, Schulte M, Godt A (2009) Three-spin correlations in double electron-
electron resonance. Phys Chem Chem Phys 11:65806591
EPR in Protein Science 117
93. Hilger D, Jung H, Padan E, Wegener C, Vogel KP, Steinhoff HJ, Jeschke G (2005) Assessing
oligomerization of membrane proteins by four-pulse DEER: pH-dependent dimerization of
NhaA Na+/H + antiporter of E-coli. Biophys J 89:13281338
94. Upadhyay AK, Borbat PP, Wang J, Freed JH, Edmondson DE (2008) Determination of the
oligomeric states of human and rat monoamine oxidases in the outer mitochondrial mem-
brane and octyl beta-D-glucopyranoside micelles using pulsed dipolar electron spin reso-
nance spectroscopy. Biochemistry 47:15541566
95. Bode BE, Margraf D, Plackmeyer J, Durner G, Prisner TF, Schiemann O (2007) Counting the
monomers in nanometer-sized oligomers by pulsed electron electron double resonance.
J Am Chem Soc 129:67366745
96. Jeschke G, Schlick S (2006) Spatial distribution of stabilizer-derived nitroxide radicals
during thermal degradation of poly(acrylonitrile-butadiene-styrene) copolymers: a unified
picture from pulsed ELDOR and ESR imaging. Phys Chem Chem Phys 8:40954103
97. Domingo Kohler S, Spitzbarth M, Diederichs K, Exner TE, Drescher M (2011) A short note
on the analysis of distance measurements by electron paramagnetic resonance. J Magn Reson
208:167170
98. Pannier M, Schops M, Schadler V, Wiesner U, Jeschke G, Spiess HW (2001) Characteriza-
tion of ionic clusters in different ionically functionalized diblock copolymers by CW EPR
and four-pulse double electron-electron resonance. Macromolecules 34:55555560
99. Drescher M, Veldhuis G, van Rooijen BD, Milikisyants S, Subramaniam V, Huber M (2008)
Antiparallel arrangement of the helices of vesicle-bound alpha-synuclein. J Am Chem Soc
130:77967797
100. Sajid M, Jeschke G, Wiebcke M, Godt A (2009) Conformationally unambiguous spin
labeling for distance measurements. Chemistry 15:1296012962
101. Borbat PP, McHaourab HS, Freed JH (2002) Protein structure determination using long-
distance constraints from double-quantum coherence ESR: study of T4 lysozyme. J Am
Chem Soc 124:53045314
102. Sale K, Sar C, Sharp KA, Hideg K, Fajer PG (2002) Structural determination of spin label
immobilization and orientation: a Monte Carlo minimization approach. J Magn Reson
156:104112
103. Fajer P, Likai S, Liu YS, Perozo E, Budil D, Sale K (2004) Molecular modeling tools for
dipolar EPR. Biophys J 86:191A
104. Polyhach Y, Bordignon E, Jeschke G (2011) Rotamer libraries of spin labelled cysteines for
protein studies. Phys Chem Chem Phys 13(6):23562366. doi:Doi 10.1039/C0cp01865a
105. Polyhach Y, Jeschke G (2010) Prediction of favourable sites for spin labelling of proteins.
Spectroscopy 24:651659
106. Hubbell WL, Altenbach C, Hubbell CM, Khorana HG (2003) Rhodopsin structure, dynamics,
and activation: a perspective from crystallography, site-directed spin labeling, sulfhydryl
reactivity, and disulfide cross-linking. Adv Protein Chem 63:243290
107. Altenbach C, Froncisz W, Hyde JS, Hubbell WL (1989) Conformation of spin-labeled
melittin at membrane surfaces investigated by pulse saturation recovery and continuous
wave power saturation electron-paramagnetic resonance. Biophys J 56:11831191
108. Altenbach C, Flitsch SL, Khorana HG, Hubbell WL (1989) Structural studies on transmem-
brane proteins 2 spin labeling of bacteriorhodopsin mutants at unique cysteines. Biochemistry
28:78067812
109. Percival PW, Hyde JS (1975) Pulsed EPR spectrometer 2. Rev Sci Instrum 46:15221529
110. Yin JJ, Pasenkiewiczgierula M, Hyde JS (1987) Lateral diffusion of lipids in membranes by
pulse saturation recovery electron-spin-resonance. Proc Natl Acad Sci USA 84:964968
111. Lacapere JJ, Pebay-Peyroula E, Neumann JM, Etchebest C (2007) Determining membrane
protein structures: still a challenge! Trends Biochem Sci 32:259270
112. Torres J, Stevens TJ, Samso M (2003) Membrane proteins: the Wild West of structural
biology. Trends Biochem Sci 28:174
118 M. Drescher
113. Torres J, Stevens TJ, Samso M (2003) Membrane proteins: the Wild West of structural
biology. Trends Biochem Sci 28:137144
114. Tompa P (2002) Intrinsically unstructured proteins. Trends Biochem Sci 27:527533
115. Timsit Y, Allemand F, Chiaruttini C, Springer M (2006) Coexistence of two protein folding
states in the crystal structure of ribosomal protein L20. EMBO Rep 7:10131018
116. Eliezer D (2009) Biophysical characterization of intrinsically disordered proteins. Curr Opin
Struct Biol 19:2330
117. Dyson HJ, Wright PE (2002) Coupling of folding and binding for unstructured proteins.
Curr Opin Struct Biol 12:5460
118. Biswas R, Kuhne H, Brudvig GW, Gopalan V (2001) Use of EPR spectroscopy to study
macromulecular structure and function. Sci Prog 84:4568
119. Morin B, Bourhis JM, Belle V, Woudstra M, Carriere F, Guigliarelli B, Fournel A, Longhi S
(2006) Assessing induced folding of an intrinsically disordered protein by site-directed spin-
labeling electron paramagnetic resonance spectroscopy. J Phys Chem B 110:2059620608
120. Belle V, Rouger S, Costanzo S, Liquiere E, Strancar J, Guigliarelli B, Fournel A, Longhi S
(2008) Mapping alpha-helical induced folding within the intrinsically disordered C-terminal
domain of the measles virus nucleoprotein by site-directed spin-labeling EPR spectroscopy.
Proteins 73:973988
121. Murakami K, Hara H, Masuda Y, Ohigashi H, Irie K (2007) Distance measurement between
Tyr10 and Met35 in amyloid beta by site-directed spin-labeling ESR spectroscopy: implica-
tions for the stronger neurotoxicity of A beta 42 than A beta 40. Chembiochem 8:23082314
122. Iurascu MI, Cozma C, Tomczyk N, Rontree J, Desor M, Drescher M, Przybylski M (2009)
Structural characterization of b-amyloid oligomer-aggregates by ion mobility mass spec-
trometry and electron spin resonance spectroscopy. Anal Bioanal Chem 395:25092519
123. Sepkhanova I, Drescher M, Meeuwenoord NJ, Limpens R, Koning RI, Filippov DV, Huber M
(2009) Monitoring Alzheimer amyloid peptide aggregation by EPR. Appl Magn Reson
36:209222
124. Lundberg KM, Stenland CJ, Cohen FE, Prusiner SB, Millhauser GL (1997) Kinetics and
mechanism of amyloid formation by the prion protein H1 peptide as determined by time-
dependent ESR. Chem Biol 4:345355
125. Igarashi R, Sakai T, Hara H, Tenno T, Tanaka T, Tochio H, Shirakawa M (2010) Distance
determination in proteins inside Xenopus laevis oocytes by double electron-electron reso-
nance experiments. J Am Chem Soc 132:82288229
126. Drescher M, Godschalk F, Veldhuis G, van Rooijen BD, Subramaniam V, Huber M (2008)
Spin-label EPR on alpha-synuclein reveals differences in the membrane binding affinity of
the two antiparallel helices. Chembiochem 9:24112416
127. Chen M, Margittai M, Chen J, Langen R (2007) Investigation of alpha-synuclein fibril
structure by site-directed spin labeling. J Biol Chem 282:2497024979
128. Der-Sarkissian A, Jao CC, Chen J, Langen R (2003) Structural organization of alpha-
synuclein fibrils studied by site-directed spin labeling. J Biol Chem 278:3753037535
129. Masuda M, Dohmae N, Nonaka T, Oikawa T, Hisanaga SI, Goedert M, Hasegawa M (2006)
Cysteine misincorporation in bacterially expressed human alpha-synuclein. FEBS Lett
580:17751779
130. George JM, Jin H, Woods WS, Clayton DF (1995) Characterization of a novel protein
regulated during the critical period for song learning in the zebra finch. Neuron 15:361372
131. Davidson WS, Jonas A, Clayton DF, George JM (1998) Stabilization of alpha-synuclein
secondary structure upon binding to synthetic membranes. J Biol Chem 273:94439449
132. Ulmer TS, Bax A, Cole NB, Nussbaum RL (2005) Structure and dynamics of micelle-bound
human alpha-synuclein. J Biol Chem 280:95959603
133. Chandra S, Chen XC, Rizo J, Jahn R, Sudhof TC (2003) A broken alpha-helix in folded
alpha-synuclein. J Biol Chem 278:1531315318
134. Bussell R, Eliezer D (2003) A structural and functional role for 11-mer repeats in alpha-
synuclein and other exchangeable lipid binding proteins. J Mol Biol 329:763778
EPR in Protein Science 119
135. Bussell R, Ramlall TF, Eliezer D (2005) Helix periodicity, topology, and dynamics of
membrane-associated alpha-synuclein. Prot Sci 14:862872
136. Jao CC, Hegde BG, Chen J, Haworth IS, Langen R (2008) Structure of membrane-bound
alpha-synuclein from site-directed spin labeling and computational refinement. Proc Natl
Acad Sci USA 105:1966619671
137. Borbat P, Ramlall TF, Freed JH, Eliezer D (2006) Inter-helix distances in lysophospholipid
micelle-bound alpha-synuclein from pulsed ESR measurements. J Am Chem Soc
128:1000410005
138. Bortolus M, Tombolato F, Tessari I, Bisaglia M, Mammi S, Bubacco L, Ferrarini A, Maniero
AL (2008) Broken helix in vesicle and micelle-bound alpha-synuclein: insights from site-
directed spin labeling-EPR experiments and MD simulations. J Am Chem Soc
130:66906691
139. Georgieva ER, Ramlall TF, Borbat PP, Freed JH, Eliezer D (2008) Membrane-bound alpha-
synuclein forms an extended helix: long-distance pulsed ESR measurements using vesicles,
bicelles, and rodlike micelles. J Am Chem Soc 130:1285612857
140. Robotta M, Braun P, van Rooijen B, Subramaniam V, Huber M, Drescher M (2011) Direct
evidence of coexisting horseshoe and extended helix conformations of membrane-bound
alpha-synuclein. Chemphyschem 12:267269
141. Georgieva ER, Ramlall TF, Borbat PP, Freed JH, Eliezer D (2010) The lipid-binding domain
of wild type and mutant alpha-synuclein compactness and interconversation between the
broken and extended helix forms. J Biol Chem 285:2826128274
142. Trexler AJ, Rhoades E (2009) Alpha-synuclein binds large unilamellar vesicles as an
extended helix. Biochemistry 48:23042306
143. Ramakrishnan M, Jensen PH, Marsh D (2003) Alpha-synuclein association with phosphati-
dylglycerol probed by lipid spin labels. Biochemistry 42:1291912926
144. Bussell R, Ramlall TF, Eliezer D (2005) Helix periodicity, topology, and dynamics of
membrane-associated alpha-synuclein. Protein Sci 14(4):862872. doi:Doi 10.1110/
Ps.041255905
145. Kamp F, Beyer K (2006) Binding of alpha-synuclein affects the lipid packing in bilayers of
small vesicles. J Biol Chem 281(14):92519259. doi:DOI 10.1074/jbc.M512292200
146. Ramakrishnan M, Jensen PH, Marsh D (2006) Association of alpha-synuclein and mutants
with lipid membranes: spin-label ESR and polarized IR. Biochemistry 45(10):33863395.
doi:10.1021/bi052344d
147. Drescher M, van Rooijen BD, Veldhuis G, Subramaniam V, Huber M (2010) A stable lipid-
induced aggregate of alpha-synuclein. J Am Chem Soc 132:40804081
148. Bordignon E (2011) Site-directed spin labeling of membrane proteins. Top Curr Chem.
doi:10.1007/128_2011_243
Top Curr Chem (2012) 321: 121158
DOI: 10.1007/128_2011_243
# Springer-Verlag Berlin Heidelberg 2011
Published online: 7 September 2011
Enrica Bordignon
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
2 Spin Labeling of Membrane Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
2.1 Spin Labeling Protocols for Membrane Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
2.2 Reconstitution Protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3 Dynamics of Spin Labeled Side Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1 EPR-Derived Molecular Architecture of the Potassium Channel KcsA . . . . . . . . . . . . 128
3.2 Light-Induced Conformational Changes in Bacteriorhodopsin Detected via
Transient Mobility Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
E. Bordignon
Laboratory of Physical Chemistry, ETH Zurich, Wolfgang-Pauli-str. 10, 8093 Zurich, Switzerland
e-mail: enrica.bordignon@phys.chem.ethz.ch
122 E. Bordignon
1 Introduction
In 1989, 4 years after the first high resolution X-ray structure of a membrane protein
[1], a paper on site-directed spin labeling (SDSL) EPR on a transmembrane protein
appeared in Biochemistry [2], followed by a second report in Science [3] titled
Transmembrane protein structure: spin labeling of bacteriorhodopsin mutants.
Hubbell suggested that a combination of EPR, site-directed mutagenesis and
nitroxide labeling could serve as a new biophysical tool to unravel structural details
of membrane proteins. This first EPR study focused on the different dynamics
(mobilities) of 18 consecutive spin labeled cysteines in the protein and their distinct
water accessibility. The EPR analysis could clearly distinguish between a short
water exposed loop and the transmembrane a-helical segment of the protein. In the
same year the first 3D structure of bacteriorhodopsin (BR) was solved by electron
microscopy [4]. The X-ray structure of BR appeared only later [5]. In 1994 in a
review article, Hubbell further confirmed the potentiality of the technique, stating
that the technology is now in place for an expanded range of applications of
SDSL [6]. Indeed, to date, a plethora of structural information has been obtained
by SDSL EPR on a variety of membrane proteins, confirming the statement of the
pioneer of the technique. The increased availability of structures with atomistic
details from X-ray and NMR did not relegate SDSL EPR to proteins of unknown
structure or to minor aspects of the spin labeled side chains accessibility and
dynamics. On the contrary, it enormously favored its application to conformational
transitions between different protein states, dimerizationoligomerization pro-
cesses, characterization of protein dynamics during the conformational changes
under physiological conditions, and recently also modeling of overall conforma-
tional transitions based on interspin distance constraints.
This chapter will introduce the most important aspects of SDSL EPR applied to
membrane proteins corroborated by prominent examples from the literature and
will address the newest developments of the technique.
Site-Directed Spin Labeling of Membrane Proteins 123
To date, the most commonly used spin labels for membrane proteins are pyrroline-
type nitroxide radicals. Several labeling strategies and spin labels are available at
present. However, the extreme specificity for the free thiol of cysteines, a stoichio-
metric reaction at most sites, and the relatively small size and flexibility of the
modified side chain make the (1-oxyl-2,2,5,5-tetramethyl-D3-pyrroline-3-methyl)
methanethiosulfonate spin label (MTSSL) the most popular choice for SDSL
applications. 3-(2-Iodoacetamido)-2,2,5,5-tetramethyl-1-pyrrolidinyloxy radical
labels (IAP) are also used in combination with cysteines. The advantage of IAP
vs MTSSL is that the covalent bond with the cysteine is irreversibly formed and this
may be more suitable for particular situations in which the MTSSL tends to be
released from the labeled sites.
The labeling strategies for membrane proteins are similar to those applicable to
water soluble proteins (a cys-less background is also required prior to the cys site-
specific mutagenesis); however more care must be taken in identifying sites which
are accessible for the spin label. In particular, sites buried in the middle of the
transmembrane core of the protein may not be labeled at all, or once labeled may
influence the protein stability. Sites located at the interface between the membrane
and the bulk water, e.g., in the short loops connecting the transmembrane helices, or
at cytoplasmic or periplasmic domains, are usually highly accessible and can be
targeted for first tests on the protein under investigation (Fig. 1a). The freely
downloadable software MMM [http://www.epr.ethz.ch/software/index] can be
used to predict the accessibility towards spin labels (currently MTSSL or IAP)
for each site of the protein prior to the site-specific mutagenesis if a structural
template of the protein exists [7].
Experimentally, the spin labeling procedure is carried out on proteins solubilized
in detergent, to maximize the reaction yield. In a few cases (protein- and detergent-
dependent) the spin labels may bind aspecifically to the cys-less protein, or to the
detergent micelles. It is then good practice to spin label prior to the SDSL study the
cys-less protein and eventually also the empty micelles, to verify that noncovalent
aspecific labeling is negligible. Otherwise, one needs to change the spin label to
protein ratio during the labeling procedure or eventually the type of detergent to
minimize this unwanted effect. It is worth mentioning that the labels trapped in the
micelles are removed by the reconstitution procedure, unless the protein is
measured in detergent-solubilized form.
In the following, a general protocol for spin labeling membrane proteins with
MTSSL (also valid for IAP) is described, with particular attention to steps which
may cause problems or may need to be slightly modified according to the protein
124 E. Bordignon
stability. Here it is assumed that the expression and purification of the cys-less
membrane protein of interest has been optimized prior to the spin labeling. The
membrane protein engineered to have one cysteine residue to be spin labeled is
usually prepared in the detergent-solubilized form and its purity is >98%. Knowl-
edge of the protein concentration is mandatory for the spin labeling protocol, and
for the subsequent determination of the spin labeling efficiency. For labeling, the
detergent-solubilized protein is diluted to a concentration of about 1030 mM. In
some cases the protein is incubated for 1 h at room temperature with 10 mM DTT
(1,4-dithio-DL-threitol) to reduce eventually oxidized cysteines. DTT must be
removed before starting the labeling procedure (1 mM final residual concentration)
to avoid nitroxide reduction and to allow the formation of the disulfide bond
between MTSSL and the cys residue. Immediately after DTT removal, tenfold
molar excess of MTSSL is added from a 100 mM stock solution (usually in DMSO,
stored at 80 C). If the final concentration of MTSSL in the water solution
exceeds 500 mM, spin label biradicals can form which are not anymore available
for the labeling of the protein. Usually, the protein-MTSSL solution is shaken
overnight at 4 C, after which the labeling should be complete. Labeling can also
be performed for 45 h at room temperature, depending on the protein stability.
Note that very accessible sites can be labeled after few minutes of incubation. In
general, the labeling efficiency increases with (1) increasing the label-to-protein
molar excess (e.g., up to 40-fold); (2) increasing the temperature (from 4 to 37 C,
depending on the protein stability); (3) slightly denaturating the protein (valid only
for very stable proteins, e.g., 40% DMSO was used to label position 46 in bacterio-
rhodopsin, Fig. 1a); and (4) preincubating the protein with reducing agents. After
Site-Directed Spin Labeling of Membrane Proteins 125
the spin labeling, one proceeds with the removal of the excess free label in the
buffer. Different methods are possible: (1) dialysis slow but efficient, requires at
least 500 mL of detergent-containing buffer, which could be expensive; (2) centrif-
ugal filter devices fast and efficient, but some proteins precipitate on the mem-
brane after centrifugation; (3) desalting columns very fast and efficient,
performed by centrifugation or gravity, requires a larger volume than centrifugal
devices.
Once the free label is removed, the protein is concentrated to the desired final
concentration. At this stage the labeling efficiency can be tested. Continuous wave
EPR is highly sensitive, enabling measurements on as little as 100 pmol of protein
(10 mL of 10 mM spin labeled protein). For optimal signal-to-noise ratio,
concentrations in the order of 100 mM are preferable. The residual free label is
clearly distinguishable in the EPR spectrum from the bound fraction due to its
narrow three-line spectrum typical of isotropic fast motional regimes. In the case
that the spectrum still contains a disturbing fraction of free label, further washing is
necessary. The labeling efficiency can be calculated directly from the continuous
wave EPR spectrum by second integral analysis and comparison with the spectrum
of a standard nitroxide radical of known concentration (typically a 300 mM water
solution of TEMPOL). The ratio between the spin concentration and the protein
concentration gives directly the labeling efficiency for single cys mutants. Labeling
efficiencies in the 70100% range are considered good; problematic sites allows
only for 050% efficiency. Efficiencies >100% pose some questions on the speci-
ficity of the labeling, which must be further tested with the cys-less background
protein. For interspin distance determination (two spin labeled cysteines per pro-
tein) an average efficiency of at least 6070% per cysteine is advisable.
After successful spin labeling, the detergent-solubilized protein can be either used
directly or reconstituted into liposomes or other bilayer-mimicking environments
(Fig. 1b). Reconstitution protocols are strictly protein-specific and several methods
can be found in the literature based on detergent extraction by biobeads, dilution
methods, etc. An important issue in membrane protein reconstitution in general and
most specifically for EPR analysis is the homogeneity of the distribution of the
proteins inside the membranes. In particular, lipid to protein ratios >500 are
advisable to avoid protein crowding which may lead, for example, to unwanted
interprotein interspin distances in the 28 nm range. Extrusion of the liposomes
before or after reconstitution can be performed to increase the homogeneity of the
sample. The type of lipids plays an important role in either determining the
efficiency of the insertion or contributing to the directionality of the insertion
process. In fact, membrane proteins can insert into liposomes both with no prefer-
ential orientation (e.g., 50% of the proteins exposes the cytoplasmic regions to the
lumen and the other 50% the periplasmic region) or with remarkable preference for
126 E. Bordignon
one specific orientation with respect to the lumen (e.g., proteins having long
cytoplasmic domains tend to insert in liposomes leaving those domains exposed
to the bulk water).
An alternative approach consists in reconstitution of the membrane protein in
bilayer-mimicking environments, i.e., self-assembling lipid bilayer nanodiscs
(Fig. 1b). Nanodiscs consist of a small portion of membrane bilayer that has been
solubilized by the addition of two amphipathic proteins, the membrane scaffold
proteins (MSP) derived from the apolipoprotein A-1 [810]. Details of the prepara-
tion can also be found at http://sligarlab.life.uiuc.edu/nanodisc/protocols.html.
These proteins wrap around the hydrophobic core of the lipids, effectively creating
a soluble portion of membrane.
The advantages of this method are: (1) different nanodiscs diameters can be
obtained with different MSP sequences; (2) the protein is in a bilayer-mimicking
environment; (3) both the cytoplasmic and the periplasmic exposed regions of the
protein are easily accessible with ligands, nucleotides, interacting proteins, etc.,
without the problems associated with the directional incorporation of the proteins in
liposomes and the compartmentalization between lumen and bulk water. The
disadvantages are: (1) the reconstitution in nanodiscs is protein-specific and
requires optimization of the protocol and of the MSP to be used; (2) the maximal
concentrations achievable are less than 100 mM.
Room temperature EPR spectra of spin labels attached at protein sites are extremely
sensitive to even minor changes in the reorientational motion of the spin label with
respect to the external magnetic field. The reorientation of the spin label in the
protein is generally anisotropic and it is a complex function of the spin label
molecular structure and the primary, secondary, tertiary, and eventually quaternary
structure of the protein under investigation. Complex distributions of motional
states can often be observed in spectra exhibiting more than one spectral compo-
nent. Simulations of the CW spectra are performed using molecular dynamics
simulations [1115] or a slow motion theory based on the stochastic Liouville
equation (SLE). The theoretical model developed by Freed based on the SLE
(SRLS, Slowly Relaxing Local Structure) is the most advanced to date [16, 17].
Two principle dynamic modes are included in SRLS, the global tumbling of the
protein and the internal spin label motion consisting of side chain isomerizations
and protein backbone fluctuations. In the case of medium-size membrane proteins
embedded in micelles or in membranes, the overall tumbling of the object generally
exceeds the range of rotational correlation times measurable by EPR (>100 ns).
Thus, one can consider the protein as immobile on the EPR timescale, and the
spin label motion only due to internal rotation around the dihedral angles of the
label and to the protein backbone fluctuations. Even with the reduced number of
rotational motions to be considered in the case of membrane proteins, a
Site-Directed Spin Labeling of Membrane Proteins 127
comprehensive theoretical description of the spectra has not yet been achieved. The
SRLS model was shown to be able to separate the effects of faster internal modes of
motion from slower overall motions only if spectra at multiple frequencies are
available. Even with the multifrequency SRSL approach, up to three distinct
spectral components are often required to account for all observed spectral features,
showing the intricate and composite interaction between the label and the protein
environment.
The complications in the spectral fitting partially stem from the fact that MTSSL
is a rather long and flexible label characterized by five rotatable bonds. Clearly a
more rigid label, with the NO group closely related to the backbone atom
positions would alleviate this problem, but it would also create steric constraints
at the spin labeled site, which may disturb the protein fold. In fact, it is exactly its
intrinsic flexibility that makes MTSSL a nonperturbing probe at the majority of the
sites in proteins.
For a semiquantitative description of spin label mobility, simple empirical
parameters extracted from the spectra at room temperature are widely used. Exper-
imentally, to detect room temperature EPR spectra, 1020 mL of sample are
inserted in glass capillaries (0.9 mm outer diameter) and the spectra are detected
with a sweep width of 15 mT, 0.15 mT maximal modulation amplitude
(corresponding to the natural linewidth of a freely tumbling MTSSL in water)
and 0.51 mW microwave power (the power saturation curve must be measured
at least once for the cavity in use). The mobility scale is based on (1) the inverse of
the spectral second moment and (2) the inverse of the central EPR linewidth. Both
parameters are increasing functions of the spin label mobility and can be used to
give insights into the protein secondary structure, based on the available correlation
between mobility and protein structure obtained from a large number of studies,
mainly on a-helical proteins [1820]. Although the semiquantitative analysis is
usually enough to draw conclusions on secondary structure and conformational
changes, interpretative problems may arise if spectra show complicated features
with multiple components, which in the literature are often identified as immobile
and mobile components.
The high sensitivity for relatively fast motions (rotational correlation times in
the 0.13 ns range) extends the structural analysis beyond the usually silent
flexible regions, which are increasingly identified as relevant for signal transduc-
tion. In general, the extent of labeling is correlated to the accessibility of the site
towards the spin label; thus very mobile positions show spin labeling efficiency
close to 100%. Nevertheless, very immobile positions can be labeled due to the
flexible nature of MTSSL, and the plasticity of the side chains surrounding a buried
spin labeled site. Care must be taken when spectra close to the rigid limit are
detected at room temperature; in fact the label can be trapped in one specific
orientation relative to the backbone. This makes general statements about side
chain mobility or distances with respect to a second spin labeled site ambiguous.
Moreover, the protein structure might have been affected by the introduced steric
hindrance. Such interpretative problems can be solved by analyzing the overall
pattern of side chain mobilities in an extended protein segment, i.e., by performing
a nitroxide site scan on consequent residues.
128 E. Bordignon
Details of the overall architecture and of the gating mechanism of the potassium
channel KcsA were elucidated in the last few years with SDSL by the group of
Perozo. Here, one long site scan (63 single cysteines spin labeled with MTSSL)
published in 1998 on this membrane protein is reviewed to highlight the potentiality
of the analysis of the nitroxide mobility to infer structural details of long protein
segments [22]. Some caveats arising from the use of homo-oligomeric proteins are
addressed. Potassium channels are integral membrane proteins which can be found
virtually in all living organisms. They catalyze an efficient flow of K+ across the
membrane bilayer, with a fundamental role in generation and modulation of the
electrical excitability in cells [23].
The data available at the time of the EPR research on the potassium channel
KcsA were: (1) mechanistic information derived from electrical measurements; (2)
tetrameric organization of the monomer (160 residue polypeptide); (3) a-helical
nature of the polypeptide chain; (4) stabilization of the open conformation by low
pH; (5) newly available X-ray structure from the MacKinnons group [24], with the
PDB file not yet released. The systematic EPR scan of the membrane embedded
region of the channel via mobility, accessibility and distance analysis was used to
generate a three-dimensional architecture of the two transmembrane (TM) helices
of the channel and to elucidate their topology with respect to the membranewater
interface. Interestingly, 66 cys mutants were prepared in the protein stretch 2252
and 86120 where the two TM segments were thought to be located (Fig. 2a) and
Site-Directed Spin Labeling of Membrane Proteins 129
only three of them were found not to be expressed, confirming that cysteine
mutagenesis is tolerated without relevant structural rearrangements in most cases;
thus they are good probes to report on local protein properties. All the cys mutants
were tested for structural and functionality studies prior to the EPR study. It is
important to note that functionality tests are mandatory for newly engineered
mutants, and it is recommended to perform them before and after labeling to
avoid interpretative problems. A nonfunctional or nonperfectly folded spin labeled
mutant cannot be used for EPR analysis.
Each single cysteine mutant was spin labeled and analyzed by RT CW EPR
spectroscopy. As an example of a nitroxide scan, spectra detected in the first
(2232) and second (86109) segments are shown in Fig. 2b. Interestingly, being
a homotetramer, each single cysteine mutant provides four spins per channel. If the
distance between the four spin labels in the tetramer is on the order of 2 nm or
higher, no dipolar interaction appears; thus the features of the room temperature
spectra purely reflect the spin label dynamics.
In contrast, for distances <2 nm, the evident spectral broadening due to spin
exchange and dipolar coupling interferes with the mobility analysis (e.g., position
108, Fig. 4b, denoted with an asterisk). To obtain mobility information, an approach
called spin dilution is used, which consists in under-labeling the protein (e.g., in
this study a 1:10 ratio spin label:channel was used) to obtain a spectrum which
reflects mostly the dynamics. Alternatively, to minimize asymmetry, it is possible
to label the protein with a mixture of MTSSL and its diamagnetic analog (several
molecules are available, e.g., the compound A167900 from Toronto Research,
Canada).
In Fig. 4d, the effect of spin dilution is illustrated. Spectral broadening is clearly
evident in the spectral wings of the fully labeled sample. The shortest distance
between two adjacent monomers based on the X-ray structure (PDB 18L8) can now
be simulated with the software MMM (http://www.epr.ethz.ch/software/index)
based on a rotamer library approach [7] showing a mean distance of 1.4 nm, in
agreement with the dipolar broadening observed.
The structural analysis on KcsA was performed based on the mobility of each
spin labeled side chain in the protein segments under investigation. It is worth
recognizing in Fig. 4b that most of the CW RT spectra show multiple spectral
components, characterized by different mobility (a few examples are highlighted
by arrows). This is a very general property of the R1 side chain in proteins. The
components reflect the anisotropy of the spin label reorientational motion, but their
appearance could also have other causes. They could arise from a slow equilibrium
between two different protein conformations or the presence of asymmetric sites in
the protein. The molecular interpretation of different spectral components is cum-
bersome. Multifrequency EPR [17], temperature analysis of the CW spectra [27],
pulse saturation recovery techniques [28], or high pressure EPR [29] can help
unravel the possible origins of the spectral components. In the case of KcsA, the
spin labels motional information was quantitatively extracted from the inverse
central linewidth (DH01, mobility parameter) and was corroborated by the mea-
sure of the accessibility of the spin labeled side chains towards lipids (O2
130 E. Bordignon
Fig. 2 A nitroxide scan on KcsA. (a) Linear representation of the putative transmembrane
topology of KcsA and the nitroxide scan (linear scale with arrows). (b) Room temperature CW
EPR spectra for two regions in TM1 and TM2. Multiple nitroxide components are highlighted by
red arrows in selected spectra. (c) Mobility and accessibility plots. Periodical pattern are visible.
On the right, helical wheel representation showing the trends of the EPR parameters extracted
from the spectra in a polar coordinate representation. (d) Example of dipolar broadening on
position 108, and effect of underlabeling on the spectral shape. On the right the shortest distance
Site-Directed Spin Labeling of Membrane Proteins 131
The dynamics of spin labeled side chains are strictly related to the secondary and
tertiary structure of the protein under investigation. If functional conformational
transitions are induced in proteins with substrates, ligands, pH, nucleotides, ions,
lipids, etc., the concomitant changes in the spin label dynamics describe the protein
rearrangements. SDSL EPR can follow the proteins in action, and being a
technique applicable to proteins in physiological environments, it gives unique
potentiality to unveil the rearrangements of membrane proteins during their func-
tion, e.g., substrate-ions transport, light activation, oligomerization, ligand binding,
etc. In the case of the seven-transmembrane helix protein bacteriorhodopsin, light
excitation induces transient isomerization of the retinal molecule bound in the core
of the protein through a Schiff base. It is the light-induced change in the conforma-
tion of the retinal molecule which results in a conformational change of the
surrounding protein and the proton pumping action during a so-called
Fig. 2 (continued) simulated with MMM on the X-ray structure (PDB 18L8). (e) EPR-derived
molecular architecture of the channel, colored according to the water accessibility scale (red, high
water accessibility; blue, low). Upper panel, intracellular view; bottom panel, lateral view of two
monomers. (f) Crystal structure of the channel (PDB 18L8). Panels ae adapted from [25]
132 E. Bordignon
Fig. 3 Light-induced helix movements in BR and SRII. (a) Schematic representation of the
conformational changes detected by EPR during the photocycle. The difference spectrum reflects
the increase in mobility in state 2 with respect to the ground state. Arrows highlight the character-
istic changes. (b) Left, periplasmic view of BR in ribbon (up) and schematic (down)
representations. The EPR-detected movements are highlighted by arrows. Central panel, EPR
detection of transient change in mobility at one fixed B field position, superimposed to the optical
traces detected at different wavelengths. Right panel, spectrum of the dark-adapted state and
difference spectrum characteristic of a mobility decrease. Adapted from [30]. (c) Analogous
scheme for SRII in complex with the transducer HtrII. Central panel, normalized transient traces
for two positions in SRII. Right panel, dark-adapted spectra with the difference spectra
superimposed (stick representation). Adapted from [33]
134 E. Bordignon
Key steps to pave the way for the application to spin labeled proteins were
performed by the group of Han [41].
X-band nitroxide EPR spectra in the rigid limit (T < 200 K) also convey
information on the nitroxide microenvironment. In fact, the Azz principal value of
the 14N hyperfine tensor (half of the splitting between the positive low field peak
and the negative high field peak of the X-band CW spectrum) is proportional to the
polarity of the nitroxide microenvironment (for MTSSL the Azz value changes from
3.3 to 3.7 mT going from an apolar to a polar environment). This information can be
complemented by detection of the propensity of the nitroxide to form H-bond
(proticity) via the gxx principal value of the g tensor. Sufficiently precise
measurements of gxx require the continuous wave detection of the low temperature
nitroxide spectrum at W band (3.4 T/95 GHz).
MsbA is a putative lipid A flippase from E. coli [42]. It belongs to the ubiquitous
protein superfamily of ABC transporters, active transporters which couple the
vectorial translocation of a variety of substrates across the membranes with ATP
hydrolysis [43]. ABC transporters share a common architecture consisting of two
conserved cytoplasmic modules which harness ATP energy for biological work and
two highly divergent transmembrane domains (TMDs) which provide the translo-
cation channel for each substrate. MsbA has sequence similarity to the subclass
of exporters linked to the development of multidrug resistance and cancer. A
systematic SDSL EPR analysis on MsbA was published in 2009 by the group of
Mchaourab [26], which shed light on the overall conformational changes during the
nucleotide cycle. Here, the mobility and accessibility study performed on the
transporter is reviewed, with particular emphasis on the water accessibility infor-
mation obtained by continuous wave power saturation (examples of power satura-
tion curves for a water exposed residue are presented in Fig. 4e).
MsbA is a homodimer of a polypeptide consisting of fused nucleotide binding
domains (NBDs) and TMDs, a common feature of bacterial ABC exporters. The
substrate is picked up from the inner leaflet of the membrane, and exported to the
outside of the membrane by a conformational change of the transporter. Crystal
structures of MsbA exist in the apo inward-facing conformation (PDB 3B5W, Ca
atoms only) and in the AMPPNP-bound outward-facing conformation (PDB 3B60,
3.7 A resolution). AMPPNP (50 -adenylyl imidodiphosphate) is a nonhydrolyzable
analog of ATP, which traps the transporter in the closed state. The apo state of the
transporter is defined by a V-shaped chamber open to the cytoplasm and the inner
leaflet of the membrane, with the two NBDs separated by about 5 nm (Fig. 4a).
Addition of AMPPNP induces a tight packing of the NBDs which sandwiches the
two nucleotides and induces an opening of the TMDs towards the outer leaflet of the
bilayer.
136 E. Bordignon
Fig. 4 Accessibility data on MsbA reveals the alternating-access mechanism. (a) Structural
model of the open (provided by Chang) and closed states (PDB 3B60). The cytoplasmic region
in the nitroxide scan (green) is highlighted. (b) Accessibility profile of the apo-state. The
cytoplasmic region shows high water accessibility, in line with the structure. (c) Plot of the
water accessibility in water the structure. (d) Changes in water accessibility in the cytoplasmic
region upon transition to the ADP-Vi intermediate. (e) Example of three saturation curves for a
water-exposed side chain to extract the accessibility parameter. Adapted from [26]
Site-Directed Spin Labeling of Membrane Proteins 137
to avoid lineshape artifacts prior to the high field experiments). The variation of
Azz (3.33.7 mT from an apolar aprotic to a polar protic solvent) and
gxx (2.00912.0081 from an apolar aprotic to a polar protic solvent) is observable
at W band. The gxx and Azz variation along the bacteriorhodopsin channel is seen as
a shift of the position of the low field maximum and of the hyperfine splitting in the
high field region, respectively (Fig. 5a). Spectral fitting can be performed with the
function pepper from Easyspin [48]. A distribution of H-bonds around the NO
group, and planar deviations of the NO group, make the gxx region of the W-band
spectra rather broad (gxx strain). Additionally, multiple components can be clearly
identified for several protein positions in the gxx region, arising from nitroxide
subpopulations having 0, 1, and 2 H-bonds [49]. Plots of gxx and Azz vs nitroxide
position along the bacteriorhodopsin proton channel (Fig. 5a, right panel) reveal a
characteristic variation in the polarity of the nitroxide microenvironment along the
channel. Residue S162R1 (R1 denotes the spin labeled side chain) is located in the
EF loop at the cytoplasmic surface, whereas residue K129 is positioned in the DE
loop on the extracellular surface. The high polarity detected is in agreement with
the structure. The polarity of positions 100, 167, and 171 is significantly less and
reaches its minimum at position 46 in the vicinity of the retinal. The plot directly
reflects the hydrophobic barrier which the proton has to overcome on its way
through the proton channel. The 2D plot gxx vs Azz is also used to discriminate
sites with different polarity and proticity [50, 51].
Fig. 5 Polarity and proticity analysis. (a) W-band rigid limit spectra of BR (PDB 1MOL) spin
labeled in the proton channel. The changes in gxx and Azz are highlighted by vertical lines. Right
panel, plots of gxx and Azz vs nitroxide position showing the hydrophobic barrier in the channel.
Adapted from [45]. (b) Structure of LHCIIb (PDB 2BHW) with two IAP spin labels attached with
MMM at positions 52 and 196. Right panel, ESEEM traces detected on a water soluble standard
(gray) and on the two spin labeled positions. Bottom, ESEEM modulation depths correlated with
the water accessibility. Adapted from [47]
less accessible (e.g., V196 located towards the TMDs interior) can be predicted
(Fig. 5b, PDB 2BHW). LHCIIb used in this study was spin labeled with IAP
(iodoacetamido proxyl spin label). It is worth pointing out that the preparation of
the antenna complex requires the mixing of the purified apo-protein solubilized in
detergent with a solution of pigments and lipids in slightly reducing conditions
140 E. Bordignon
(necessary for the pigments stability). IAP was chosen as best labeling probe based
on the following criteria: (1) first attempts to use iodoacetamido TEMPO
(6-membered ring) revealed that the piperidine- was more easily reduced than the
analogous pyrrolidine-nitroxide; (2) the disulfide bond between the cys residue and
MTSSL was unstable, resulting in release of label from the sites. The reversibility
of the disulfide bond formation between the cys residue and MTSSL makes it
possible to remove the label (e.g., with 10 mM DTT, dithiothreithol) to recover
the single cys mutant. However, the weakness of the disulfide bonds also makes it
possible to release the label (1) upon relevant protein conformational changes
involving the spin label site, (2) in the presence of a neighboring unlabeled free
cysteine, and (3) upon temperature increase. Although the MTSSL is the most
widely used, the choice of the best spin label probe is necessarily protein- and
conditions-dependent.
Experimentally, the three-pulse ESEEM experiment is performed at 50 K with a
p/2-t-p/2-T-p/2-t-echo pulse sequence with standard phase cycling. The pulse
length of the p/2 pulse is usually 16 ns and the constant interpulse delay t 344 ns,
corresponding to a proton blind spot. Suppression of the proton modulations at the
blind spot improves precision of the fitting of the deuterium modulation. The
second interpulse delay, T, with an initial value of 80 ns, is incremented in steps
of 8 ns. The three-pulse ESEEM time domain traces for two selected positions are
presented in Fig. 5b (right panel). The differences in water accessibility are clearly
seen in the normalized primary data by comparing the depth of the ESEEM
modulations, defined as the peak-to-peak amplitude between the first maximum
and the first minimum of the deuterium modulation (higher for 52 than for 196).
Different methods to extract quantitatively the deuterium modulation depth taking
into account its dependence on the interpulse delay are discussed in the original
work [47]. The ESEEM water accessibility method is shown to be robust and to
complement the more traditional accessibility analysis.
The interaction between two spin labels attached to either a single polypeptide or
adjacent polypeptides (e.g., protein complexes, oligomers, etc.) is composed of
static dipolar interaction, which can be modulated by the residual motion of the spin
label side chains, and exchange interaction [55]. By measuring the spinspin
interaction at cryogenic temperature in the so-called rigid limit case (<200 K),
the residual motion of the label is minimized and the exchange and dipolar
interactions can be precisely extracted. Interspin distances in the 15 nm range
are routinely measured by EPR on membrane proteins with an accuracy varying
with the mean distance and distance distribution to be measured and with the
chosen experimental conditions. The short distances (12 nm) are extracted from
Site-Directed Spin Labeling of Membrane Proteins 141
continuous wave EPR spectra in the rigid limit via lineshape analysis [56, 57]. The
long distances (1.76 nm) can be measured by pulse EPR techniques, namely
Double Electron Electron Resonance (DEER, also known as PELDOR, pulsed
electron-electron double resonance) [58, 59] or Double-Quantum EPR [60]. The
interspin distance determination is a powerful strategy to deduce proximity of
selected secondary structural elements in membrane proteins or protein complexes
of unknown structure, and to follow conformational changes correlated with the
physiological activity of membrane proteins.
Since the linewidth of the spectra is a steep function of the interspin distance,
empirical or semiempirical parameters such as spectral amplitude ratios or spectral
second moment values were used to extract distances semiquantitatively and to
answer structural questions in the past. In the following, the software available to
extract distances and the caveats hidden in the analysis are presented.
The dipolar broadening produced by interspin distances in the 1.52 nm range is
extracted from low temperature CW spectra by convolution or deconvolution
methods given that the spectra of the two noninteracting spin labels are available.
In the case of homodimeric proteins, the noninteracting spectrum can be approxi-
mated by spin dilution methods (e.g., the protein is spin labeled with 20% MTSSL
and 80% nondiamagnetic analog, leading to approximately 4% of double labeled
homodimers in the final sample). In the case of higher order homo-oligomers,
molecular engineering can help to solve the problems of overlapping distances in
CW lineshape analysis (discussed in the next section).
The most popular pieces of software available to extract short distances from
CW EPR spectra are DIPFIT (from Steinhoffs group) [57], ShortDistances
(from Altenbach, https://sites.google.com/site/altenbach/labview-programs/short-
distances) or user-based fitting routines with Easyspin [48].
In DIPFIT one first needs to simulate the rigid limit spectrum of the noninter-
acting species (to extract the linewidth parameters pseudo Voigtian lineshape
which will be kept constant in the fitting of the dipolar spectrum). The
software performs a convolution of the simulated noninteracting spectrum with a
Pake function (in the strong coupling regime) characterized by a distance r and
Gaussian distance distribution with a standard deviation (sr) to fit the dipolar-
broadened spectrum. The distance distribution obtained can be considered reliable
if the mean distance is in the 1.42.0 nm range. For r > 2 nm, the accuracy of
the convolution methods progressively decreases, as the dipolar splitting gets
closer to the natural EPR linewidth, and for r < 1.4 nm the presence of exchange
coupling influences the accuracy of the distances extracted solely based on the
dipolar terms [61].
A problem arises if the spin labeling efficiency is not 100%, a case generally
encountered. DIPFIT allows the fraction of the noninteracting spectrum to be fitted.
However, the three main parameters to be varied, namely the distance, its distribu-
tion, and the fraction of noninteracting species, are strictly correlated. The knowl-
edge of the spin labeling efficiency is mandatory for an accurate determination for
the distance (and it should be >60% for reliable distance extraction).
142 E. Bordignon
The software ShortDistances is also solely based on dipolar terms, requires the
noninteracting spectrum as template, and performs the convolution based on a Pake
function (strong coupling regime) which allows for different distance distribution
functions to be used. Due to the large amount of parameters to be fitted, a simple
Gaussian distribution is to be preferred, because more complicated functions not
supported by modeling or structural information may lead to results difficult to
interpret. The possibility to extract distances from room temperature spectra, which
is in principle allowed in ShortDistances cannot be generalized to arbitrary pairs of
interacting spins, and each case must be evaluated thoroughly before convolution
with a Pake function to simulate interspin distances which are modulated by
the dynamics of the side chains [62]. Decreasing the temperature to 160 K is the
suggested general procedure to solve at least the problem of the residual spin label
mobility.
In general, convolution of the noninteracting spectrum provides only an approx-
imation to the lineshape of the dipolar-broadened spectrum even in the rigid
limit. First, this corresponds to treating the dipoledipole interaction as a small
perturbation, neglecting the pseudosecular term. This approximation is valid only if
the dipoledipole interaction is much smaller than the typical frequency difference
between the two coupled spins. Second, orientation correlation between the
spinspin vector and the molecular frame of the nitroxide is neglected. Both these
approximations can be relaxed by simulating the dipoledipole coupled spectrum
with Easyspin [48]. However, such an approach is computationally much more
expensive than Pake convolution. Furthermore, the CW EPR lineshape at a single
frequency may not be sufficient to extract both the interspin distance and
the relative orientation of the labels. In any case these complications introduce
errors in distances extracted by Pake convolution methods, which have not yet been
systematically studied by theoretical approaches.
Many of the problems encountered by lineshape analysis are not present when
using DEER [58] for the longer distance range determination. In fact there is no
need to have the spectrum of each noninteracting spin system to obtain a reliable
interspin distance determination (one needs only one or two singly spin labeled
proteins to confirm the background decay function to be used throughout the study
on the same protein); the amount of noninteracting species in the sample does not
contribute to the form factor F(t), and thus not to the accuracy of the distance
extracted (only the modulation depth is affected); a large range of distances can be
determined with high precision including the 45 nm width of the membrane
bilayer barrier. The first DEER application to membrane proteins appeared in
2004 on the membrane protein Na+/proline transporter PtuP showing that it was
possible to measure the distance between two labels positioned across the mem-
brane bilayer (about 5 nm) [63]. Jeschke suggested DEER as a powerful method to
investigate structure and conformational changes of integral membrane proteins
reconstituted in liposomes. In fact, since then, a tremendous number of publications
using DEER as main driving technique for structural investigation on membrane
Site-Directed Spin Labeling of Membrane Proteins 143
proteins followed. The most common software used to extract distance distribution
from DEER traces is DeerAnalysis [64] (http://www.epr.ethz.ch/software/index).
Other software is available from Freeds (http://www.acert.cornell.edu/index_files/
acert_ftp_links. php) and Fajers (http://www.sb.fsu.edu/~fajer/Programs/DEFit/
defit.html) labs.
One of the major problems facing distance determination by pulsed EPR on spin
labeled membrane proteins is the short relaxation time Tm (generally around 1 ms).
Solvent deuteration is routinely used (1020 vol% deuterated glycerol as cryopro-
tectant for membrane-embedded or detergent-solubilized samples, or deuterated
buffer) to slow down the nitroxide relaxation, thus extending the range of distance
measurement and sensitivity. However, the problems connected to the too fast
relaxation time called for a systematic analysis of all factors modifying the Tm
[65]. Recently, deuteration of water-soluble histone proteins was shown to extend
Tm to a considerable degree (up to 30 ms) [66].
Other sensitivity problems arise from the low concentrations usually available
for membrane proteins, and the molecular crowding in the membrane bilayer
leading to very fast background decay functions which impair the DEER analysis,
especially for long distances. This problem can be minimized by increasing the
lipid to protein ratio (at least 5001,000 lipid per protein). All these factors call for
the development of even more sensitive techniques, e.g., Q-band DEER [67, 68] at
high microwave power, and new strategies, e.g., spin labels containing metal ions
which can also be used in combination with nitroxide labels [6971]. The DEER
experiment is rather expensive in terms of measurement time, typically one trace
with good signal to noise on membrane proteins requires 1224 h, and thus even a
small sensitivity increase (e.g., a factor 2 in S/N) can speed up a protein study four
times.
The presence of multiple spins in the same protein or protein complex creates
interpretative problems in DEER, due to the appearance of artifact peaks in the
distance distribution. Methods to identify artifact distances in multispin systems
are under development [72]. Strategies to overcome problems due to the super-
position of distances in the same polypeptide can make use of orthogonal
labeling (e.g., using different spin labels binding to cys and unnatural amino
acids [73]). In the easier case of protein complexes with spin labels in different
subunits, one can use different nitroxide isotopes, as demonstrated on model
compounds [74].
Having an accurate distance determination between the NO groups of the spin
labeled side chains still does not directly convey the structural information at the
level of the backbone of the protein which would be required for modeling
structures and complexes at high resolution. To correlate the spinspin distance
constraints to the backbonebackbone distances requires modeling. At the present
time, modeling approaches combining sparse accurate distance constraints to
macromolecular structures are under development [7, 75].
144 E. Bordignon
Fig. 6 Structural details obtained by lineshape analysis. (a) In KcsA, labeling at any residue
position renders a tetramer with potentially four spin labels. (i) Tandem dimer construct (ii) with
cys residues in both protomers (control used to evaluate the effects of the intersubunit linker) and
(iii) with only one of the protomers containing a cys (used in the analysis). (b) Rigid-limit X-band
EPR spectra obtained at pH 7 (thick line, closed state) and at pH 4 (thin line, open state). Right
panel, absorption spectra obtained from integration and relative fits obtained with convolution
superimposed. (c) Simulated spin- and amplitude-normalized spectra for the two interspin
distances in the figure (100% spin labeling efficiency). (d) Helical wheel representation of residues
100119. Both closed (top) and open (bottom) states are represented as pairs of helical wheel
146 E. Bordignon
(Fig. 6d), and gives the first insights into a possible rotation and translation
movement of TM2 away from the symmetry axis. Perozo developed a simple
computational approach (restraint-driven Cartesian transformation, ReDCaT)
based on the exhaustive sampling of rigid-body movement in Cartesian space to
determine the type, direction, and magnitude of the conformational changes in TM2
using limited distance information. The EPR-driven model of the conformational
change of the TM2 helices from the closed to the open channel is presented in
Fig. 6e, and supports a scissoring-type motion.
Fig. 6 (continued) diagrams. The arrows inside the wheels in the open state represent distance
changes from close to open (outward is increase and inward, decrease). (e) Modeling the confor-
mational rearrangements in TM2. Up, graphical representation showing the type and extent of
individual helical movements (red closed and blue open), bottom, calibrated cross-sectional
representation of the conformational changes. Adapted from [76]
Site-Directed Spin Labeling of Membrane Proteins 147
Fig. 7 Modeling a high-resolution dimeric structure. (a) Estimate of the mean distance on the
example of spin label K-221R1. The primary DEER trace (V(t)/V(0)), the form factor and the
distance distribution obtained by Tikhonov regularization (the L curve is shown in the inset) with
the software DeerAnalysis are presented. (b) Fits of primary experimental DEER data (black lines)
by simulated data (red lines) corresponding to the final structure of the NhaA dimer and a
distribution of spin label conformations modeled by a rotamer library. (c) C2 symmetry axis of
the dimer created with the EPR constraints. (d) Comparison between the EPR structure and the
electron density projection to the membrane plane obtained by cryo-EM on 2D crystals. (e)
Comparison between the EPR structure and the dimer modeled on the refined cryo-EM data
(PDB 3FI1). Adapted from [77]
the primary data V(t) are fitted directly (Fig. 7b), and the information on the
distance distribution are used in the docking approach to match a modeled confor-
mational distribution of the spin labels.
In the first stage of the modeling the two component molecules are treated as
rigid bodies. The relative arrangement of a dimer requires in principle six
148 E. Bordignon
parameters (three Euler angles for the rotation and three components of a transla-
tion vector). In this case, the C2 symmetry of the resulting homodimer reduces the
necessary parameters to four (Fig. 7c). The nine distance constraints experimentally
obtained can then be considered enough to produce a reliable docking of the two
proteins. Details of the grid search algorithm and the final energy minimization
which yield the dimeric arrangement of NhaA are thoroughly described in the
original work [77]. The dimer obtained by DEER distance constraints in
proteoliposomes is presented in Fig. 7d, superimposed to the low resolution cryo-
EM density showing the rather good agreement, in particular in the interfaces. More
recently, a higher resolution dimeric model based on the monomeric structure was
produced (PDB 3FI1) using a better resolved 2D cryo-EM crystal. The new model
is shown to be in good agreement with the EPR-model (Fig. 7e), confirming the
high potentiality of EPR-driven modeling for membrane protein complexes in
quasi-native environments.
Fig. 8 EPR detection of membrane insertion of spin labeled proteins. (a) NMR models of inactive
Bax and Bid (PDB 1F16, 2BID, respectively). The buried helical hairpin is shown in orange, and
the BH3 domain in yellow. Positions spin labeled in Bax are highlighted. (b) Kinetics of Bax
conformational changes at 37 C. Spectra of double labeled Bax were recorded each 43 s at 37 C
150 E. Bordignon
GPCRs convey 80% of signal transduction across cell membranes [83]. The
receptors consist of seven transmembrane helices (analogous to bacteriorhodopsin)
connected by intra- and extracellular loops and they can be activated by diverse
ligands (light-sensitive compounds, amino acids, peptides, neurotransmitters, etc.).
GPCR in turn can activate an associated G-protein which further affects
Fig. 8 (continued) with different triggering agents. The intensity of the central EPR line vs
incubation time (logarithmic scale) is plotted. The inset shows the spectra at time zero (black
dotted) and at the end of the incubation time. (c) DEER traces and distance distribution obtained
with DeerAnalysis in different conditions, as written in the figure. The asterisk highlights the
glimpse of long distances reflecting the interdimer interaction. (d) Model of three adjacent Bax
dimers, based on the interspin distances detected. Adapted from [80]
Site-Directed Spin Labeling of Membrane Proteins 151
6 Conclusions
SDSL EPR is sensitive to flexible regions of proteins and to dynamical changes, and
can be used to measure water accessibility profiles and accurate distances between
spin labeled side chains. The wealth of information that can thus be obtained makes
SDSL EPR a direct tool to access conformational changes of proteins. The bridge
152 E. Bordignon
Fig. 9 Light-induced opening of helix TM6 in rhodopsin. (a) View of the cytoplasmic face of
inactive rhodopsin (PDB entry 1GZM) showing modeled R1 side chains for all sites investigated.
(b) Representative selection of DEER results in the inactive (black) and active (red) states. Left,
normalized dipolar evolution after removal of the exponential background. Middle, DC centered
Fourier transform of the data. Right, distribution calculated by Tikhonov regularization. (c)
Projection contours of the spin locations calculated from the measured distance distributions for
representative sites (blue, inactive; red, active). The outward movement of TM6 is highlighted by
an arrow. Right, after defining a central reference point halfway between the dark location of 74
and 252, the radial distribution profiles along a line connecting the center and the most probable
location is calculated for each site. Adapted from [86]
Site-Directed Spin Labeling of Membrane Proteins 153
References
1. Deisenhofer J, Epp O, Miki K, Huber R, Michel H (1985) Structure of the protein subunits in
the photosynthetic reaction centre of Rhodopseudomonas viridis at 3A resolution. Nature 318
(6047):618624
2. Altenbach C, Flitsch SL, Khorana HG, Hubbell WL (1989) Structural studies on transmem-
brane proteins. 2. Spin labeling of bacteriorhodopsin mutants at unique cysteines. Biochemis-
try 28(19):78067812
3. Altenbach C, Marti T, Khorana HG, Hubbell WL (1990) Transmembrane protein-structure spin
labeling of bacteriorhodopsin mutants. Science 248(4959):10881092
4. Henderson R, Baldwin JM, Ceska TA, Zemlin F, Beckmann E, Downing KH (1990) Model for
the structure of bacteriorhodopsin based on high-resolution electron cryo-microscopy. J Mol
Biol 213(4):899929
5. PebayPeyroula E, Rummel G, Rosenbusch JP, Landau EM (1997) X-ray structure of bacterio-
rhodopsin at 2.5 angstroms from microcrystals grown in lipidic cubic phases. Science 277
(5332):16761681
6. Hubbell WL, Altenbach C (1994) Investigation of structure and dynamics in membrane-
proteins using site-directed spin-labeling. Curr Opin Struct Biol 4(4):566573
7. Polyhach Y, Bordignon E, Jeschke G (2011) Rotamer libraries of spin labelled cysteines for
protein studies. Phys Chem Chem Phys 13(6):23562366
8. Denisov IG, Grinkova YV, Lazarides AA, Sligar SG (2004) Directed self-assembly of
monodisperse phospholipid bilayer nanodiscs with controlled size. J Am Chem Soc 126
(11):34773487
9. Skar-Gislinge N, Simonsen JB, Mortensen K, Feidenhansl R, Sligar SG, Lindberg Mller B,
Bjrnholm T, Arleth L (2010) Elliptical structure of phospholipid bilayer nanodiscs
encapsulated by scaffold proteins: casting the roles of the lipids and the protein. J Am Chem
Soc 132(39):1371313722
10. Grinkova YV, Denisov IG, Sligar SG (2010) Engineering extended membrane scaffold
proteins for self-assembly of soluble nanoscale lipid bilayers. Protein Eng Des Select 23
(11):843848
11. Steinhoff HJ, Hubbell WL (1996) Calculation of electron paramagnetic resonance spectra
from Brownian dynamics trajectories: application to nitroxide side chains in proteins. Biophys
J 71(4):22012212
12. Beier C, Steinhoff H-J (2006) A structure-based simulation approach for electron paramag-
netic resonance spectra using molecular and stochastic dynamics simulations. Biophys J 91
(7):26472664
13. Oganesyan VS (2011) A general approach for prediction of motional EPR spectra from
molecular dynamics (MD) simulations: application to spin labelled protein. Phys Chem
Chem Phys 13:47244737
14. Sezer D, Freed JH, Roux B (2009) Multifrequency electron spin resonance spectra of a spin-
labeled protein calculated from molecular dynamics simulations. J Am Chem Soc 131
(7):25972605
15. Oganesyan VS (2007) A novel approach to the simulation of nitroxide spin label EPR spectra
from a single truncated dynamical trajectory. J Magn Reson 188(2):196205
154 E. Bordignon
16. Liang ZC, Freed JH (1999) An assessment of the applicability of multifrequency ESR to study
the complex dynamics of biomolecules. J Phys Chem B 103(30):63846396
17. Zhang Z, Fleissner MR, Tipikin DS, Liang Z, Moscicki JK, Earle KA, Hubbell WL, Freed JH
(2010) Multifrequency electron spin resonance study of the dynamics of spin labeled T4
lysozyme. J Phys Chem B 114(16):55035521
18. Isas JM, Langen R, Haigler HT, Hubbell WL (2002) Structure and dynamics of a helical
hairpin and loop region in annexin 12: a site-directed spin labeling study. Biochemistry 41
(5):14641473
19. Mchaourab HS, Lietzow MA, Hideg K, Hubbell WL (1996) Motion of spin-labeled side chains
in T4 lysozyme, correlation with protein structure and dynamics. Biochemistry 35
(24):76927704
20. Bordignon E, Steinhoff HJ (2007) Membrane protein structure and dynamics studied by site-
directed spin-labeling ESR. In: Hemminga MA, Berliner LJ (eds) ESR spectroscopy
in membrane biophysics, vol 27. Biological magnetic resonance. Springer, New York,
pp 129164
21. Freed DM, Horanyi PS, Wiener MC, Cafiso DS (2010) Conformational exchange in a
membrane transport protein is altered in protein crystals. Biophys J 99(5):16041610
22. Perozo E, Cortes DM, Cuello LG (1998) Three-dimensional architecture and gating mecha-
nism of a K+ channel studied by EPR spectroscopy. Nat Struct Biol 5(6):459469
23. Hille B (2001) Ion channels of excitable membranes, 3rd edn. Sinauer Associates, Sunderland
24. Doyle DA, Morais Cabral J, Pfuetzner RA, Kuo A, Gulbis JM, Cohen SL, Chait BT,
MacKinnon R (1998) The structure of the potassium channel: molecular basis of K+ conduc-
tion and selectivity. Science 280(5360):6977
25. Perozo E, Cortes DM, Cuello LG (1998) Molecular architecture of the Streptomyces K+
channel first transmembrane segment. A site directed spin-labeling study. Biophys J 74(2):A44
26. Zou P, Mchaourab HS (2009) Alternating access of the putative substrate-binding chamber
in the ABC transporter MsbA. J Mol Biol 393(3):574585
27. Doebber M, Bordignon E, Klare JP, Holterhues J, Martell S, Mennes N, Li L, Engelhard M,
Steinhoff HJ (2008) Salt-driven equilibrium between two conformations in the HAMP domain
from N. pharaonis: the language of signal transduction? J Biol Chem 283(42):2869128701
28. Bridges MD, Hideg K, Hubbell WL (2010) Resolving conformational and rotameric exchange
in spin-labeled proteins using saturation recovery EPR. Appl Magn Reson 37(14):363
29. McCoy J, Hubbell WL (2011) High-pressure EPR reveals conformational equilibria and
volumetric properties of spin-labeled proteins. Proc Natl Acad Sci USA 108(4):13311336
30. Steinhoff HJ, Mollaaghababa R, Altenbach C, Hideg K, Krebs M, Khorana HG, Hubbell WL
(1994) Time-resolved detection of structural-changes during the photocycle of spin-labeled
bacteriorhodopsin. Science 266(5182):105107
31. Radzwill N, Gerwert K, Steinhoff HJ (2001) Time-resolved detection of transient movement
of helices F and G in doubly spin-labeled bacteriorhodopsin. Biophys J 80(6):28562866
32. Wegener AA, Chizhov I, Engelhard M, Steinhoff HJ (2000) Time-resolved detection of
transient movement of helix F in spin-labelled pharaonis sensory rhodopsin II. J Mol Biol
301(4):881891
33. Bordignon E, Klare JP, Holterhues J, Martell S, Krasnaberski A, Engelhard M, Steinhoff HJ
(2007) Analysis of light-induced conformational changes of Natronomonas pharaonis sensory
rhodopsin II by time resolved electron paramagnetic resonance spectroscopy. Photochem
Photobiol 83(2):263272
34. Klare JP, Bordignon E, Engelhard M, Steinhoff HJ (2004) Sensory rhodopsin II and bacterio-
rhodopsin: light activated helix F movement. Photochem Photobiol Sci 3(6):543547
35. Wegener AA, Klare JP, Engelhard M, Steinhoff HJ (2001) Structural insights into the early
steps of receptor-transducer signal transfer in archaeal phototaxis. EMBO J 20(19):53125319
36. Gordeliy VI, Labahn J, Moukhametzianov R, Efremov R, Granzin J, Schlesinger R, B uldt G,
Savopol T, Scheidig AJ, Klare JP, Engelhard M (2002) Molecular basis of transmembrane
signalling by sensory rhodopsin II-transducer complex. Nature 419(6906):484487
Site-Directed Spin Labeling of Membrane Proteins 155
37. Altenbach C, Greenhalgh DA, Khorana HG, Hubbell WL (1994) A collision gradient-method
to determine the immersion depth of nitroxides in lipid bilayers application to spin-labeled
mutants of bacteriorhodopsin. Proc Natl Acad Sci USA 91(5):16671671
38. Altenbach C, Froncisz W, Hemker R, Mchaourab H, Hubbell WL (2005) Accessibility of
nitroxide side chains: absolute Heisenberg exchange rates from power saturation EPR.
Biophys J 89(3):21032112
39. Pyka J, Ilnicki J, Altenbach C, Hubbell WL, Froncisz W (2005) Accessibility and dynamics of
nitroxide side chains in T4 lysozyme measured by saturation recovery EPR. Biophys J 89
(3):20592068
40. Volkov A, Dockter C, Polyhach Y, Paulsen H, Jeschke G (2010) Site-specific information on
membrane protein folding by electron spin echo envelope modulation spectroscopy. J Phys
Chem Lett 1(3):663667
41. Pavlova A, McCarney ER, Peterson DW, Dahlquist FW, Lew J, Han S (2009) Site-specific
dynamic nuclear polarization of hydration water as a generally applicable approach to monitor
protein aggregation. Phys Chem Chem Phys 11(31):68336839
42. Reyes CL, Ward A, Yu J, Chang G (2006) The structures of MsbA: insight into ABC
transporter-mediated multidrug efflux. FEBS Lett 580(4):10421048
43. Locher KP (2009) Structure and mechanism of ATP-binding cassette transporters. Philos
Trans R Soc Lond B Biol Sci 364(1514):239245
44. Dong J, Yang G, McHaourab HS (2005) Structural basis of energy transduction in the transport
cycle of MsbA. Science 308(5724):10231028
45. Mobius K, Savitsky A, Wegener C, Plato M, Fuchs M, Schnegg A, Dubinskii AA, Grishin YA,
Grigorev IA, Kuhn M, Duche D, Zimmermann H, Steinhoff HJ (2005) Combining high-field
EPR with site-directed spin labeling reveals unique information on proteins in action. Magn
Reson Chem 43:S4S19
46. Wegener C, Savitsky A, Pfeiffer M, M obius K, Steinhoff HJ (2001) High-field EPR-detected
shifts of magnetic tensor components of spin label side chains reveal protein conformational
changes: the proton entrance channel of bacteriorhodopsin. Appl Magn Reson 21
(34):441452
47. Volkov A, Dockter C, Bund T, Paulsen H, Jeschke G (2009) Pulsed EPR determination of
water accessibility to spin-labeled amino acid residues in LHCIIb. Biophys J 96(3):11241141
48. Stoll S, Schweiger A (2006) EasySpin, a comprehensive software package for spectral
simulation and analysis in EPR. J Magn Reson 178(1):4255
49. Bordignon E, Brutlach H, Urban L, Hideg K, Savitsky A, Schnegg A, Gast P, Engelhard M,
Groenen EJJ, Mobius K, Steinhoff HJ (2010) Heterogeneity in the nitroxide micro-environment:
polarity and proticity effects in spin-labeled proteins studied by multi-frequency EPR. Appl
Magn Reson 37(14):391403
50. Plato M, Steinhoff HJ, Wegener C, T orring JT, Savitsky A, M
obius K (2002) Molecular orbital
study of polarity and hydrogen bonding effects on the g and hyperfine tensors of site directed
NO spin labelled bacteriorhodopsin. Mol Phys 100(23):37113721
51. Brutlach H, Bordignon E, Urban L, Klare JP, Reyer H-J, Engelhard M, Steinhoff HJ (2006)
High-field EPR and site directed spin labeling reveal a periodical polarity profile: the sequence
8894 of the phototransducer NpHtrII in complex with sensory rhodopsin, NpSRII. Appl
Magn Reson 30:359372
52. Jeschke G, Schweiger A (eds) (2001) Principles of pulse electron paramagnetic resonance.
Oxford University Press, Oxford
53. Standfuss J, Terwisscha van Scheltinga AC, Lamborghini M, Kuhlbrandt W (2005)
Mechanisms of photoprotection and nonphotochemical quenching in pea light-harvesting
complex at 2.5 A resolution. EMBO J 24(5):919928
54. Liu Z, Yan H, Wang K, Kuang T, Zhang J, Gui L, An X, Chang W (2004) Crystal structure of
spinach major light-harvesting complex at 2.72 A resolution. Nature 428(6980):287292
55. Berliner LJ, Eaton SS, Eaton GR (eds) (2000) Distance measurements in biological systems by
EPR. Biological magnetic resonance. Kluwer Academic/Plenum, New York
156 E. Bordignon
56. Rabenstein MD, Shin YK (1995) Determination of the distance between 2 spin labels attached
to a macromolecule. Proc Natl Acad Sci USA 92(18):82398243
57. Steinhoff HJ, Radzwill N, Thevis W, Lenz V, Brandenburg D, Antson A, Dodson G, Wollmer
A (1997) Determination of interspin distances between spin labels attached to insulin:
comparison of electron paramagnetic resonance data with the X-ray structure. Biophys J 73
(6):32873298
58. Pannier M, Veit S, Godt A, Jeschke G, Spiess HW (2000) Dead-time free measurement of
dipole-dipole interactions between electron spins. J Magn Reson 142(2):331340
59. Milov AD, Ponomarev AB, Tsvetkov YD (1984) Electron-electron double resonance in
electron spin echo: model biradical systems and the sensitized photolysis of decalin. Chem
Phys Lett 110(1):6772
60. Borbat PP, McHaourab HS, Freed JH (2002) Protein structure determination using long-
distance constraints from double-quantum coherence ESR: study of T4 lysozyme. J Am
Chem Soc 124(19):53045314
61. Banham JE, Baker CM, Ceola S, Day IJ, Grant GH, Groenen EJJ, Rodgers CT, Jeschke G,
Timmel CR (2008) Distance measurements in the borderline region of applicability of CW
EPR and DEER: a model study on a homologous series of spin-labelled peptides. J Magn
Reson 191(2):202218
62. Altenbach C, Oh KJ, Trabanino RJ, Hideg K, Hubbell WL (2001) Estimation of inter-residue
distances in spin labeled proteins at physiological temperatures: experimental strategies and
practical limitations. Biochemistry 40(51):1547115482
63. Jeschke G, Wegener C, Nietschke M, Jung H, Steinhoff HJ (2004) Interresidual distance
determination by four-pulse double electron-electron resonance in an integral membrane
protein: the Na+/proline transporter PutP of Escherichia coli. Biophys J 86(4):25512557
64. Jeschke G, Chechik V, Ionita P, Godt A, Zimmermann H, Banham J, Timmel CR, Hilger D,
Jung H (2006) DeerAnalysis2006 a comprehensive software package for analyzing pulsed
ELDOR data. Appl Magn Reson 30(34):473498
65. Dastvan R, Bode BE, Karuppiah MP, Marko A, Lyubenova S, Schwalbe H, Prisner TF (2010)
Optimization of transversal relaxation of nitroxides for pulsed electron-electron double
resonance spectroscopy in phospholipid membranes. J Phys Chem B 114(42):1350713516
66. Ward R, Bowman A, Sozudogru E, El-Mkami H, Owen-Hughes T, Norman DG (2010) EPR
distance measurements in deuterated proteins. J Magn Reson 207(1):164167
67. Zou P, Mchaourab HS (2010) Increased sensitivity and extended range of distance
measurements in spin-labeled membrane proteins: Q-band double electron-electron resonance
and nanoscale bilayers. Biophys J 98(6):L18L20
68. Hofer P, Heilig R, Schmalbein D (2000) The superQ-FT accessory for pulsed EPR, ENDOR
and ELDOR at 34 GHz. Bruker Spin Rep 152(153):3743
69. Lueders P, Jeschke G, Yulikov M (2011) Double electronelectron resonance measured
between Gd3+ ions and nitroxide radicals. J Phys Chem Lett 2(6):604609
70. Potapov A, Song Y, Meade TJ, Goldfarb D, Astashkin AV, Raitsimring A (2010) Distance
measurements in model bis-Gd(III) complexes with flexible bridge. Emulation of biological
molecules having flexible structure with Gd(III) labels attached. J Magn Reson 205(1):3849
71. Raitsimring AM, Gunanathan C, Potapov A, Efremenko I, Martin JM, Milstein D, Goldfarb D
(2007) Gd3+ complexes as potential spin labels for high field pulsed EPR distance
measurements. J Am Chem Soc 129(46):1413814139
72. Jeschke G, Sajid M, Schulte M, Godt A (2009) Three-spin correlations in double electron-
electron resonance. Phys Chem Chem Phys 11(31):65806591
73. Fleissner MR, Brustad EM, Kalai T, Altenbach C, Cascio D, Peters FB, Hideg K, Peuker S,
Schultz PG, Hubbell WL (2009) Site-directed spin labeling of a genetically encoded unnatural
amino acid. Proc Natl Acad Sci USA 106(51):2163721642
74. Jeschke G, Zimmermann H, Godt A (2006) Isotope selection in distance measurements
between nitroxides. J Magn Reson 180(1):137146
Site-Directed Spin Labeling of Membrane Proteins 157
75. Hirst SJ, Alexander N, McHaourab HS, Meiler J (2011) RosettaEPR: an integrated tool for
protein structure determination from sparse EPR data. J Struct Biol 173(3):506514
76. Liu YS, Sompornpisut P, Perozo E (2001) Structure of the KcsA channel intracellular gate in
the open state. Nat Struct Biol 8(10):883887
77. Hilger D, Polyhach Y, Padan E, Jung H, Jeschke G (2007) High-resolution structure of a
Na+/H+ antiporter dimer obtained by pulsed election paramagnetic resonance distance
measurements. Biophys J 93(10):36753683
78. Bhatnagar J, Borbat PP, Pollard AM, Bilwes AM, Freed JH, Crane BR (2010) Structure of the
ternary complex formed by a chemotaxis receptor signaling domain, the CheA histidine
kinase, and the coupling protein CheW as determined by pulsed dipolar ESR spectroscopy.
Biochemistry 49(18):38243841
79. Bhatnagar J, Freed JH, Crane BR (2007) Rigid body refinement of protein complexes with
long-range distance restraints from pulsed dipolar ESR. In: Methods in enzymology, vol 423.
Academic, London, pp 117133
80. Bleicken S, Classen M, Padmavathi PVL, Ishikawa T, Zeth K, Steinhoff HJ, Bordignon E
(2010) Molecular details of BAX activation, oligomerization, and membrane insertion. J Biol
Chem 285(9):66366647
81. Gavathiotis E, Suzuki M, Davis ML, Pitter K, Bird GH, Katz SG, Tu HC, Kim H, Cheng EHY,
Tjandra N, Walensky LD (2008) BAX activation is initiated at a novel interaction site. Nature
455(7216):10761081
82. Bode BE, Margraf D, Plackmeyer J, Durner G, Prisner TF, Schiemann O (2007) Counting the
monomers in nanometer-sized oligomers by pulsed electron-electron double resonance. J Am
Chem Soc 129(21):67366745
83. Millar RP, Newton CL (2010) The year in G protein-coupled receptor research. Mol
Endocrinol 24(1):261274
84. Palczewski K, Kumasaka T, Hori T, Behnke CA, Motoshima H, Fox BA, Trong IL, Teller DC,
Okada T, Stenkamp RE, Yamamoto M, Miyano M (2000) Crystal structure of rhodopsin: a G
protein-coupled receptor. Science 289(5480):739745
85. Altenbach C, Yang K, Farrens DL, Farahbakhsh ZT, Khorana HG, Hubbell WL (1996)
Structural features and light-dependent changes in the cytoplasmic interhelical E-F loop region
of rhodopsin: a site-directed spin-labeling study. Biochemistry 35(38):1247012478
86. Altenbach C, Kusnetzow AK, Ernst OP, Hofmann KP, Hubbell WL (2008) High-resolution
distance mapping in rhodopsin reveals the pattern of helix movement due to activation. Proc
Natl Acad Sci USA 105(21):74397444
Top Curr Chem (2012) 321: 159198
DOI: 10.1007/128_2011_300
# Springer-Verlag Berlin Heidelberg 2011
Published online: 13 December 2011
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
2 Spin Labeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
2.1 Nucleobases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
2.2 Sugar Phosphate Backbone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
3 Theoretical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
3.1 CW EPR Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
3.2 Hyperfine Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
3.3 Pulsed Electron-Electron Double Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
4 Applications to Nucleic Acids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.1 CW EPR Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.2 Pulse-EPR applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
5 Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
1 Introduction
Nucleic acid (NA) molecules play essential roles in such biological functions as
translation and transcription, most importantly by storing the genetic code (DNA)
and transferring it for protein synthesis (RNA). Additionally, they are involved in
catalytic and regulatory functions in cells, as has been recently discovered in
ribozymes, riboswitches, small regulatory RNAs, and the ribosome. In all these
cases, the interactions of NAs with small molecules or proteins are of central
importance for their function. Different from most proteins, NAs can adopt several
tertiary structures, depending on their surrounding and external conditions. This
large conformational flexibility is on the one hand an important and crucial factor
for a detailed understanding of their specific function and on the other a limiting
factor for classical structural methods, such as for example X-ray crystallography.
Methods capable of investigating structural and dynamic aspects of
macromolecules are therefore especially significant for the functional characteriza-
tion of NAs. NMR is one of the most promising methods for working with NAs: it
can determine the tertiary structure and the fast local dynamics of NAs with atomic
resolution, but is at its limit for larger NA molecules or NAprotein complexes
and slow conformational dynamics [1, 2]. Fluorescence Resonance Energy Transfer
(FRET) can detect such long range distance changes by attaching two fluorophore-
labels to the NA molecule [3, 4]. Its high sensitivity allows the performance of such
measurements at room temperature down to a single molecule level; however, due
to the linker flexibility and other unknown parameters, the derived distance infor-
mation is mostly qualitative in nature. EPR spectroscopy is a method which can fill
this gap between the local high resolution structures, derived by NMR spectros-
copy, and the dynamical information obtained by FRET spectroscopy. Within the
last few years the bandwidth of EPR methods used to investigate dynamics and
structural constraints on NAs has strongly increased and has established EPR as a
new and promising method in the field of nucleic acid research. In this review we
want to summarize some of the more recent results obtained by EPR spectroscopy
on NAs. We first review the most common nitroxide spin labeling strategies used
for NAs, briefly introduce different EPR experiments used for the investigation of
structure and conformational flexibility of NAs, and discuss selected applications
which illustrate and highlight the potential of the EPR method on NAs. Finally, we
conclude by pointing out some new trends and potential perspectives for the future.
2 Spin Labeling
Site-specific spin labeling of nucleic acids differs substantially from the procedure
used for proteins; therefore the most common methods to attach nitroxide radicals
to nucleic acids will be described and compared in this chapter. Several reviews
Structure and Dynamics of Nucleic Acids 161
[57] and monographs [8, 9] can provide the reader with more in-depth
information.
First side-directed spin labeling (SDSL) studies date back to the early 1970s.
There, tRNAs were subjected to nitroxide labeling via either naturally occurring
rare nucleobases, e.g., 2-thio-uridine, enzymatically introduced nucleobases such as
2-thiocytidine and 4-thio-uridine, or the 30 -ends [1014]. It is worthwhile to note
that labeling of the 30 -end, 1 (Fig. 1), causes strong structural perturbations due to
the transformation of the sugar moiety to a morpholine derivative. Moreover, the
general drawback of the above-mentioned investigations is their requirement of
tRNA and, depending on the label, an introduction of a positive charge in the case
of 2-thiocytidine. The more recent approaches can be subdivided into strategies
2.1 Nucleobases
Approximately 25 years later, Hubbell and coworkers [15], as well as Varani and
Ramos [16], developed a labeling strategy that involves reaction of 4-thiouridine
substituted RNA with the nitroxides 2,2,5,5-tetramethylpyrroline-1-oxyl-3-methyl-
methanethiosulfonate (MTSL) or 2,2,5,5-tetramethyl-pyrroline-1-oxyl-3-iodo-
acetamide leading to 2 and 3 (Fig. 1), respectively. Both labeling strategies report
almost quantitative yields as monitored via UV-spectroscopy, but compound 3
(Fig. 1) is chemically more stable than 2 due to thioether formation instead of a
disulfide bridge generation. However, a crucial step is the choice of the labeling site
to minimize disturbances of the intermolecular interface, which might arise from
the loss of the imino proton in 4-thiouridine upon spin labeling. None of the authors
found severe affection of the RNA structure at the chosen positions resulting from
nitroxide modification. Yet the approach is limited to 4-thiouridine. Hopkins and
Robinson circumvent the abstraction of the imino proton, thus lowering structural
perturbations by usage of 5-iodouridine, 2,2,5,5-tetramethyl-pyrroline-1-oxyl-
3-acetylene (TPA) and Pd-catalyzed Sonogashira chemistry during solid-state
synthesis [17]. This protocol has been extended by Engels and coworkers to
5-iododesoxyuridine, 5-iodocytidine, and 2-iodoadenosine [1821]; 46 (Fig. 1)
depict the respective structures. This on column coupling shortens the time-
consuming NAs synthesis of already spin-labeled phosphoramidites previously
used in DNA dynamic studies [22] and involves simple purification. Another
advantage is the coupling yield of 95% if a 20 -acid-labile orthoester strategy in
combination with a new class of 50 -silyl ethers (Dharmacon) instead of a standard
tert-butyldimethylsilyl protection scheme is applied. The latter conditions led to
quenching of the nitroxide radical and further byproducts. Also, the corresponding
tetrahydropyridine analog of TPA, spin label 7 (Fig. 1), was recently introduced [23, 24].
It can be readily prepared in half the number of steps that are needed for its spectroscopic
ancestor 4. In addition, the length of the acetylene chain connecting the nitroxide to the
base has been investigated [25] proving small amplitudes of internal motion for 4 as
compared to a diacetylene tether. This is of great importance for PELDOR
measurements discussed in Sect. 4.2.2. Alternatively, ethynyl side chain bearing
derivatives of 7-deaza-2deoxyadenosine or 20 -deoxyuridine may undergo a Cu+
catalyzed [3 + 2]-cycloaddition with 2,2,6,6-tetramethyl-piperidine-1-oxyl-4-azide
(so-called click chemistry) to yield ~60% of 9 and 10 (Fig. 1) [26]. Note that usage
of Cu2+ instead of Cu+ leads to a reduction of the nitroxide to the hydroxylamine species.
Enzymatic strategies based on Klenow filling were also applied by Bobst
and coworkers to incorporate different types of spin labels attached to the 5- or
4-position of uridine, or the 5-position of cytidine of RNA and DNA [9, 27, 28].
None of the labels affected the structure of the biomolecule when 12% nucleotides
Structure and Dynamics of Nucleic Acids 163
were spin labeled. Nitroxide labeled deoxyguanine, 8 (Fig. 1), can be created post-
synthetically with 2,2,6,6-tetramethylpiperidine-1-oxyl-4-amine and a 2-fluoro-
hypoxanthine containing oligodeoxynucleotide [29]. Melting point analysis
indicates that nitroxide labeling of the guanine base does not destabilize the duplex.
A more general strategy employed by H obartner and coworkers allows direct attach-
ment of the paramagnetic probe to the exocyclic amino groups of nucleobases cytosine
and adenine in the major groove, and of guanine in the minor groove of WatsonCrick
base-paired RNA duplexes. The convertible nucleosides O4-(4-chlorophenyl)uridine,
O6-(4-chlorophenyl)inosine, and 2-fluoroinosine were labeled with 2,2,6,6-
tetramethylpiperidine-1-oxyl-4-amine leading to compounds 1113 (Fig. 1) [30].
Thermal destabilization was found to be in the order of 56 C per spin label for A,
C, and G. It is worth noting that cleavage from the polystyrene support, deprotection of
the phosphate backbone and nucleobases, as well as nucleobase substitution take place
at the same time in this procedure.
Nitroxides may also be attached to the site of choice via more than one covalent
bond, causing extremely rigid spin labels. As mentioned before, this improves the
information content of PELDOR experiments. One route discussed in the literature
is the synthesis of a quinolonyl based, heavily modified C-derived phosphoramidite
14 (Fig. 1). Because of the formation of a base pair with 2-aminopurine [31], its
applications are limited. Nevertheless, to overcome the size limitations of synthesized
oligonucleotides, 14 can be chemically incorporated into DNA and subjected to
Klenow polymerase filling [32]. Another extremely rigid C-analog, 15 (Fig. 1), has
been developed by Sigurdsson and coworkers [33], inspired by a phenoxazine based
deoxycytidine derivative [34]. Thermal denaturation and circular dichroism (CD)
experiments revealed a negligible effect on DNA duplex stability and conformation
[35]. Moreover, simple treatment of 15 with sodium sulfide at 45 C for 14 h causes
complete conversion to the corresponding amine [36]. This highly fluorescent species
can therefore also be used for optical spectroscopy.
The same group has reported that removal of the sugar moiety in 15 leads to 16
(Fig. 1) showing unique properties [37] regarding specific interactions with abasic
sites in DNA. At 30 C, 1 equiv. of 16 is reported to readily bind to the abasic site in
a noncovalent manner until saturation occurs. The high specificity of binding was
demonstrated via the small change of binding constants upon introduction of a tenfold
excess of unmodified DNA. Minor nonspecific binding was found to be in the order
of 5% at low temperatures. This novel labeling strategy is named noncovalent-
SDSL and has promising applications on natural occurring DNA molecules.
3 Theoretical Background
Fig. 3 EPR spectroscopy provides detailed information regarding metal ion coordination sphere,
interaction with small molecules, local and global dynamics, as well as overall conformation of
nucleic acid molecules
166 I. Krstic et al.
used to probe the local surrounding up to about 0.8 nm of the unpaired electron spin
in great detail. This is especially interesting for metal ions like Mn2+ replacing
physiologically crucial Mg2+ ions that play an important role in catalytic function
and tertiary structure stabilization of ribozymes. Finally, dipolar spectroscopy can
be applied to determine long range distances in NAs. This is a well established
method to study complex structures of NAs and their changes upon binding to other
molecules, and for the determination of the conformational flexibility of NAs under
various conditions. In the following, a brief introduction to the basic concepts of
these subclasses will be given. More detailed information regarding these methods
can be found in standard textbooks of modern EPR [48].
The shape of a nitroxide radical CW EPR spectrum reveals high sensitivity to its
rotational motion in liquid solutions. Thus, it can be used as a valuable tool for
monitoring the local and global dynamics of NAs on time scales defined by the EPR
frequency.
Nitroxide EPR spectra are determined by Zeeman interactions of the unpaired
electron and nuclear spins with the external magnetic field B and the hyperfine interac-
tion of the electron spin with the nitrogen nuclear spin (I1 for 14N and I1/2 for 15N)
[49]. In solid phase the magnetic energy of the electron-nucleus spin system can be
described by the static Hamiltonian
^ mN gN B^I ^IAS
H^ mB BgS ^ (1)
where mB and mN are Bohr and nuclear magneton, respectively and S and I are the
electron and nuclear spin vector operators, respectively. The g-tensor accounts for
the anisotropy of the electron spin Zeeman interaction, whereas the A-tensor
describes the field independent electron nuclear hyperfine interaction. The magnetic
tensors g and A are both approximately diagonal in the coordinate frame reflecting
the symmetry of the nitroxide radical. In order to obtain the Hamiltonian in the
laboratory coordinate system with the z-axis parallel to the magnetic field B, g- and
A-tensors have both to be transformed from the nitroxide molecular axis system to
the laboratory frame. Calculation of the energy levels and corresponding transition
probabilities for all possible radical orientations with respect to the magnetic field
with their appropriate statistical weight results in the so-called powder spectrum
represented in Fig. 4. This spectrum has a width of approximately 70 Gauss at
typical X-band frequencies (9.5 GHz) and represents the experimentally observed
spectra of nitroxide radicals in a disordered frozen solution sample.
However, the method of obtaining the ensemble spectra by a superposition of
spectra from molecules with all possible random orientations with respect to the
external magnetic field, as used above for frozen samples, is not applicable for
Structure and Dynamics of Nucleic Acids 167
mobile radicals in liquid solution. In this case the effective values of the g- and
A-tensors determining the Hamiltonian H in the laboratory frame are all time
dependent and do not result in stationary solutions for the eigenvalues and
eigenvectors of (1). In this case the density matrix has to be found for the system
based on the Liouville/von Neumann equation with a time dependent Hamiltonian
H(t) to describe the spectrum [50, 51]:
@^
r i^
Ht; ^
r (2)
@t
h
The general solution of the density matrix for arbitrary rotational correlation
time and ordering potential is non-trivial. In the case of very fast motion, meaning
that the rotational correlation time of the radical is much shorter than the combined
inverse anisotropies of the g- and A-tensor, the g- and A-tensors in (1) can be
substituted by their averaged values giso (gxx + gyy + gzz)/3 and Aiso (Axx +
Ayy + Azz)/3, respectively. In the high field approximation the Hamiltonian H
becomes diagonal, which allows solving the Liouville/von Neumann equation
exactly. The time evolution of the transversal magnetization M(t) is given by the
trace of the product of the density matrix with the spin rising operator
S^ S^x iS^y . Neglecting the nuclear Zeeman interaction, we obtain
mB giso B Aiso t
^ ^ exp i
Mt / Tr S r t 1 2 cos t exp (3)
h h T2
168 I. Krstic et al.
This expression describes the free induction decay (FID) signal of the spin
system after an intense and short mw-pulse, which usually cannot be realized
experimentally for a nitroxide. Nevertheless, for low microwave excitation power
the CW EPR spectra are given as the Fourier transformation of the FID signal (3).
For fast rotational motion this spectrum consists of three narrow hyperfine lines
of equal intensities at the spectral positions mB giso B Aiso =h, mB giso B=h; and
mB giso B Aiso =h. The width of the lines in this fast tumbling case is defined by
(p T2)1, where T2 is the transversal relaxation time of the electron spin. For
nitroxide radicals and magnetic field values of about 0.3 T this is valid for rotational
correlation times faster than 10 ps.
If the radical motion slows down a gradual transition from a highly resolved
three line spectrum of a liquid to the broad powder spectrum of a solid is observable
(Fig. 5). For rotational correlation times tc in the range of 0.011 ns, the Lorenzian
form of individual spectral lines can be described by the width parameter
T21 Do2 tc , being proportional to the ensemble average of the squared resonance
frequency fluctuations Do. Any dynamical process which causes the resonance
frequency to change will therefore contribute to the line width. Approximating
that anisotropy can be described by axial tensors [52], the modulation of the
10 ns
5 ns
1 ns
100 ps
10 ps
Fig. 5 Nitroxide EPR spectra as a function of the isotropic rotational correlation time for X-band
frequencies. Simulation is performed by EasySpin [53]
Structure and Dynamics of Nucleic Acids 169
mB B DA
Do Dg MI (4)
h
h
where MI is the nuclear spin quantum number and Dg and DA represent the
variation of Zeeman and hyperfine splitting, respectively. Inserting (4) in the
expression for T21 we obtain
1
a bMI cMI2 (5)
T2
which predicts the line widths dependence on the nuclear spin quantum number MI.
The coefficients a, b, and c are defined as follows:
h 2 t c
a Dg2 mB B= (6)
b DgDA mB B=
h2 t c (7)
c DA2 tc =
h2 (8)
^z
H^hfi Aiso^Iz S (9)
170 I. Krstic et al.
This interaction, which does not depend on the orientation of the molecule with
respect to the external magnetic field, gives valuable information on the local
structure, such as for example the bond strength and geometry of hydrogen bonds
to the radical or on the covalency of metal ligand bonds.
The second interaction is the magnetic dipoledipole interaction between the
unpaired electron and nuclear spins. It is short-ranged due to its R3 dependency
and the weak magnetic moment of the interacting nuclei. This term is strongly
anisotropic, meaning that it depends on the angle between the vector connecting the
unpaired electron and the nuclear spin and the external magnetic field. Whereas this
interaction is only a small perturbation for the electron spin at magnetic fields above
0.1 T with respect to the electron Zeeman interaction, it has a severe influence on
the nuclear spins. The nuclear spin eigenstates are mixed by the additional hyper-
fine field arising from the strong magnetic moment of the unpaired electron leading
to forbidden transitions further complicating CW EPR spectra. The according
spin Hamiltonian can be expressed as
^ Azz^Iz S
H^hfi ^I A S ^z Axz^Ix S
^z Ayz^Iy S
^z (10)
H^qi ^I Q ^I (11)
All these effects usually cause an inhomogeneous broadening of the EPR lines in
frozen disordered samples or large and slowly tumbling macromolecules leading to
broad and unresolved EPR spectra. In such cases, advanced pulsed and double
resonance EPR methods have to be utilized to disentangle these interactions to
nearby nuclei spins.
strong dependence on the external magnetic field. Typically the effect vanishes for
field strengths above 1 T, where the nuclear Zeeman-splitting exceeds the hyperfine
splitting. Furthermore, it can only be observed in solids or frozen samples, because
the anisotropy of the hyperfine interaction is averaged to zero in liquid solutions.
The method allows one to obtain distances to close by nuclei. Quantitative simula-
tion of the ESEEM effect can be used to determine the number of coupled nuclear
spins [59]. The simplest experiment is the 2-pulse Hahn-echo ESEEM (Fig. 6).
It has a low resolution as the signal decays with the short transverse relaxation time
T2 but it is very simple and thus very efficient for a small number of coordinated
Fig. 6 Pulse-Sequences, different pulse sequences for hyperfine and dipolar spectroscopy
(ESEEM, ENDOR, and PELDOR)
172 I. Krstic et al.
nuclear spins. The number of coupled nuclei spins is also extractable in favorable
cases. The 3-pulse ESEEM, based on a stimulated echo sequence, has a higher
spectral resolution. It suffers from so-called blind spots; certain hyperfine
frequencies are suppressed due to the selected pulse delay between the first and
second pulse. Repeating the 3-pulse ESEEM experiment with different pulse delays
usually circumvents this drawback. In addition, 4-pulse ESEEM and, in the 2D
version, HYSCORE further improve the spectral resolution of the ESEEM method.
The off-diagonal signals in HYSCORE correlate hyperfine frequencies arising from
one nuclear spin in the two different electron spin manifolds (mS 1/2). Espe-
cially in these 2D experiments a high sensitivity is obtained via a magnetic field
setting where the nuclear Zeemann splitting has approximately the same value as
the hyperfine coupling (cancelation condition) [60]. Otherwise matched pulses can
be applied to increase the modulation depth [61].
Here mk is the transversal magnetization of the kth spin at the time of the
refocused echo and lj is the inversion efficiency of the jth spin by the pump
pulse. Rkj and Ykj define the length and orientation with respect to the static
magnetic field of the vector connecting both electron spins, respectively. D is the
dipolar interaction constant, given as (2p 52.18) MHznm3 for two radicals.
In the case where the paramagnetic centers are randomly distributed within
the sample the average of (12) leads to an exponential expression of the signal
decay:
174 I. Krstic et al.
Fig. 7 PELDOR signal analysis. (a) Time domain PELDOR signal as a function of the delay time
T of the pump pulse. The dashed line shows the exponentially decaying intermolecular dipolar
contribution to the signal. (b) Time domain PELDOR signal after division of the original PELDOR
time domain data by the fit-function representing the intermolecular decay. (c) Fourier transform
of the PELDOR time trace (b) representing the dipolar Pake-pattern. (d) Distance distribution
function obtained from the PELDOR time traces (b) by Tikhonov regularization. From the last
representation the distances for spin pairs A-B can be the most easily extracted
4p2 l c
Vinter T V0 exp p DT (13)
9 3
where c is the spin concentration in the sample and l is the fraction of spins excited
by the inversion pulse. Typically, this formula is used to describe the intermolecular
part of the PELDOR signal decay for spin labeled molecules (dashed curve in
Fig. 7a). More interesting is the specific intramolecular interaction between the
spin labels attached to one individual NA or a specific NAprotein complex.
This intramolecular PELDOR signal for N spin labels per randomly oriented
macromolecules in the sample can be calculated according to [67] by using
p=2 ! !
1 XN Y N
DT
Vintra T 1ll cos 3 1 3cos y sin y dy
2
(14)
N k1 j6k 0 Rjk
Structure and Dynamics of Nucleic Acids 175
Z1 Zp=2
DT
VT 1 l l f R cos 1 3cos 2
y sin y dy dR (15)
R3
0 0
where f(R) describes the distance distribution function between the two spin labels.
This equation is most frequently used for the quantitative analysis of experimental
PELDOR time traces. It allows a parameter free extraction of the distance distribu-
tion function f(R) and therefore the determination of the average distance between
the two unpaired spins of the spin-labeled macromolecule from PELDOR
measurements. A typical example of such a data analysis is depicted in Fig. 7. In
a first step the intermolecular decay is separated from the original signal. The
experimental decay function at long times T is fitted with an exponential function
and then the experimental PELDOR time trace is divided by this function (Fig. 7b).
After that, the remaining PELDOR signal, which represents the intra-molecular
interactions Vintra, is analyzed to determine the average distance R or the distance
distribution function f(R). The singularities of the dipolar Pake pattern, obtained
after Fourier transformation, can be used to extract the average distance R (Fig. 7c),
or the Fredholm equation of the first kind (15) can be solved using Tikhonov
regularization methods to obtain the distance distribution function f(R).
However, it should be mentioned that this procedure is only valid for spin-pairs
(N 2) with a random mutual orientation. If the two spin labels have a fixed
relative orientation, the inversion efficiency depends on the specific orientation of
the molecule with respect to the external magnetic field, leading to a more elaborate
calculation of the PELDOR time trace:
Z Z
1 p=2
DT
VT 1 f R ly cos 1 3cos y 1 sin ydydR
2
(16)
0 0 R3
In general, the separation of the distance distribution function f(R) from the part
containing the mutual angular orientation between the two spin labels is more
cumbersome and not straightforward [6871]. However, it allows one to obtain
additional valuable information regarding structure and flexibility of the macro-
molecule under investigation, as will be shown in Sect. 4.2.2.
176 I. Krstic et al.
Fig. 8 The secondary structure of the neomycin-responsive riboswitch (left) and CW EPR data of
the single-labeled RNA samples at 0 C in 20 wt% sucrose solution (right)
Structure and Dynamics of Nucleic Acids 177
label 4 (Fig. 1) at either the duplex (U26), bulge (U7), or terminal loop region
(U15) [73].
In each region the nitroxide displays a different spectral width, reflecting the
different mobility of these parts of the RNA molecule. The most immobilized
nucleotide is located in the closing stem (U26). The spectrum of this site is
characterized by a splitting of both low field and high field peaks, due to partial
averaging of the anisotropic hyperfine tensor. The nucleotide U15 resides in the
terminal loop, which is a well known RNA motif called U-turn, defined by a
sequence UNRN (N-any nucleotide, R-purine). The conformation of this motif is
stabilized by hydrogen bonds between the imino proton of U and a phosphate
oxygen of the last nucleotide in the motif and between 20 -OH of U and N7 in the
purine base (R) as inferred from crystallographic data [74]. Thus, the spin label at
U15 reveals only a slightly narrower hyperfine splitting due to the high order of the
nucleotide U15 in the loop. A much higher mobility is expected for a bulge region,
which is the least structured part of the riboswitch, as known from NMR [75].
Indeed, the spin label attached to U7 shows the highest level of hyperfine averaging
and the narrowest EPR spectrum.
Conformational dynamics and tertiary structure of RNA molecules are essential
for their functions as catalysts, regulatory elements, or structural scaffolds. Tertiary
structural elements between helical and unpaired regions of the molecule, e.g.,
pseudoknots or kissing loops, are stabilized through van der Waals interactions,
specific hydrogen bonds, metal ions, or p-stacking interactions. The motional
dynamics of such motifs and the changes induced by their interaction with metal
ions, small organic molecules, and peptides can be probed by CW EPR spectros-
copy. Some illustrative examples of such applications will be described in the
following.
The interaction between the trans-activation responsive (TAR) RNA of the
human immunodeficiency virus (HIV) with the trans-activator of transcription
(Tat) protein is essential for production of full-length RNA transcripts during
viral replication and has been studied via CW EPR in detail by Sigurdsson and
collaborators [47, 7679]. The TAR RNA consists of two helical regions,
connected by a tri-nucleotide bulge which acts as a joint between them. The motion
between the two helices is strongly reduced upon addition of Ca2+ or cognate
peptide, leading to a coaxial stacking. The changes in the spectral width of the
CW EPR spectrum of spin label 21 (Fig. 2) were measured at four different labeling
positions (Fig. 9) upon interaction with metal ions, inhibitors, and the Tat motif.
From the dynamic signature obtained for each interaction partner, the authors
concluded that the conformational change does not require divalent metal ions
because Na+ induces the same dynamic response as Ca2+. They suggested that TAR
is able to adapt its conformation to accommodate binding of metal ions with
different size and coordination properties in the bulge region [77]. The dynamic
response predicts that Arginin 52 (R52) is essential for Tat binding and that
argininamide has a binding motive similar to that of the wild type Tat protein
[78, 79]. Further investigations from the same authors with point mutations at the
C- and N-terminal also anticipated a possible role of R56 for the Tat protein
178 I. Krstic et al.
Fig. 9 Changes in CW EPR spectral width at four different labeling positions of TAR as a
function of the interaction with calcium, sodium, argininamide, and Tat protein. The figure is from
[79] with permission of the journal
binding. This point mutation affected the mobility of nucleotides U23 and U38,
which are part of a triple base interaction with the Tat protein [76].
DeRose and coworkers have explored conformational changes of TAR RNA
upon binding of divalent metal ions (Ca2+) by measuring the dipolar coupling
between two attached spin labels 20 using CW EPR (Fig. 2). The U25U40
distances obtained from Fourier deconvolution methods are 11.9 0.3 A for
TAR RNA in the absence of divalent metal cations and 14.2 0.3 A when
50 mM Ca2+ was added [45]. These results are in accordance with the proposed
coaxial stacking of the two TAR helices upon addition of metal ions based on the
X-ray crystal structure [80].
The group I intron of Tetrahymena thermophila, the first identified catalytic
RNA molecule, is another example of a tertiary structured RNA which was
investigated by CW EPR [39]. The exon substrate of this ribozyme forms a
double-stranded region (duplex P1) with an internal guide sequence. Docking of
P1 controls the substrate cleavage. Qin and coworkers have examined the nanosec-
ond dynamics of the P1 duplex, with spin label R5a (bromo-substituted 18; see
Fig. 2) conjugated to the Sp-phosphorothioate diastereomer. CW EPR spectra of
two spin-labeled RNA sequences, designated as SCSL and SOSL in Fig. 10, favoring
either the open or closed conformation, revealed lower nitroxide mobility in the
closed state of the ribozyme. The authors attributed this decreased P1 duplex
dynamics to the requirement for proper positioning of the bound substrate for the
catalytic reaction. Additionally, they have shown that in the open state of the
ribozyme the length of the J1/2 junction can influence the P1 duplex mobility [39].
Interaction between the GNRA hairpin loop and the asymmetric internal loop
(receptor) is one of the most abundant long-range tertiary interaction motifs in large
RNAs, such as group I and II introns. Hubbell, Qin, and coworkers have
investigated the hairpinreceptor complex formation in the presence of Mg2+ ions
by CW EPR employing spin label 18 (Fig. 2) attached to an RNA hairpin containing
Structure and Dynamics of Nucleic Acids 179
Fig. 10 CW EPR probing of P1 helix docking. (a) Schematic renditions of the open and closed
ribozyme complex. The substrate sequences, the R5a spin label (asterisk), the cleavage site
(triangle), and the J1/2 junction are marked. (b) EPR spectra of various substrate/ribozyme
complexes. The figure is from [39] with permission of the journal
the GAAA tetraloop [41]. Monitoring the changes in the rotational correlation time
of the labeled RNA tetraloop, the formation of tetraloopreceptor complex due to
the increase in molecular size upon binding to the receptor has been detected. The
presence of Mg2+ ions was required for the complex formation, and the binding
constant was determined to be 0.40 0.05 mM. A free energy of 4.6 kcal/mol
has been calculated for this weak tetraloopreceptor interaction, assuming the
absence of other tertiary constraints.
Using spin-labeled nucleobase 2 (Fig. 1) incorporated in a tetraloop receptor
sequence (designated as TLR), Hubbell, Qin, and coworkers have studied confor-
mational changes induced by the GAAA tetraloop (designated as TL) docking in
the presence of Mg2+ [81]. As shown in Fig. 11, in the bound state of the receptor,
the mobility of the U19 base significantly increases, indicating unstacking of the
base upon tetraloop binding. This finding is consistent with the crystal structure of
a group I intron ribozyme domain containing the tetraloopreceptor complex, in
which the base equivalent to U19 is unstacked. On the other hand, spectral changes
were not detected at U5 and U17 within the receptor sequence, suggesting no
structural rearrangement in this local environment.
Most catalytically active RNAs require divalent metal ions, primarily Mg2+, to
accomplish full catalytic activity. CW EPR has been utilized to study the metal ion
180 I. Krstic et al.
Fig. 11 Conformational changes of the tetraloop receptor upon interaction with GAAA tetraloop
detected by CW EPR. (a) Schematic representations of receptor structures, with the dotted box
indicating the 11 nucleotide GAAA tetraloop receptor motif. (b) Spectra observed in the absence
(TL, black traces) and presence (TL, green traces) of tetraloop. The difference spectra (Diff,
red traces) were obtained by subtracting the TL spectra from the corresponding TL spectra.
This figure is from [81] with permission of the journal
binding affinities in the hammerhead ribozyme (HHRz) and the DielsAlder ribo-
zyme. For such studies, the naturally occurring diamagnetic Mg2+ has to be
substituted by the paramagnetic Mn2+ ion (S 5/2, I 5/2). Despite the fact
that the catalytic activity of ribozymes is reduced by metal exchange (even though
Structure and Dynamics of Nucleic Acids 181
both metal ions have similar ionic radii, charge, coordination geometry, and ligand
exchange rates [82]), it has been shown by X-ray crystallography that Mn2+ occupies
the same binding sites as Mg2+ in the minimal HHRz (mHHRz) [83, 84]. CW EPR
spectroscopy can easily distinguish the free Mn2+ ions in a hexaaqua complex in
liquid solution from Mn2+ ions bound to the ribozyme, thus allowing one to count
binding sites and to determine binding constants by titration of Mn2+ into the solution.
The mHHRz contains a single Mn2+ binding site with an affinity of 10 mM in 1 M
NaCl solution, whereas at lower ionic strength (0.1 M NaCl) four binding sites are
identified with an overall Kd of 4 mM [85]. The aminoglycoside antibiotic neomycin
B inhibits the catalytic activity of this ribozyme [86]. Prisner, Schiemann, and
coworkers could show that the bound Mn2+ ion is released upon binding of
neomycin B (Kd 1.2 mM) to the hammerhead ribozyme. The tertiary stabilized
extended HHRz possesses a single high-affinity Mn2+ binding site with a Kd 10 nM
at an NaCl concentration of 0.1 M. This dissociation constant is more than two orders
of magnitude smaller than the Kd determined for the single high-affinity Mn2+ site in
the mHHRz [87]. The addition of two loops that interact between stems I and II leads
to an enhanced cleavage activity of HHRz at low Mg2+ concentrations [88]. Kim
et al. have probed the conformational dynamics of this structural element as
a function of concentration of added Mg2+ by using the spin-labeled nucleotides 21
(Fig. 2) located in the substrate stem. The splitting of both low field and high field
lines of the nitroxide spectra observed for the spin-labeled U1.6 (position 6 in stem I)
are ascribed to the docking of stems I and II originating from the folding of the HHRz.
Spectra recorded at different Mg2+ concentrations were fitted by two sites with
distinctive mobility. Thus, it was concluded that the docking of stems I and II occurs
at low Mg2+ concentrations (1 mM) [88]. A detailed description of the structure of
the metal binding site in HHRz has been achieved by using hyperfine spectroscopy as
will be described in more detail in Sect. 4.2.1.
Similar CW EPR titration experiments have been performed by Schiemann and
coworkers on the DielsAlderase ribozyme [89]. It was demonstrated that five high-
affinity Mn2+ binding sites with an upper Kd of 0.6 0.2 mM exist in this ribozyme.
From the spectral changes on stoichiometric mixtures of Mn2+ and Cd2+ the binding
sites could be assigned to three different types: inner sphere, outer sphere, and
a dimeric site.
RNAprotein interactions play a central role in cellular processes; therefore the
investigation of structure, conformational dynamics of such complexes, and their
relation to biological function is of major importance. Again, CW EPR spectros-
copy can help to provide detailed understanding of such processes and we will state
two selected examples in the following.
The interactions between a 22-amino acid long peptide (called N-peptide) and
a nascent mRNA hairpin element (called boxB) are of a crucial function in tran-
scription anti-termination. Qin and coworkers have studied this N-peptide/boxB
complex by monitoring the nanosecond dynamics of nitroxide spin labels attached
to different positions of the N-peptide [90]. Analysis of CW EPR spectra recorded
in viscous aqueous solutions at 5 C indicates that the bound N-peptide exists in
a dynamic equilibrium between two different conformational states, with the
182 I. Krstic et al.
peptide C-terminus either stacking on the RNA loop or pointing away from the
peptide/RNA interface. At the same time, the N-terminus of the N-peptide adopts
a single well defined conformation, confirming that the arginine-rich motif accounts
for boxB binding. Based on these findings, the authors suggested that the stacked
state is responsible for the function of the N-peptide/boxB [90]. Another example is
the interaction between a 20-mer HIV-1 RNA stem loop 3 and the HIV-1 nucleo-
capsid Zn-finger protein NCp7, investigated by Scholes and coworkers using
a stopped-flow CW EPR setup, with a time resolution of about 4 ms [91]. In their
experiments, spin-labeled RNA stem loop 3 was mixed with the protein NCp7 in
a ratio of 4:1 and the decrease in amplitude of the central line of the nitroxide EPR
spectrum recorded as a function of reaction time. The stopped-flow CW EPR signal
showed that the main interaction of NCp7 with the RNA stem loop 3 occurs within
the dead time of the apparatus (<4 ms); a second contribution with a time constant
of ~30 ms and a much slower immobilization over seconds (possibly concomitant
with large complex formation).
Throughout the last few years, pulse EPR experiments have been increasingly
applied to study NAs. As explained in Sect. 3, such methods can be subdivided
into two major categories: hyperfine and dipolar spectroscopy. Hyperfine spectros-
copy has so far been used mainly in the context of NAs to probe the binding of Mn2+
metal ions to RNA structures, whereas dipolar spectroscopy has found more
widespread applications in the investigation of secondary and tertiary structure
and conformational changes on NAs.
184 I. Krstic et al.
Fig. 12 Summary of hyperfine spectroscopy results on Mn2+ bound to HHRz. Top right: Q-band
phosphor ENDOR showing a hyperfine splitting consistent with a inner-sphere coordination to the
phosphor-diester oxygen. Lower right: Fourier transform of X-band ESEEM measurements
showing the coordination to the nitrogen of dG10.1. Lower left: X-band ESEEM spectroscopy
of HHRz in deuterated water. A quantitative comparison with simulations allows to determine the
number of coordinated water molecules. Figure adapted from [106] with permission of the journal
Structure and Dynamics of Nucleic Acids 185
metal ion retains four water ligands as measured by 1H/2H HYSCORE spectroscopy
for tsHHRz and mHHRz [86, 87, 103].
Another application of ENDOR and HYSCORE techniques on NAs represents
the study of the coordination environment of Cu2+ ions in poly (dG-dC) (dG-dC)
polymer reported by Jeschke and coworkers [107]. Two different copper species
were detected and analyzed. Copper I was characterized by a weak 14N coupling
and a moderate 1H coupling, which are assigned to a Cu2+ binding to N7 and H8
atoms of a single guanine. The authors proposed that species II forms coordinative
bonds with N7 and H8 atoms of a guanine and with N3 and H5 of cytosine in a
metal mediated Hoogsteen base pair.
Within the last 10 years PELDOR [108110] has demonstrated its applicability and
reliability to provide structural information on NAs. In the following we want to
highlight some applications of this method on double-helical and more complex
doubly spin-labeled DNA and RNA molecules to determine distances and distance
changes with high accuracy.
Fig. 14 Orientation selective PELDOR experiments on ds-DNA molecules with a rigid spin
label 15 incorporated. The geometry between the out-of-plane vectors with respect to the
interconnecting vector can be easily followed by measuring PELDOR as a function of the observer
frequency, as shown on the right side. Figure adapted from [113]
This spin label acts as a rigid cantilever, attached to the DNA molecule, and
therefore allows a very precise determination of the conformational flexibility of
double-stranded DNA as stretching, twisting, and bending [114].
Selectively attached spin probe 8 (Fig. 1) was employed by Sicoli et al. to
monitor quantitatively B-A conformational change of double-stranded DNA
[115]. The B-A transition was induced by a high concentration of trifluoroethanol
(higher than 70 vol.%) and very good agreement between the distances measured
by PELDOR and the average distance obtained from MD simulations was achieved.
These authors used the same spin label to study subtle structural changes in DNA
induced by lesions as a model for DNA damaging. Distance measurements on a
set of undamaged and damaged DNA duplexes that contain 8-oxoguanine, a nick,
a gap, a bulge, an abasic site analog, and an anucleosidic site were supported by
molecular dynamics studies [116].
Hobartner and coworkers have recently employed spin label 11 (Fig. 1) to detect
simultaneously two competing structures of the incompletely self-complementary
RNAs, the hairpin and duplex [30]. In the hairpin conformation the two spin labels
are 6 bp apart, resulting in a distance of 1.8 nm (Fig. 15). Upon addition of
complementary RNA strand the hairpin structure becomes disrupted and a continu-
ous 20-bp duplex is formed. In the newly formed helical structure the two TEMPO
groups are 11 bp apart, yielding a distance distribution centered at 3.1 nm. By
increasing the amounts of complementary RNA the ratio between the two coexisting
structures shifts completely towards the RNA duplex.
Krstic et al. applied PELDOR spectroscopy for the first time to map the global
structure of a tertiary folded RNA molecule, precisely the neomycin-responsive
188 I. Krstic et al.
Fig. 15 Two RNA sequences, named 7 and 12, individually fold into a hairpin conformation but
form an extended duplex when mixed. PELDOR experiments for different ratios of 7 and 12 are
shown together with the distance distribution functions obtained from Tikhonov regularization of
the PELDOR time traces. Integration of the peaks in the distance distribution allows a precise
quantification of the different RNA structures in the mixture. Figure adapted from [30]
Fig. 16 (a) Background corrected PELDOR time traces and distance distribution functions of
a double-labeled neomycin-responsive riboswitch in the presence (red) and absence (black) of the
ligand. The measured distances are in good agreement with the NMR structure determined for the
neomycin-bound state of the riboswitch. Figure adapted from [73]. (b) Model of labeled neomy-
cin-sensitive riboswitch based on the NMR structure (pdb-code: 2kxm) with spin labels attached to
C5-positions of U4, U14, U15, and U26
good agreement with the structure of the ligand-bound state of the riboswitch
determined by NMR spectroscopy [73]. This indicates the intrinsic propensity of
the global RNA architecture toward its energetically favored ligand-bound form
[73].
The structure of human telomeric RNA repeat in the presence of K+ ions has
been investigated by Sicoli et al. [30]. Spin label 12 (Fig. 1) was introduced into the
trinucleotide loop regions and efficient quadruplex formation was confirmed by UV
melting and circular dichroism experiments. The distance of 3.7 0.2 nm obtained
from high quality PELDOR data is in accordance with the NMR structure of
unmodified RNA, supporting the formation of a parallel-stranded G-quadruplex
conformation. Drescher and coworkers have studied conformations of human
telomeric DNA repeat in solutions containing different counter ions, employing
spin label 7 (Fig. 1). They inferred that the quadruplex in the presence of K+ ions
experiences both the propeller and the basket conformation, whereas in Na+ ions
containing solution the antiparallel form is the only conformation present [118].
Spin probe 21 (Fig. 1) was introduced into stems I and II of the extended
hammerhead ribozyme and folding process was investigated by measuring changes
in spin-spin distances upon addition of Mg2+ ions [119]. DeRose and coworkers
190 I. Krstic et al.
Fig. 17 Overlaid distance distribution functions for spin-labeled extended HHRz in 0.1 M NaCl,
pH 7.0 at Mg2+ concentrations of 0 (black), 1 (red), 10 (blue), 25 (green), and 50 mM (gold). The
black dashed line is for the isolated spin-labeled substrate in 1 mM Mg2+ and 0.1 M NaCl (left).
Model of spin-labeled HHRz created from the crystal structure (PDB ID 2GOZ). Substrate is
shown in blue and the enzyme is in magenta. The spin labels are shown in green with dashed
circles (right). Figure adapted from [119]
reported that at low ionic strength the two stems of HHRz are randomly oriented
resulting in very broad distance distribution. In contrast, at 10 mM and higher Mg2+
concentrations this RNA molecule is globally folded and the measured distance of
~ 2.4 nm is in accordance to a simple model constructed from a crystal structure of
this tertiary folded ribozyme (Fig. 17).
Recently, Steinhoff and coworkers have used the same labeling procedure to
study the conformational transition of a tetracycline aptamer upon ligand binding.
They concluded that this synthetic riboswitch exhibits a thermodynamic equilib-
rium between two conformations in the ligand-free state and captures one confor-
mation upon ligand binding [120].
Fig. 19 Secondary structure (with spin-labeled nucleotides in red), baseline corrected PELDOR
time traces and distance distribution for 12-bp double-labeled DNA in vitro (green) and in-cell
(incubation time indicated in legend). Data were fitted with two Gaussian functions. Figure
adapted from [121]
Acknowledgments Our own EPR work relied on the synthesis of spin-labeled RNA and DNA
molecules performed by Nelly Piton and Olga Romainczyk from the group of Joachim W. Engels
(Institute of Organic Chemistry and Chemical Biology, Goethe University Frankfurt) and the work
by Pavol Cekan from the group of Snorri Th. Sigurdsson (University of Iceland) on the rigid spin
labels for DNA. Vasyl Denysenkov is thanked for high-field G-band PELDOR experiments on
DNA samples. Olav Schiemann (University of St. Andrews) is thanked for his major impact in the
initial phase of this work as Habilitand in Frankfurt. Funding from the German Research Society
(DFG) within the Collaborative Research Center 579 RNA-Ligand Interaction is gratefully
acknowledged as well as support from the Center of Biomolecular Magnetic Resonance
(BMRZ) and the Center of Excellence Frankfurt Macromolecular Complexes (DFG).
References
32. Okonogi TM, Alley SC, Reese AW, Hopkins PB, Robinson BH (2000) Sequence-dependent
dynamics in duplex DNA. Biophys J 78(5):25602571
33. Barhate N, Cekan P, Massey AP, Sigurdsson ST (2007) A nucleoside that contains a rigid
nitroxide spin label: a fluorophore in disguise. Angew Chem Int Ed Engl 46(15):26552658.
doi:10.1002/anie.200603993
34. Lin KY, Jones RJ, Matteucci M (1995) Tricyclic 20 -deoxycytidine analogs syntheses and
incorporation into oligodeoxynucleotides which have enhanced binding to complementary
RNA. J Am Chem Soc 117(13):38733874
35. Cekan P, Smith AL, Barhate N, Robinson BH, Sigurdsson ST (2008) Rigid spin-labeled
nucleoside C: a nonperturbing EPR probe of nucleic acid conformation. Nucleic Acids Res 36
(18):59465954. doi:10.1093/nar/gkn562
36. Cekan P, Sigurdsson ST (2008) Single base interrogation by afluorescent nucleotide: each of
the four DNA bases identified by fluorescence spectroscopy. Chem Commun
(29):33933395. doi:10.1039/B801833b
37. Shelke SA, Sigurdsson ST (2010) Noncovalent and site-directed spin labeling of nucleic
acids. Angew Chem Int Ed Engl 49(43):79847986. doi:10.1002/anie.201002637
38. Nagahara S, Murakami A, Makino K (1992) Spin-labeled oligonucleotides site specifically
labeled at the internucleotide linkage separation of stereoisomeric probes and EPR
spectroscopical detection of hybrid formation in solution. Nucleos Nucleot 11(24):889901
39. Grant GP, Boyd N, Herschlag D, Qin PZ (2009) Motions of the substrate recognition duplex
in a group I intron assessed by site-directed spin labeling. J Am Chem Soc 131(9):31363137.
doi:10.1021/ja808217s
40. Qin PZ, Haworth IS, Cai Q, Kusnetzow AK, Grant GPG, Price EA, Sowa GZ, Popova A,
Herreros B, He H (2007) Measuring nanometer distances in nucleic acids using a sequence-
independent nitroxide probe. Nat Protoc 2(10):23542365. doi:10.1038/nprot.2007.308
41. Qin PZ, Butcher SE, Feigon J, Hubbell WL (2001) Quantitative analysis of the isolated
GAAA tetraloop/receptor interaction in solution: a site-directed spin labeling study. Bio-
chemistry-US 40(23):69296936
42. Cai Q, Kusnetzow AK, Hideg K, Price EA, Haworth IS, Qin PZ (2007) Nanometer distance
measurements in RNA using site-directed spin labeling. Biophys J 93(6):21102117.
doi:10.1529/biophysj.107.109439
43. Cai Q, Kusnetzow AK, Hubbell WL, Haworth IS, Gacho GPC, Van Eps N, Hideg K,
Chambers EJ, Qin PZ (2006) Site-directed spin labeling measurements of nanometer
distances in nucleic acids using a sequence-independent nitroxide probe. Nucleic Acids
Res 34(17):47224730. doi:10.1093/nar/gkl546
44. Flaender M, Sicoli G, Fontecave T, Mathis G, Saint-Pierre C, Boulard Y, Gambarelli S,
Gasparutto D (2008) Site-specific insertion of nitroxide-spin labels into DNA probes by click
chemistry for structural analyses by ELDOR spectroscopy. Nucleic Acids Symp Ser
52:147148. doi:10.1093/nass/nrn075
45. Kim NK, Murali A, DeRose VJ (2004) A distance ruler for RNA using EPR and site-directed
spin labeling. Chem Biol 11(7):939948. doi:10.1016/j.chembiol.2004.04.013
46. Edwards TE, Sigurdsson ST (2007) Site-specific incorporation of nitroxide spin-labels into
20 -positions of nucleic acids. Nat Protoc 2(8):19541962. doi:10.1038/nprot.2007.273
47. Edwards TE, Okonogi TM, Robinson BH, Sigurdsson ST (2001) Site-specific incorporation
of nitroxide spin-labels into internal sites of the TAR RNA; structure-dependent dynamics of
RNA by EPR spectroscopy. J Am Chem Soc 123(7):15271528
48. Schweiger A, Jeschke G (2001) Principles of pulse electron paramagnetic resonance. Oxford
University Press, USA
49. Atherton NM (1993) Principles of electron spin resonance. Ellis Horwood, New York
50. Abragam A (1961) The principles of nuclear magnetism. The international series of
monographs on physics. Clarendon, Oxford
51. Freed JH, Fraenkel GK (1963) Theory of linewidths in electron spin resonance spectra.
J Chem Phys 39(2):326
Structure and Dynamics of Nucleic Acids 195
52. Kivelson D (1960) Theory of ESR linewidths of free radicals. J Chem Phys 33(4):10941106
53. Stoll S, Schweiger A (2006) EasySpin, a comprehensive software package for spectral
simulation and analysis in EPR. J Magn Reson 178(1):4255. doi:10.1016/j.jmr.2005.08.013
54. Schneider DJ, Freed JH (1989) Calculating slow motional magnetic resonance spectra:
a users guide. In: Berliner LJ, Reuben J (eds) Spin labeling: theory and applications.
Biological magnetic resonance. Plenum Press, New York
55. Mims WB, Nassau K, Mcgee JD (1961) Spectral diffusion in electron resonance lines. Phys
Rev 1(6):20592069
56. Rowan LG, Hahn EL, Mims WB (1965) Electron-spin-echo envelope modulation. Phys Rev
137(1A):6171
57. Mims WB (1972) Electron spin echoes. In: Geschwind S (ed) Electron paramagnetic reso-
nance. Plenum Press, New York
58. Hofer P, Grupp A, Nebenfuhr H, Mehring M (1986) Hyperfine sublevel correlation
(HYSCORE) spectroscopy a 2D electron-spin-resonance investigation of the squaric acid
radical. Chem Phys Lett 132(3):279282
59. Ichikawa T, Kevan L, Bowman MK, Dikanov SA, Tsvetkov YD (1979) Ratio analysis of
electron spin echo modulation envelopes in disordered matrixes and application to the
structure of solvated electrons in 2-methyltetrahydrofuran glass. J Chem Phys 71
(3):11671174
60. Flanagan HL, Singel DJ (1987) Analysis of 14N ESEEM patterns of randomly oriented solids.
J Chem Phys 87(10):56065616
61. Jeschke G, Rakhmatullin R, Schweiger A (1998) Sensitivity enhancement by matched
microwave pulses in one- and two-dimensional electron spin echo envelope modulation
spectroscopy. J Magn Reson 131(2):261271
62. Feher G (1956) Observation of nuclear magnetic resonances via the electron spin resonance
line. Phys Rev 103(3):834834
63. Davies ER (1974) A new pulse endor technique. Phys Lett A 47(1):12
64. Mims WB (1965) Pulsed electron nuclear double resonance (E.N.DO.R.) experiments. Proc
Roy Soc (London) 283(Ser. A;1395):452457
65. Epel B, Manikandan P, Kroneck PMH, Goldfarb D (2001) High-field ENDOR and the sign of
the hyperfine coupling. Appl Magn Reson 21(34):287297
66. Cook RJ, Whiffen DH (1964) Relative signs of hyperfine coupling constants by a double
ENDOR experiment. Proc Phys Soc Lond 84(6):845845
67. Bode BE, Margraf D, Plackmeyer J, Durner G, Prisner TF, Schiemann O (2007) Counting the
monomers in nanometer-sized oligomers by pulsed electron-electron double resonance. J Am
Chem Soc 129(21):67366745. doi:10.1021/ja065787t
68. Marko A, Margraf D, Yu H, Mu Y, Stock G, Prisner T (2009) Molecular orientation studies
by pulsed electron-electron double resonance experiments. J Chem Phys 130(6):064102.
doi:10.1063/1.3073040
69. Margraf D, Cekan P, Prisner TF, Sigurdsson ST, Schiemann O (2009) Ferro- and antiferro-
magnetic exchange coupling constants in PELDOR spectra. Phys Chem Chem Phys 11
(31):67086714. doi:10.1039/b905524j
70. Marko A, Margraf D, Cekan P, Sigurdsson ST, Schiemann O, Prisner TF (2010) Analytical
method to determine the orientation of rigid spin labels in DNA. Phys Rev E 81(2)
71. Margraf D, Bode BE, Marko A, Schiemann O, Prisner TF (2007) Conformational flexibility
of nitroxide biradicals determined by X-band PELDOR experiments. Mol Phys 105
(1516):21532160
72. Zhang XJ, Cekan P, Sigurdsson ST, Qin PZ (2009) Studying RNA using site-directed spin-
labeling and continuous-wave electron paramagnetic resonance spectroscopy. Method
Enzymol 469:303328. doi:10.1016/S0076-6879(09)69015-7
73. Krstic I, Frolow O, Sezer D, Endeward B, Weigand JE, Suess B, Engels JW, Prisner TF
(2010) PELDOR spectroscopy reveals preorganization of the neomycin-responsive
riboswitch tertiary structure. J Am Chem Soc 132(5):14541455. doi:10.1021/ja9077914
196 I. Krstic et al.
74. Moore PB (1999) Structural motifs in RNA. Annu Rev Biochem 68:287300. doi:10.1146/
annurev.biochem.68.1.287
75. Duchardt-Ferner E, Weigand JE, Ohlenschlager O, Schtnidtke SR, Suess B, Wohnert J (2010)
Highly modular structure and ligand binding by conformational capture in a minimalistic
riboswitch. Angew Chem Int Ed Engl 49(35):62166219
76. Edwards TE, Robinson BH, Sigurdsson ST (2005) Identification of amino acids that promote
specific and rigid TAR RNA-Tat protein complex formation. Chem Biol 12(3):329337.
doi:10.1016/j.chembiol.2005.01.012
77. Edwards TE, Sigurdsson ST (2003) EPR spectroscopic analysis of TAR RNA-metal ion
interactions. Biochem Biophys Res Commun 303(2):721725
78. Edwards TE, Sigurdsson ST (2002) Electron paramagnetic resonance dynamic signatures of
TAR RNA-small molecule complexes provide insight into RNA structure and recognition.
Biochemistry-US 41(50):1484314847
79. Edwards TE, Okonogi TM, Sigurdsson ST (2002) Investigation of RNA-protein and RNA-
metal ion interactions by electron paramagnetic resonance spectroscopy. The HIV TAR-Tat
motif. Chem Biol 9(6):699706
80. Ippolito JA, Steitz TA (1998) A 1.3-A resolution crystal structure of the HIV-1 trans-
activation response region RNA stem reveals a metal ion-dependent bulge conformation.
Proc Natl Acad Sci USA 95(17):98199824
81. Qin PZ, Feigon J, Hubbell WL (2005) Site-directed spin labeling studies reveal solution
conformational changes in a GAAA tetraloop receptor upon Mg2+-dependent docking of
a GAAA tetraloop. J Mol Biol 351(1):18. doi:10.1016/j.jmb.2005.06.007
82. Feig AL (2000) The use of manganese as a probe for elucidating the role of magnesium ions
in ribozymes. Met Ions Biol Syst 37:157182
83. Pley HW, Flaherty KM, Mckay DB (1994) 3-Dimensional structure of a hammerhead
ribozyme. Nature 372(6501):6874
84. Scott WG, Murray JB, Arnold JRP, Stoddard BL, Klug A (1996) Capturing the structure of
a catalytic RNA intermediate: the hammerhead ribozyme. Science 274(5295):20652069
85. Horton TE, Clardy DR, DeRose VJ (1998) Electron paramagnetic resonance spectroscopic
measurement of Mn2+ binding affinities to the hammerhead ribozyme and correlation with
cleavage activity. Biochemistry-US 37(51):1809418101
86. Schiemann O, Fritscher J, Kisseleva N, Sigurdsson ST, Prisner TF (2003) Structural investi-
gation of a high-affinity MnII binding site in the hammerhead ribozyme by EPR spectroscopy
and DFT calculations. Effects of neomycin B on metal-ion binding. Chembiochem 4
(10):10571065. doi:10.1002/cbic.200300653
87. Kisseleva N, Khvorova A, Westhof E, Schiemann O (2005) Binding of manganese(II) to
a tertiary stabilized hammerhead ribozyme as studied by electron paramagnetic resionance
spectroscopy. RNA 11(1):16
88. Kim NK, Murali A, DeRose VJ (2005) Separate metal requirements for loop interactions and
catalysis in the extended hammerhead ribozyme. J Am Chem Soc 127(41):1413414135.
doi:10.1021/ja0541027
89. Kisseleva N, Kraut S, Jaschke A, Schiemann O (2007) Characterizing multiple metal ion
binding sites within a ribozyme by cadmium-induced EPR silencing. HFSP J 1(2):127136
90. Zhang X, Lee SW, Zhao L, Xia T, Qin PZ (2010) Conformational distributions at the
N-peptide/boxB RNA interface studied using site-directed spin labeling. RNA 16
(12):24742483. doi:10.1261/rna.2360610
91. Xi X, Sun Y, Karim CB, Grigoryants VM, Scholes CP (2008) HIV-1 nucleocapsid protein
NCp7 and its RNA stem loop 3 partner: rotational dynamics of spin-labeled RNA stem loop
3. Biochemistry-US 47(38):1009910110. doi:10.1021/bi800602e
92. Spaltenstein A, Robinson BH, Hopkins PB (1989) DNA structural data from a dynamics
probe the dynamic signatures of single-stranded, hairpin-looped, and duplex forms of DNA
are distinguishable. J Am Chem Soc 111(6):23032305
Structure and Dynamics of Nucleic Acids 197
112. Ward R, Keeble DJ, El-Mkami H, Norman DG (2007) Distance determination in heteroge-
neous DNA model systems by pulsed EPR. Chembiochem 8(16):19571964. doi:10.1002/
cbic.200700245
113. Schiemann O, Cekan P, Margraf D, Prisner TF, Sigurdsson ST (2009) Relative orientation of
rigid nitroxides by PELDOR: beyond distance measurements in nucleic acids. Angew Chem
Int Ed Engl 48(18):32923295. doi:10.1002/anie.200805152
114. Marko A, Denysenkov VP, Margraf D, Cekan P, Schiemann O, Sigurdsson ST, Prisner TF
(2011) Conformational flexibility of DNA. J Am Chem Soc. doi:10.1021/ja201244u
115. Sicoli G, Mathis G, Delalande O, Boulard Y, Gasparutto D, Gambarelli S (2008) Double
electron-electron resonance (DEER): a convenient method to probe DNA conformational
changes. Angew Chem Int Ed Engl 47(4):735737. doi:10.1002/anie.200704133
116. Sicoli G, Mathis G, Aci-Seche S, Saint-Pierre C, Boulard Y, Gasparutto D, Gambarelli S
(2009) Lesion-induced DNA weak structural changes detected by pulsed EPR spectroscopy
combined with site-directed spin labelling. Nucleic Acids Res 37(10):31653176.
doi:10.1093/nar/gkp165
117. Weigand JE, Sanchez M, Gunnesch EB, Zeiher S, Schroeder R, Suess B (2008) Screening for
engineered neomycin riboswitches that control translation initiation. RNA 14(1):8997.
doi:10.1261/rna.772408
118. Singh V, Azarkh M, Exner TE, Hartig JS, Drescher M (2009) Human telomeric quadruplex
conformations studied by pulsed EPR. Angew Chem Int Ed Engl 48(51):97289730.
doi:10.1002/anie.200902146
119. Kim NK, Bowman MK, DeRose VJ (2010) Precise mapping of RNA tertiary structure via
nanometer distance measurements with double electron-electron resonance spectroscopy.
J Am Chem Soc 132(26):88828884. doi:10.1021/Ja101317g
120. Wunnicke D, Strohbach D, Weigand JE, Appel B, Feresin E, Suess B, M uller S, Steinhoff HJ
(2010) Ligand-induced conformational capture of a synthetic tetracycline riboswitch
revealed by pulse EPR. RNA. doi:10.1261/rna.2222811
121. Krstic I, Hansel R, Romainczyk O, Engels JW, D otsch V, Prisner TF (2011) Long-range
distance measurements on nucleic acids in cells by pulsed EPR spectroscopy. Angew Chem
Int Ed Engl 50(22):50705074. doi:10.1002/anie.201100886
Top Curr Chem (2012) 321: 199234
DOI: 10.1007/128_2011_303
# Springer-Verlag Berlin Heidelberg 2011
Published online: 11 November 2011
J. van Slageren
Contents
1 Introduction: Overview of EPR in Molecular Nanomagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
2 Unconventional EPR Measurement Methods in Molecular Nanomagnetism . . . . . . . . . . . . 204
2.1 Frequency Domain Magnetic Resonance Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
2.2 Interferometer-Based Fourier-Transform Terahertz Spectroscopy . . . . . . . . . . . . . . . . . 211
2.3 Terahertz Time-Domain Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
2.4 Comparison of EPR Measurement Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
2.5 Magnetization Detected EPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
3 Pulse EPR in Molecular Nanomagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
3.1 SpinLattice Relaxation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
3.2 Phase-Memory Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.3 Future of Pulse EPR in Molecular Magnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
4 Emerging Trends and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
effect, because of their large spin state degeneracy [22]. Because these magnetic
cooling MNMs have small anisotropies to provide the necessary large degeneracies,
EPR plays a less prominent role in this area, beyond materials characterization.
A final area where EPR has not played a decisive role so far, but possibly could, is
that of single-chain magnets (SCMs). SCMs are spin chains that combine ferro-
[23] or ferrimagnetic [24] interactions with Ising-type anisotropy, resulting in slow
relaxation of the magnetization that is governed by concerted action of both the
exchange coupling and the zero-field splitting of the individual units.
Most of the EPR studies of MNMs have focused on the investigation of the
magnetic anisotropy in the form of zero-field splitting, which is responsible for both
the energy barrier towards relaxation of the magnetization and for the quantum
tunneling process of the magnetization. Because ZFS is a field-independent inter-
action, it is advantageous to carry out EPR-measurements at a range of different
frequencies. Extrapolation of the dependence of frequency on resonance field to
zero field yields the splitting of the spin multiplet in zero field, i.e., the ZFS (see
Fig. 1 for a recent example). Different aspects of the origin of ZFS in MNM have
been described in various books [2730]. The ZFS of the ground spin multiplet can
Fig. 1 Top: High-field EPR spectra recorded along the easy axis of a single crystal
of [Mn6IIIO2(Et-sao)6(O2CPh(Me)2)2(EtOH)6] at 331 GHz and different temperatures as indicated,
and frequencies as indicated. Bottom: Diagram of microwave frequency as a function of resonance
field. adapted from [25]. Used with permission. #2009 Elsevier. See also Fig. 2
202 J. van Slageren
X
^ D S^2 1 SS 1 E S^2 S^2
H mB B g S z 3 x y Bqk O^qk (1)
k 4;q
where the first term is the Zeeman interaction between electron spin and magnetic
field, and the second and third terms are the second order axial and transverse ZFS,
respectively. Here, second order reflects the fact that the spin Hamiltonian contains
the spin operators to the second power. The last term is the sum of higher (than
second) order ZFS terms, which were shown to parametrize the effects of the ZFS-
induced mixing between spin states on the level splittings within the ground multi-
plet (S-mixing) [31]. In the case of SMMs or, generally, MNMs with large, negative
D-values, powder spectra or single crystal spectra recorded with the external field
along the zero-field splitting quantization axis (the easy axis, if D < 0) are often
not very informative regarding the transverse ZFS (the third term in (1) and the
elements of the fourth term with q 6 0), because the energies of the MS states with
large MS quantum number are little affected by transverse ZFS. It is this transverse
ZFS that is responsible for quantum tunneling of the magnetization. Much more
informative in that regard are single-crystal EPR spectra that are recorded with
the field perpendicular to the quantization axis (in the hard plane). For example,
in a recent study on [Mn12O12(tBuCH2CO2)16(CH3OH)4]CH3OH, this approach
allowed the determination of the transverse ZFS terms up to sixth order [32].
There have been a number of reviews of EPR studies of MNMs in the past
5 years [3337] especially by Barra/Gatteschi and McInnes, and it is not the aim of
the present chapter to repeat the content of those reviews in detail. From these
reviews, it is clear that certain EPR studies of MNMs have become classics in
their field and they are briefly outlined below. The reader is referred to the cited
reviews and primary literature for further detail. In addition, the quasi-biannual
reports by Collison and McInnes provide useful comprehensive overviews of EPR
studies on exchange coupled clusters, including biological systems, for the relevant
years [3841].
Studies on a six-membered, ferromagnetically coupled copper(II) ring were an
early triumph of high-frequency EPR in molecular nanomagnetism because of
the textbook-quality, perfectly axial spectra [35, 42]. Rather more investigations
have been carried out on antiferromagnetic rings which, for an even number of
equal metal ions, have S 0 ground states. EPR measurements have been suc-
cessfully employed to determine the ZFSs of the spin excited states in [Fe6(tea)6]
(tea3 is triethanolaminate) [35, 43], [Cr8F8Piv16] (Cr8, Piv pivalate) [35, 44],
and [CsFe8(tea)8]Cl [45]. In contrast, the antiferromagnetic rings (R2NH2)+
[CrIII7MIIF8(Piv)16] , and (Cr7M) have spin ground states of S 1/2, 1, or 3/2
[46]. Again, EPR investigations were very fruitful and yielded a wealth of
information on ground and excited states [36, 47, 48]. The ZFSs in Cr8 were
shown to result from a major contribution of the combined single-ion
New Directions in Electron Paramagnetic Resonance Spectroscopy 203
correlation between the axial zero-field splitting value D and the helical pitch of the
molecular structure was established.
To conclude this section, it is clear that over the past two decades EPR spectro-
scopic studies on MNMs have moved from sample characterization to detailed
investigations of the origins of ZFS and the establishment of magnetostructural
correlations for the ZFS. All these studies were carried out conventionally, in the
sense that the external field was swept at constant radiation frequency, while
monitoring the change in radiation intensity. There are a number of alternative
measurement methods that can be broadly divided into those where the measure-
ment is performed at fixed field and the frequency dependence of the transmitted
or reflected radiation intensity is obtained directly, and, second, those where the
magnetic resonance transitions are detected by a change in the thermodynamic
properties of the sample, such as the magnetization or the magnetic torque. These
methods are reviewed in Sect. 2. The third section of this review summarizes the
pulse EPR studies that have been performed on MNMs. This is an exciting new area
in molecular magnetism, where studies on MNMs can contribute to the understand-
ing of quantum coherence and decoherence and their dependence on molecular and
electronic structure and complexity.
The zero-field splittings of the spin ground states of a number of MNMs were
determined by means of FDMR (Table 1). In most of these studies, the data were
analyzed in the giant spin Hamiltonian framework, i.e., considering the ground spin
206
Table 1 Zero-field splittings of molecular nanomagnets determined by frequency domain magnetic resonance spectroscopy
Compound S D/cm1 E/cm1 105B40/cm1 105B44/cm1 Ref.
[Mn12O12(OAc)16(H2O)4] 10 0.461 0 2.2 0 [66]
[Mn12O12(O2CCMe3)16(H2O)4] 10 0.464 0 1.9 0 [67]
[Mn12O12(O2CPh)16(H2O)4] 10 0.47 0 2 0 [68]
(PPh4)[Mn12O12(O2CEt)16(H2O)4] 19/2 0.454 0 1 0 [69]
[(tacn)6Fe8O2(OH)12Br7(H2O)]Br(H2O)8 10 0.203 0.00307 0.004 0 [70]
[Mn9O7(O2CCH3)11(thme)(py)3(H2O)2] 17/2 0.247 0 0.46 0 [71]
[Ni4(MeOH)4(sea)4]a 4 0.93 0.023 43 210 [72]
[Mn6O2(sao)6(O2CMe)2(EtOH)4] 4 2.12 0 15 0 [26]
[Mn6O2(Et-sao)6(O2CPh)2(EtOH)4] 12 0.386 0 0.4 0 [26]
[Mn6O2(Et-sao)6(O2CPhMe2)2(EtOH)4] 12 0.362 0 0.6 0 [26]
[Ni(hmp)(MeOH)Cl]4 4 0.592b, 0.771 0 0 0 [73]
[Ni(hmp)(dmb)Cl]4 4 0.621, 0.631 0 0 0 [73]
[Ni(hmp)(dmp)Cl]4 4 0.610 0 0 0 [73]
The parameters D and B40 are related to the parameters D2 and D4 of the spin Hamiltonian H D2Sz2 + D4Sz4 that is also often used to describe the ZFS,
30
according to B04 D4 =35 and D D2 25 35 D4 35 SS 1D4 [74]
a
Including inelastic-neutron-scattering data
b
Resonance lines due to two different species were observed
J. van Slageren
New Directions in Electron Paramagnetic Resonance Spectroscopy 207
multiplet only. The first (1998) such study was performed on pressed powder pellet
samples of Mn12Ac [66]. The ZFS spin Hamiltonian parameter values of D2
0.389 cm1 and D4 7.65 104 cm1 (Table 1) were obtained. These para-
meters corresponded well to the parameters obtained by high-frequency EPR.
Furthermore, FDMR measurements showed the axial ZFSs of other Mn12 deri-
vatives, such as [Mn12O12(O2CCMe3)16(H2O)4] (Mn12Piv, Piv pivalate) [67] and
[Mn12O12(O2CPh)16(H2O)4] (Mn12Bz, Bz benzoate) [68], to be quite similar
(Table 1). One-electron reduction of Mn12 causes reduction of one of the aniso-
tropic MnIII ions to give a more isotropic MnII ion, with a concurrent decrease in
ZFS, evidenced by FDMR measurements which gave a slightly decreased D-value
of D 0.454 cm1, and, surprisingly, a change in sign in the case of B40 (Table 1).
The measurement also confirmed the ground state spin of S 19/2 [69].
Another early FDMR investigation further validated the usefulness of the tech-
nique. In this study, four resonance lines were observed for [(tacn)6Fe8O2(OH)12
Br7(H2O)]Br(H2O)8 (Fe8) [70]. Here the often excellent agreement between reso-
nance frequencies determined by FDMR and those determined by INS was also
noted for the first time. An attempt was made to extract the second order transverse
ZFS parameter from the data, and a value of |E| 0.0307 cm1 was obtained,
which is lower than that obtained from the INS data. It was noted that the simu-
lations are not very sensitive to the E-value, and not at all to the sign of E. FDMR
measurements on MNMs with large spins and large negative D-values are generally
not very sensitive to transverse ZFS, because transverse ZFS mostly affects levels
with small MS values, i.e., those close to the top of the energy barrier, and transitions
between these are often not observed in FDMR (see Sect. 2.4). The good correspon-
dence between INS and FDMR was also noted in the case of [Mn9O7(O2CCH3)11
(thme)(py)3(H2O)2]MeCNEt2O (H3thme 1,1,1-tris(hydroxymethyl)-ethane) [71].
Magnetization measurements demonstrated that the metallocubane clusters
[Ni4(ROH)4(sea)4] (H2sea salicylidene-2-ethanolamine; R Me or Et) with
approximate S4 symmetry have S 4 ground states and large negative D values,
and yet the expected slow relaxation of the magnetization was not observed, even
down to 40 mK [72]. A combined INS and FDMR study proved that this is due to
large transverse ZFS both in second and fourth order, which causes mixing of the
MS-states and hence efficient quantum tunneling of the magnetization. The origin of
the ZFS in S4 symmetry clusters was the subject of a theoretical study, which
proved that the first excited total-spin states belong to a degenerate representation
of the S4 point group, which corresponds to a quasi-orbital angular momentum [75].
First order spinorbit coupling then leads to an effective antisymmetric exchange
term in the spin Hamiltonian. An attempt was made to fit the INS spectra of
[Ni4(MeOH)4(sea)4] using this approach, although the nonzero E-value shows
that the effective cluster symmetry is lower than S4. In EPR studies on other Ni4
clusters, no antisymmetric exchange interactions were found [8].
Finally, a recent study reported FDMR and INS spectra of a series of hexa-
nuclear manganese(III) clusters, with both S 4 and S 12 ground states (Fig. 2)
[26]. Magnetically, these clusters consist of two exchange coupled triangles,
and ground state spin value depends on the nature of the exchange interaction
208 J. van Slageren
Fig. 2 FDMR-spectra
recorded on a powder
pellet of [Mn6O2(Et-sao)6
(O2CPhMe2)2(EtOH)4]
4EtOH at different
temperatures as indicated.
Adapted from [26]. See
also Fig. 1
2.1.4 Lineshapes
It was recognized early on that the lineshape formula that best describes the FDMR
lines in spectra recorded on MNMs is usually the Gaussian lineshape rather than the
Lorentzian [70]. This suggests that the linewidth is dominated by (Gaussian)
distributions in the spin Hamiltonian parameters, especially in D. For Mn12Ac
evidence for a discrete number of isomers rather than a continuous distribution
was found [78] in agreement with high-frequency EPR results [7, 79]. In two cases,
(PPh4)[Mn12O12(O2CEt)16(H2O)4] [69] and [Ni4(MeOH)4(sea)4] [72], the reso-
nance lines were observed to be better described by Lorentzian functions. This
suggests, that the linewidth may be determined by spin relaxation, in particular T2
relaxation. This was supported by the fact that the linewidth increased towards
higher temperatures. Assuming that the linewidth is determined by T2 relaxation, an
estimate of T2 5058 ps was obtained, which is much shorter than TM times that
were later obtained for other MNMs by pulsed EPR on dilute frozen solutions
(Sect. 3.1). Temperature-dependent Gaussian linewidths were also found in high-
frequency EPR measurements on Fe8, which were attributed to the change in
thermal fluctuations of the intermolecular magneto-dipolar interaction [80, 81].
The Lorentzian linewidth for [Ni4(ROH)4(sea)4] is largely temperature-independent,
suggesting that the spin relaxation is temperature-independent in the investigated
temperature range, or that the lineshape has a different origin. A priori, the
lineshape of an FDMR line is not necessarily Lorentzian or Gaussian. However,
when implementing a MachZehnder interferometer, one can also measure the
frequency dependence of the phase shift due to the sample (i.e., the dispersion,
rather than the absorption). This allows direct determination of both real and
imaginary parts of the magnetic permeability or dynamic magnetic susceptibility,
without assumptions regarding the lineshape [62]. In a magnetized medium, the
material parameters (dielectric permittivity and magnetic permeability) become
tensorial in nature [82] and this can have an influence on the propagation of
radiation in such media, leading to magneto-optical effects. FDMR measurements
involve magnetic dipole transitions and, thus, only the magnetic permeability tensor
needs to be taken into account. When the radiation propagation is perpendicular to
the magnetization direction (Voigt geometry), the magneto-optical effects are
rather small and were reported to lead to slight shifts in the apparent resonance
frequency, as well as causing an asymmetry of the resonance lines, for measure-
ments on magnetized single-crystal mosaics of Mn12Ac [78]. In HFEPR measure-
ments in finite fields, asymmetric lines were also observed, due to distributions
in the angle between molecular ZFS tensor and external field [83]. If the radiation
propagation is parallel to the direction of the magnetization (Faraday geometry), the
effects are much more pronounced. In this case, a strong rotation (>150 /mm
sample thickness) of the plane of polarization of linearly polarized radiation
was observed off-resonance (the familiar Faraday effect), again on magnetized
Mn12Ac single-crystal mosaics (Fig. 3) [84]. At resonance, one circular component
of the linearly polarized radiation is virtually fully absorbed, and the transmitted
210 J. van Slageren
Because field and frequency can be changed at will, more elaborate measurements
can be carried out. The relaxation of the magnetization in Mn12Ac was investigated
in some detail, using the measurement procedure of cooling in a certain field,
followed by inversion of the external field to a certain value in the opposite
direction and subsequent recording of spectra at various time delays. FDMR
measurements on powder samples after inverting the field from +0.45 to 0.45 T
showed the gradual disappearance of one resonance line at 9.57 cm1, assigned to
a transition from the metastable MS +10 state, and the appearance of a resonance
line due to the transition from the 10 ground state. At the higher field of 0.9 T, the
relaxation no longer proceeds homogeneously, but a pronounced hole develops
in the resonance line, concurrently with the development of a peak at higher
frequencies. These observations were attributed to quantum tunneling of the mag-
netization [86]. The relaxation of the magnetization was also performed on single-
crystal mosaic samples of Mn12Ac [87]. The populations of the MS 10 states
New Directions in Electron Paramagnetic Resonance Spectroscopy 211
were extracted from the intensities of the resonance lines, and from the time-
dependence of these intensities the relaxation time was determined. These measure-
ments were repeated at many different longitudinal fields. The field dependence of
the relaxation time shows pronounced minima which were assigned to quantum
tunneling of the magnetization occurring at those longitudinal fields where
MS levels on the left and right of the potential energy barrier are accidentally
degenerate. The shift of the apparent resonance frequency during the relaxation
process and dependence of this shift on the longitudinal field were discussed in
terms of internal dipolar fields and quantum tunneling effects.
Most FDMR investigations of MNMs (Sect. 2.1) were performed in the THz-Far
Infrared (FIR) region of 90600 GHz. This frequency range is also accessible by
Fourier-transform infrared spectrometers, based on MartinPuplett or Michelson
interferometers. The advantage of this measurement method is that the entire
spectrum can be acquired with a single radiation source, in contrast to current
implementations of FDMR, which require about a dozen BWOs to cover the entire
frequency range from 30 to 1,200 GHz. In addition, higher frequencies are easily
accessed. Below ca. 300 GHz, the intensity of conventional far-infrared sources
drops drastically, which decreases sensitivity. With the use of coherent synchrotron
radiation, instead of a conventional mercury lamp, this disadvantage can be over-
come [88]. To enhance sensitivity, lock-in detection was employed, locked to the
pulse repetition rate of the synchrotron source (1.25 MHz). The FT THz method
was successfully used to measure the ZFS in Mn12Ac [88]. The spectra were
presented as the frequency dependence of detector signal at low-temperatures
divided by that at high temperatures, to eliminate the frequency dependence of
the source intensity, as well as absorption lines due to electric dipole transitions in
air. Note that the physical principles of FT THz and FDMR spectroscopies are
equal. FT-THz measurements on the SMM (NEt4)[Mn2III(5-Brsalen)2(MeOH)2
RuIII(CN)6] (5-BrsalenN,N0 -ethylenebis(5-bromosalicylidene) aminato anion)
showed features at 9.0, 12.8, 18.6, and 19.3 cm1, of which only the first two
were attributed to magnetic resonance transitions, because they split on application
of a magnetic field [89]. The combination of these magnetic resonance measure-
ments and susceptibility data allowed establishing of a consistent energy level
diagram. The analogous chromium complex, (NEt4)[Mn2III(5-Brsalen)2(MeOH)2
CrIII(CN)6] was also investigated by FT-THz spectroscopy, as well as by INS, and
gave similar results [90].
In addition to the studies cited above, Fourier Transform IR-spectrometers have
been used for the study of magnetic excitations since the 1960s. Thus, collective
excitations were observed in, among others, FeF2 [91, 92], FeIIIdithiocarbamate
212 J. van Slageren
[93] and LaMnO3 [94]. Crystal field splittings were observed for lanthanide ions
[95]. The ZFS was studied in mononuclear complexes of FeIII [96] FeII [97], and
square planar CoIII [98, 99] among others. Even hemoglobin and myoglobin were
successfully investigated in this manner [93, 100].
The following is the personal view of the author, based on his experience with the
discussed techniques. The advantage of frequency-domain techniques, described in
Sects. 2.12.3 is that no external field is required and ZFSs are obtained directly.
The ability to apply a small field is useful, especially at frequencies above ca.
20 cm1, in order to distinguish magnetic resonance transitions from molecular
vibrations and other phonon-type excitations. The advantage of field-swept tech-
niques is that field-swept spectra tend to have flatter baselines, which increases
sensitivity. Cavity and other resonator methods, which are only easily implemented
in field-swept experiments, are much more sensitive. Therefore, single-crystal
New Directions in Electron Paramagnetic Resonance Spectroscopy 213
measurements are more easily carried out, which give access to a great deal of detail
that is not easily available from polycrystalline samples.
The linewidths of the resonance lines in FDMR and FT-THz measurements
(Fig. 4) are limited by the sample, although the intrinsic resolution of FDMR is
higher. FT-THz has the advantage that a single period of the interferometer covers
the entire frequency range available (typically up to 40100 cm1, depending on
source and detector), with the possibility of accumulating a number of scans (using
a mercury lamp source typically dozens of scans are acquired). FDMR, on the other
hand, is better suited to measurements below 10 cm1, where the optics of most IR
spectrometers starts to cut off some of the intensity. THz TDS suffers from much
lower resolution, which is limited by the length along which the delay stage can be
moved. Its advantage lies in the ability to record both amplitude and phase and
the inherent fast time resolution available, although both factors have yet to be
exploited in molecular magnetism.
INS also has a lower resolution and also suffers from poor sensitivity. Its
clear advantages (that should not be underestimated) are that transitions between
spin multiplets are allowed and that both the frequency- and wavevector (Q)-
dependence of the scattering intensity are recorded. The Q-dependence allows, in
principle, the distinguishing of magnetic from nonmagnetic excitations and also
gives information on the spin wavefunction [60]. Finally, because the scattering
intensity only depends on the population of the initial state, INS is very useful at
high temperatures, where the EPR intensities are low, because of limited population
differences between initial and final states.
Not all molecular systems are amenable to EPR in the field-, frequency-, or
time-domain. For example, fast spin relaxation can lead to excessively broad lines.
Furthermore, the sensitivity of nonresonator techniques, including all non-field-swept
measurement methods, is generally insufficient to allow for detailed single crystal
measurements. Finally, during the wide-range field sweep of a field-swept, conven-
tional (high-frequency) EPR measurement, the actual physical properties, including
the spin ground state may change, e.g., in the case of field-induced spin ground state
changes in antiferromagnetic (ring) systems. For these reasons, effort has been
invested into the development of other detection schemes, based on the monitoring
of a thermodynamic quantity of the system, such as the magnetization or the magnetic
torque. Many of the instruments described below were devised to investigate the spin
dynamics of MNMs, for which purpose pulse EPR spectrometers are also currently
used (Sect. 3).
2.5.1 Implementations
Fig. 5 Magnetization curves measured on Fe8 in the presence of microwave radiation with
n 118 GHz and T 10 K. The dips are due to transitions between the MS states of the
S 10 ground state, as indicated. Adapted from [115]. Used with permission. #2005 American
Physical Society. Readers may view, browse, and/or download material for temporary copying
purposes only, provided these uses are for noncommercial personal purposes. Except as provided
by law, this material may not be further reproduced, distributed, transmitted, modified, adapted,
performed, displayed, published, or sold in whole or part, without prior written permission from
the American Physical Society
than previously reported values g||c 1.98 and gc 1.95 [110] or g||c 1.98 and
gc 1.968 [111] (both X-band EPR), g 1.97 (57.8 GHz) [112], g||c 1.981,
and gc 1.953 (FDMR) [113]. The same setup was used to detect EPR in
(H2NMe2)[Cr7NiF8(O2CCMe3)16] (Cr7Ni) [114] In later measurements, on Fe8,
a Hall bar was employed rather than a micro-SQUID (Fig. 5) [115].
A second instrument, developed by Friedman et al., employed a Hall bar sensor
to measure the magnetization of the sample, which has the advantage over micro-
SQUIDs that the operating magnetic field can be much higher [116, 117]. The
authors used this and later setups to investigate magnetic resonance and spin
dynamics of Fe8 in a series of papers. High-power BWOs and a rectangular cavity
were used to enhance the microwave field at the sample, and the Hall bar sensor was
mounted outside the cavity [117]. Heating of the sample to 4 K, at a nominal
cryostat temperature of 2 K by nonresonant heating of the resonator by the radia-
tion was observed, with additional heating observed at resonance. The D-value of
0.2944 K, obtained from measurements at different frequencies between 127 and
136 GHz, is in good agreement with published values. Further measurements were
performed on [Mn4O3Cl(O2CCH3)3(dbm)3] (Hdbm dibenzoylmethane) [118].
An improved setup employed a cylindrical rather than a rectangular cavity and
a thin film pick-up loop rather than a Hall bar sensor for detection [119]. The pick-
up coil makes the instrument well suited for the investigation of fast spin dynamics
(see below) but less for spectroscopic purposes. The emf induced in the loop is
measured with a SQUID voltmeter coupled to a room temperature amplifier.
216 J. van Slageren
A Hall bar sensor, but no resonator, was also used by del Barco, Kent et al. at
temperatures down to 0.38 K with the sample in vacuum [120]. Measuring
the longitudinal field dependence of the magnetization of [Ni(hmp)(dmb)Cl]4
(Ni4, Hhmp 2-hydroxymethylpyridine, dmb 3,3-dimethyl-1-butanol) with
and without radiation (at single frequencies between 24 and 43 GHz) allowed the
observation of magnetic resonance transitions, with extracted ZFS values in good
agreement with published values. In transverse fields, the plots of resonance fields vs
frequencies show a clear curvature near zero-field, which is a signature of the ground
state being a quantum superposition of states with opposite magnetization, as is also
required for quantum tunneling of the magnetization. This setup was later improved,
when a strip line resonator was implemented on the chip with the Hall sensor, where
different resonators allow for measurements between 10 and 30 GHz, and both the
transmitted radiation and the magnetization can be measured [121, 122].
A commercial SQUID magnetometer was used for magnetization detection in
combination with Klystron sources at 95 and 141 GHz, where the advantage lies in
the ability to determine the absolute magnetization quantitatively [123]. Investiga-
tion of the power dependence allowed extraction of T1 in copper sulfate. Direct
time-resolved measurements appear to be difficult to implement in view of the slow
measurement time scale of commercial SQUID magnetometers. Again, the method
was used to investigate Fe8, and clear magnetic resonances were observed [124].
The magnetic resonance transitions can also be observed by measuring the
magnetic torque in magnetically anisotropic systems [125]. The setup employed
BWO radiation sources in combination with an optical split coil superconducting
magnet, allowing excitation frequencies between 30 GHz and 1 THz, and fields up
to 7 T. A sample rotator is incorporated which allows measurements at arbitrary
angles with the magnetic field. The use of an optical cryostat enables arbitrary
polarization of the radiation with the external field. Preliminary measurements on
a star-shaped Fe4 single-molecule magnet in both field and frequency domains were
presented. The combination of EPR and torque magnetometry was also used by
Ohmichi et al., but has not yet been used for investigations of MNMs [126, 127].
broadened lines, the half-width in CW-EPR is related to the real spinspin relaxa-
tion time T2, of which TM is a lower limit, by Dn (2pT2)1, assuming T1 >> T2
[133]. Although EPR lines of MNMs are generally inhomogeneously broadened,
a lower limit of TM can still be obtained from the linewidth. This approach yielded
values in the range of 1010 to 109 s, including 0.2 ns [121], 0.5 ns [120], 2 ns [107],
and 10 ns [114]. Pulse EPR measurements (Sect. 3) revealed that phase memory
times can fortunately be much longer than these estimates.
Much effort has been devoted to the study of the dynamics of the magnetization in
single molecule magnets, where the axial magnetic anisotropy creates an energy
barrier towards inversion of the magnetic moment [29]. The relaxation time is
generally governed by an Arrhenius law t t0exp(DE/kBT). Microscopically,
relaxation of the magnetization is a multistep process, in which the molecule
undergoes multiple phonon-induced transitions to increasingly higher lying
MS-states, until the highest-lying MS state is reached. This is then followed by
multiple downward transitions on the other side of the energy barrier. For direct
spin-phonon interactions, the oscillating strain caused by a phonon causes modula-
tion of the ZFS, which can induce transitions between MS states with the selection
rule DMS 1, 2 [29, 74]. Because the phonon density decreases with decreas-
ing energy, the direct spin-phonon interaction process becomes slowest close to the
top of the energy barrier, where the spacing between MS levels is smallest. The rate
of the slowest spin-phonon transition is reflected by the Arrhenius prefactor t0. In
EPR terms, these individual phonon-induced transitions are nothing but spinlattice
relaxation. In addition to the direct process mentioned above, in ESR-spectroscopy,
two two-phonon processes, the Orbach and the Raman process, are generally
considered, where in the former the intermediate state is real while in the latter it
is virtual [74]. Although there is ample information on the spinlattice relaxation in
single-spin systems [74, 134, 135], much less is available on exchange coupled
systems [28]. One important finding is that spin relaxation in polynuclear systems is
generally faster, because of the presence of a number of additional relaxation
pathways. Thus modulation of the exchange interaction can lead to Orbach-type
spinlattice relaxation, with an excited spin state as the intermediate state. For
dimers only the nonisotropic exchange interactions are effective in this way, but
in larger systems, modulation of the isotropic exchange interaction can lead to
relaxation [28, 136].
The spinspin relaxation T2, or the experimentally determined phase memory
time TM, is a parameter of interest in molecular magnetism, because it is the time
available for a quantum computation. It was recognized several years ago that
MNMs may be used to implement quantum bits [137, 138]. There are currently
essentially three proposals for using MNMs for quantum information processing
[21, 139, 140]. The first is an elaborate scheme to use high-spin MNMs, such as
New Directions in Electron Paramagnetic Resonance Spectroscopy 219
SMMs, where it is important that the MS states are not equidistant. This is the
case for a ZFS split ground state [19]. A combination of multifrequency pulses at
high frequencies where the radiation magnetic field is perpendicular to the ZFS
quantization axis, and a low-frequency pulse where the field is parallel to this axis,
can create a simultaneous coherent population of many MS states. In addition, the
field strengths and phases of the high-frequency components must be judiciously
controlled. If those conditions are met, Grovers algorithm, which, in certain cases,
allows finding an entry in a database with one single query, can be implemented.
For this scheme, entanglement of many-particle systems is not necessary [19].
A second proposal deals with more conventional two-level systems for use as
qubits [139]. The advantages of MNMs for such purposes are generally listed as
follows. The larger physical size of MNMs compared to single sites makes resolu-
tion requirements for local addressing less strict. Second, the excited spin states
can, in carefully engineered systems, be used to implement effective switching of
the interaction between two qubits [141]. Engineering the interaction between pairs
of MNMs is currently receiving a great deal of attention [50, 142] but is far from
trivial. The interaction should be much weaker than the interactions within the
MNM qubit so that the two can be considered separate entities, but the coupling
energy should be much larger than the thermal energy [139]. The third proposal
concerns the electric control of the exchange interaction in triangular MNMs, where
an electric field can couple the two S 1/2 states, that can show a small energy
splitting. These two states are characterized by their chirality [140, 143]. The
advantage of electric field control is that electric fields can be applied much more
locally. In all of these schemes, the quantum operations need to be performed
within the lifetime of the generated coherent superposition states, i.e., within the
quantum coherence time, for which the experimentally determined phase memory
time TM is a lower limit.
The spin dynamics are also critically important in biophysical EPR. This field
revolves around the elucidation of geometric and electronic structures and dynam-
ics of biological systems, by the determination of g-tensors, of distances between
electron spins and of couplings between nuclear and electron spins. For many of
the measurement techniques employed, sufficiently long TM times are required for
these techniques to be fruitfully applied.
It is clear from the above that the spin dynamics of MNMs is a subject of great
current interest. The techniques to implement quantum operations and those of
pulse EPR are essentially the same. In both cases, pulses of arbitrary length and
phase, that are resonant with the energy splitting of the relevant states of the inves-
tigated system, are required. For a two-state system, the two states can be pictured
as opposite poles on a sphere (the Bloch sphere), where all possible coherent
superposition states are points on that sphere. Applying a resonant microwave
pulse causes a rotation of the spin in a direction perpendicular to that of the
microwave magnetic field [144]. The turning angle is given by the microwave
field strength (B1) and the pulse duration (tp) according to y o1tp, with o1
gmBB1/ h. For a given turning angle, short, intense pulses excite spins in a larger
frequency range than do long, less intense pulses, because of the Heisenberg
220 J. van Slageren
uncertainty relation, where the FWHM is given by Dn 1.207 tp1 for a rectan-
gular pulse [144]. Applying series of pulses of judiciously chosen lengths and vary-
ing the interpulse delays allow determination of the spin relaxation time constants
of MNMs, as detailed in the next two sections. The work by Takui et al. on pulse
EPR spectroscopy on high-spin organic radicals is not reviewed here [145, 146].
Some of the earliest spinlattice relaxation time measurements were carried out
using superheterodyne detection by partial saturation of the magnetic resonance by
a microwave pulse, and subsequent monitoring of the transient recovery on an
oscilloscope [147], which is essentially a type of saturation recovery measurement
[144]. In this manner, the temperature dependence of T1 in iridium(IV) pairs in
(NH4)2PtCl6 and copper(II) pairs in Zn(dtc)2 [dtc bis(diethy1-dithiocarbamate)]
was investigated in detail. These early measurements have been reviewed [28, 148]
and are not discussed further here. Another version of the saturation-recovery
experiment involves a long saturation pulse or sequence of pulses, followed after
variable time delay T by the primary echo pulse sequence p/2tfixptfixecho to
detect the longitudinal magnetization. A third method to measure T1 is by inversion
recovery, using the pulse sequence pTp/2tfixptfixecho. This last method is
more sensitive to spectral and spin diffusion than the former [144]. Finally,
stimulated echo decay can also be used to measure T1, by variation of T in the
pulse sequence p/2tfixp/2Tp/2tfixecho. This method is even more sensitive
to spin and spectral diffusion. In the following, we discuss spinlattice relaxation
times measured on MNMs and related systems (Table 2). We restrict ourselves to
pulse EPR measurements, omitting measurements of the electron spin dynamics by
NMR [134], or a.c. magnetic susceptibility [164].
The antiferromagnetic ring (H2NMe2)[Cr7NiF8(O2CCMe3)16] (Cr7Ni) has a spin
ground state of S 1/2, and was the first MNM to be investigated by pulse EPR
[18]. The spinlattice relaxation time was investigated by means of inversion
recovery measurements on frozen solutions in toluene at X-band. T1 reaches
a value of about 1 ms at the lowest temperature of 1.8 K (Table 2), but decreases
rapidly with increasing temperatures, down to 102 ns at 9.5 K, suggesting inter-
action with phonons, i.e., Raman or Orbach processes. Measurements on (H2NMe2)
[Cr7MnF8(O2CCMe3)16] (Cr7Mn), which has an S 1 ground state, gave similar
T1 values, where the authors note that the T1 values show a variation of a factor of
two at different points in the spectrum, indicating that the ZFS plays a role in the
relaxation. W-Band (95 GHz) measurements on the same compound revealed that
the spin-relaxation time is little dependent on microwave frequency, with T1 2.9
ms (W-band, 4.5 K) [149] compared to T1 3.5 ms (X-band, 4.5 K) [18].
The spinlattice relaxation was studied in more detail in a ferric triangle,
[Fe3(m3-O)(O2CPh)5(salox)(EtOH)(H2O)] (Fe3salox, H2salox salicylaldehyde
oxime) [150]. Inversion recovery measurements at X-band frequencies yielded
Table 2 Spin relaxation data of exchange-coupled metal ion species, determined by pulse EPR
Species Abbrev. S T1 / ms TM / ms T/K Ref
(H2NMe2)[Cr7NiF8(O2CCMe3)16] Cr7Ni 1/2 103 0.55 1.8 [18]
(H2NMe2)[Cr7NiF8(O2CCMe3)16] (X-band) 1/2 ~3.5 0.379 4.5 [18]
(H2NMe2)[Cr7NiF8(O2CCMe3)16] (W-band) 1/2 ~2.0 0.357 5.0 [149]
(H2NMe2)[Cr7NiF8(O2CCMe3-d9)16] 1/2 3.8 1.8 [18]
(H2NMe2)[Cr7NiF8(O2CCMe3-d9)16] 1/2 2.21 4.5 [18]
(H2NMe2)[Cr7MnF8(O2CCMe3)16] 1 103 ~0.55 1.8 [18]
[Fe3(m3-O)(O2CPh)5(salox)(EtOH)(H2O)] Fe3 1/2 693 2.6 5.5 [150]
K6[V15As6O42(H2O)]8H2O V15 3/2 0.340 4 [151]
[Fe4(acac)6(Br-mp)2] Fe4 5 1.06 0.63 4.3 [152]
[(tacn)6Fe8O2(OH)12Br7(H2O)]Br(H2O)8 Fe8 10 0.093 1.93 [153]
[(tacn)6Fe8O2(OH)12Br7(H2O)]Br(H2O)8 10 9.5 102 0.714 1.27 [153]
III IV
[Mn Mn (m-O)2bipy4]ClO4 1/2 1,380 25 [154]
[MnIIIMnIV(m-O)2bipy4]ClO4 1/2 11 75 [154]
[MnIIMnIII(m-OH)(m-piv)2(Me3tacn)2](ClO4)2 1/2 35 4.3 [154]
[MnIIMnIII(m-OH)(m-piv)2(Me3tacn)2](ClO4)2 1/2 0.42 9 [154]
[Fe2S2(S2-o-xylyl)2]3 1/2 ~55 103 5 [155]
{Mn4OxCa} Photosystem II S0 state 1/2 9.4 4.3 [154]
{Mn4OxCa} Photosystem II S0 state 1/2 0.85 6.5 [154]
1/2 1,220 4.2 [156]
New Directions in Electron Paramagnetic Resonance Spectroscopy
T1 times between 693 ms (5.5 K) and 0.55 ms (11.0 K). The temperature dependence
of T1 could be fitted equally well to the equations describing Orbach or Raman
processes. The unrealistic obtained exponent value in case of the Raman process led
the authors to prefer the Orbach process. The Orbach energy gap was derived to be
57 cm1. This energy gap corresponds to the energy of the first excited spin state,
which allowed choosing between two sets of exchange coupling parameters that
fitted the susceptibility curve equally well.
A similar strong temperature dependence of T1 was found in inversion recovery
measurements on frozen toluene solutions of [Fe4(acac)6(Br-mp)2] (Fe4, acac
acetyl acetonate, or 1,3-pentanedioate, Br-mpH3 2-(bromomethyl)-2-(hydroxy-
methyl)-1,3-propanediol) [152] which has an S 5 ground state [165], with
T1 1.06 ms at the lowest temperature (T 4.3 K) employed. The strong temper-
ature dependence suggests that the spinlattice relaxation proceed via a two-phonon
process. Although the limited temperature range available did not allow the draw-
ing of definite conclusions, an Orbach fit gave a value of about 5 cm1 for the
Orbach energy gap, which is well within the ground spin multiplet.
Investigations of the spinlattice relaxation time of single crystals of Fe8 at
T 1.27 K, utilizing the stimulated echo sequence, revealed two relaxation
processes with time constants of 1.0 ms and 0.95 ms, respectively [153]. The authors
attribute the former time constant to spectral diffusion, and speculate that the
latter may reflect true spinlattice relaxation, although the time constant is two
orders of magnitude longer than values found by magnetization-detected EPR
(see above) [129, 131].
The results described above demonstrate that there is a large variation in T1 times
for different MNMs, as well as a strong temperature dependence of T1 (Table 2). An
extreme example of this variation in T1 is given by the two S 1/2 dimers
[MnIIIMnIV(m-O)2bipy4]ClO4 (bipy 2,20 -bipyridine) and [MnIIMnIII(m-OH)(m-
piv)2(Me3tacn)2](ClO4)2 (Me3tacn N,N0 ,N00 -trimethyl-1,4,7-triazacyclononane)
[154]. The spinlattice relaxation of the former is very slow and characterized by
a Raman process below 50 K, while that of the latter is best described by an Orbach
process. Apart from the metal oxidation states involved, the main difference
between these complexes is the strength of the exchange interaction which leads
to energy gaps between ground and first excited state of DE 450 cm1 and
DE 25.5 cm1, respectively, for the two complexes. The Orbach energy gap
(D ~ 26 K) in the latter complex was indeed found to match the energy gap to the
first excited state very well. Similar results (Table 2) are obtained for polynuclear
clusters in biological systems, e.g., the {Mn4OxCa} cluster in Photosystem II
[154, 156, 157] and ironsulfur clusters in different oxidation states [158161].
Phase memory times in solids are usually determined by the Hahn or primary echo
sequence, p/2tptecho, by variation of t [144]. For quantum computation
New Directions in Electron Paramagnetic Resonance Spectroscopy 223
purposes, the quantum coherence time must be sufficient to allow extensive quan-
tum manipulations, including error corrections. The quantum coherence time is the
same as the spinspin relaxation time T2, for which the experimentally determined
phase memory time TM is a lower limit.
For a long time it was unclear whether quantum coherence times in MNMs
would be long enough to make MNMs viable qubit candidates. Indeed, this was the
very question that the title of the first paper in this area asked [18]. The phase
memory time was determined to be TM 379 ns by X-band Hahn echo
measurements on frozen toluene solutions of Cr7Ni, which is two orders of magni-
tude longer than previous estimates of the lower limit of TM [18]. Interestingly, the
related compound (H2NMe2)[Cr7MnF8(O2CCMe3)16], which has an S 1 ground
state, has a very similar TM, which demonstrates that the ZFS or ground state spin
plays a limited role in the decoherence process. The echo decay for Cr7Ni displays a
pronounced modulation when using short, intense microwave pulses. This effect is
called electron spin echo envelope modulation (ESEEM), and is caused by formally
forbidden nuclear spin flips by the p-pulse of the Hahn echo sequence [144]. The
oscillation frequency, therefore, corresponds to the Larmor frequency of the nuclei
that the electron spin couples to via superhyperfine coupling. The ESEEM in Cr7Ni
was found to be due to coupling to the ligand protons. No coupling to the bridging
19
F nuclei was reported. It is this coupling to proton nuclear spins that is the main
decoherence pathway. By deuteration of the ligand it proved possible to enhance
TM by a factor of 6 (TM 2.21 ms at T 4.5 K), in agreement with the smaller
nuclear magnetic moment of D, compared to H. The phase memory time was found
to increase with decreasing temperature (Table 2), but not as strongly as T1 (see
above). A very similar TM of 379 ns was found in W-band measurements on Cr7Ni
[149]. W-band pulse ENDOR measurements on Cr7Ni demonstrated that the
coupling between electron spin and proton nuclear spin is dipolar in nature, and
its strength is up to ~2 MHz [149].
In X-band Hahn echo measurements on frozen solutions of Fe3 in acetone, clear
spin echoes were observed, from which TM 2.18 ms at 7 K was extracted, which
is clearly longer than for Cr7Ni under similar conditions. The observation of
ESEEM again demonstrated hyperfine coupling of the electron spin to nuclear
spins. Interestingly, TM becomes temperature independent below ca. 7 K, at
which point TM reflects the true spinspin relaxation rate, whereas at higher temper-
atures it is influenced by spinlattice relaxation. The magnetization can be rotated
by a microwave pulse with length tp around an arbitrary angle y o1tp, where o1
is the microwave field strength. After a delay time, which ensures the decay of all
quantum coherences, the magnetization along the z-axis is measured. The measure-
ment of the z-magnetization as a function of tp is called a nutation measure-
ment. The corresponding oscillations of the magnetization are the so-called Rabi
oscillations. A nutation measurement performed on Fe3 showed that part of the
observed oscillations was due to nutation of the electron spin, but another part
was attributed to ESEEM-like effects (Fig. 6). The oscillations due to transient
nutation were observed to decay very quickly, presumably due to microwave
field inhomogeneities.
224 J. van Slageren
deuterium, proving that at this field the main interaction is with the nuclear spins of
the solvent. In echo detected ESR spectra, ESEEM-like oscillations were observed
at low field, which demonstrated that hyperfine couplings to the nuclear spins of the
cluster are also present. Measurements in CS2, which is largely nuclear-spin-free,
show an increased phase memory time of Fe4 of TM 527 ns at T 4.3 K.
The transient nutation measurement showed pronounced Rabi oscillations. These
oscillations cannot be due to ESEEM-like effects, due to coupling to proton nuclei,
because the measurements were performed in zero external field, where the proton
nuclear Larmor frequency is virtually zero. On the other hand, quadrupole nuclei
(I > 1/2) may exhibit nuclear ESEEM at zero field [167]. However, the only
quadrupole nuclei in Fe4 are the two bromine atoms of the Br-mp-ligands. Strong
coupling of the electron spin to these bromine atoms is not expected because they
are far away from the spin carrying iron ions.
Measurements of the quantum coherence are usually performed in dilute
systems to prevent decoherence due to fluctuating intermolecular magnetic-dipolar
electronelectron interactions. In SMMs these fluctuations can also be frozen out at
low temperatures, below the blocking temperature of the magnetization. A single-
crystal study on Fe8 made use of this fact and, indeed, phase memory times of up to
712 ns were observed at 1.27 K [153]. Raising the temperature to 1.93 K results
in a drastic reduction of TM to 93 ns. Simulations showed that electron spin
electron spin interactions can account quantitatively for this behavior. A second
decoherence process was identified from these simulations, with a decoherence
time of about 1 ms, which was attributed to hyperfine-induced decoherence.
Spinspin relaxation was also studied for polynuclear clusters in biomolecules,
especially for Fe4S4 clusters. Interestingly a large range of values was found for
chemically similar species (Table 2) [158160].
From the previous subsections it is clear that important first results from pulse EPR
investigation of MNMs have been obtained. The phase memory times are generally
two orders of magnitude longer than initially predicted, and were shown to increase
significantly at very low temperatures. Much remains to be understood about the
details of spin relaxation and decoherence in these materials, including the effects
of spectral and spin diffusion. Furthermore, no attempts have yet been made to
utilize nuclear spin in coherent spin manipulations. Indeed, no coherent manipu-
lations beyond spin echo and nutation measurements (Rabi oscillations) have been
reported. Here, the molecular magnetism community will be able to learn a great
deal from interaction with the biophysical EPR and quantum information proces-
sing communities. Progress along these lines will depend in part on material
development. To improve phase memory times, weakly coupled nuclear spins
should be removed completely or moved as far as possible in space from the
electron spin to limit dipolar hyperfine interactions. For detailed investigations of
226 J. van Slageren
Most MNMs have well defined spin ground states. The splitting of the ground state
by ZFS can be excellently studied by EPR techniques, as the many reviews on this
subject attest. In several cases, this has proved to be impossible, often in systems
that combine high molecular symmetry with competing exchange interactions,
such as Na2[Mo72VIFe30IIIO252(CH3COO)20(H2O)92]ca. 150 H2O (Fe30) [168] or
(C5H6N+)5[Fe13F24(OCH3)12O4]CH3OH4H2O (Fe13) [169]. In such a case, in EPR
spectra a single broad line is observed, which broadens and shifts downfield
upon cooling [169, 170]. Such behavior is also observed in magnetic nanoparticles
(MNP) [171] and exploitation of the classical models used in the MNP field may aid
the description and analysis of the EPR spectra of the above-mentioned MNMs.
The spin Hamiltonian framework seems to be unable to account for these
observations, and for Fe30 a different nature of the low-lying excitations (spin
waves) has been invoked, but only partial agreement with experimental data
from INS was obtained [172]. Fittipaldi et al. have reviewed this area [173] and
have recently published new results [174].
of lanthanide ions in lattices were recently reported, with phase memory times of
TM 50 ms at 2.5 K for Er3+:CaWO4 (105 atom% Er) [201] and TM values of up to
134 ms at 4 K for Yb3+:CaWO4 (0.0025% Yb) [202]. Indeed, some of the very first
spin-echo experiments were carried out on lanthanides in lattices, e.g., Er3+:CaWO4
(TM 14 ms at ~1.8 K) [203] and Er3+/Ce3+:CaWO4 (TM 25 ms at ~1.8 K for
the Ce3+ ion) [204]. In molecular systems, TM times of 430 ns were found for
{[Nd2(a-C4H3OCOO)6(H2O)2]}n [199].
References
59. Gregoli L, Danieli C, Barra A-L, Neugebauer P, Pellegrino G, Poneti G, Sessoli R, Cornia A
(2009) Chem Eur J 15:6456
60. Waldmann O, G udel HU (2005) Phys Rev B 72:094422
61. Zhong YC, Pilbrow JR (1999) In: Poole CP, Farach HA (eds) Handbook of electron spin
resonance, vol. 2. Springer, New York
62. van Slageren J, Vongtragool S, Gorshunov B, Mukhin AA, Karl N, Krzystek J, Telser J,
Muller A, Sangregorio C, Gatteschi D, Dressel M (2003) Phys Chem Chem Phys 5:3837
63. Kirchner N, van Slageren J, Dressel M (2007) Inorg Chim Acta 360:3813
64. Kozlov GV, Volkov AA (1998) Top Appl Phys 74:51
65. Robson PN (1967) In: Martin DH (ed) Spectroscopic techniques for far infra-red
submillimetre and millimetre waves. North-Holland, Amsterdam
66. Mukhin AA, Travkin VD, Zvezdin AK, Lebedev SP, Caneschi A, Gatteschi D (1998)
Europhys Lett 44:778
67. El Hallak F, van Slageren J, Gomez-Segura J, Ruiz-Molina D, Dressel M (2007) Phys Rev B
75:104403
68. Carbonera C, Luis F, Campo J, Sanchez-Marcos J, Camon A, Chaboy J, Ruiz-Molina D, Imaz
I, van Slageren J, Dengler S, Gonzalez M (2010) Phys Rev B 81:014427
69. Kirchner N, van Slageren J, Brechin EK, Dressel M (2005) Polyhedron 24:2400
70. Mukhin A, Gorshunov B, Dressel M, Sangregorio C, Gatteschi D (2001) Phys Rev B
63:214411
71. Piligkos S, Rajaraman G, Soler M, Kirchner N, van Slageren J, Bircher R, Parsons S, G udel
HU, Kortus J, Wernsdorfer W, Christou G, Brechin EK (2005) J Am Chem Soc 127:5572
72. Sieber A, Boskovic C, Bircher R, Waldmann O, Ochsenbein ST, Chaboussant G, G udel HU,
Kirchner N, van Slageren J, Wernsdorfer W, Neels A, Stoeckli-Evans H, Janssen S, Juranyi F,
Mutka H (2005) Inorg Chem 44:4315
73. Moro F, Piga F, Krivokapic I, Burgess A, Lewis W, McMaster J, van Slageren J (2010) Inorg
Chim Acta 363:4329
74. Abragam A, Bleany B (1986) Electron paramagnetic resonance of transition ions. Dover,
New York
75. Kirchner N, van Slageren J, Tsukerblat B, Waldmann O, Dressel M (2008) Phys Rev B
78:094426
76. Petukhov K, Hill S, Chakov N, Abboud K, Christou G (2004) Phys Rev B 70 p 054426
77. Rogez G, Rebilly JN, Barra AL, Sorace L, Blondin G, Kirchner N, Duran M, van Slageren J,
Parsons S, Ricard L, Marvilliers A, Mallah T (2005) Angew Chem Int Ed 44:1876
78. Vongtragool S, Mukhin A, Gorshunov B, Dressel M (2004) Phys Rev B 69:104410
79. Takahashi S, Edwards RS, North JM, Hill S, Dalal NS (2004) Phys Rev B 70:094429
80. Hill S, Maccagnano S, Park K, Achey R, North J, Dalal N (2002) Phys Rev B 65:224410
81. Park K, Novotny M, Dalal N, Hill S, Rikvold P (2002) Phys Rev B 66:144409
82. Lax B, Button KJ (1962) Microwave ferrites and ferrimagnets. McGraw-Hill, New York
83. Park K, Novotny MA, Dalal NS, Hill S, Rikvold PA (2002) J Appl Phys 91:7167
84. van Slageren J, Vongtragool S, Mukhin A, Gorshunov B, Dressel M (2005) Phys Rev B
72:020401
85. Smith GM, Lesurf JCG, Mitchell RH, Riedi PC (1998) Rev Sci Instrum 70:1787
86. Dressel M, Gorshunov B, Rajagopal K, Vongtragool S, Mukhin AA (2003) Phys Rev B
67:060405(R)
87. van Slageren J, Vongtragool S, Gorshunov B, Mukhin A, Dressel M (2009) Phys Rev B
79:224406
88. Schnegg A, Behrends J, Lips K, Bittl R, Holldack K (2009) Phys Chem Chem Phys 11:6820
89. Pedersen KS, Dreiser J, Nehrkorn J, Gysler M, Schau-Magnussen M, Schnegg A, Holldack K,
Bittl R, Piligkos S, Weihe H, Tregenna-Piggott P, Waldmann O, Bendix J (2011) Chem
Commun 6918
90. Dreiser J, Schnegg A, Holldack K, Pedersen KS, Schau-Magnussen M, Nehrkorn J,
Tregenna-Piggott P, Mutka H, Weihe H, Bendix J, Waldmann O (2011) Chem Eur J 17:7492
New Directions in Electron Paramagnetic Resonance Spectroscopy 231
127. Ohmichi E, Mizuno N, Kimata M, Ohta H, Osada T (2009) Rev Sci Instrum 80:013904
128. Petukhov K, Bahr S, Wernsdorfer W, Barra AL, Mosser V (2007) Phys Rev B 75:064408
129. Bahr S, Petukhov K, Mosser V, Wernsdorfer W (2007) Phys Rev Lett 99:147205
130. Bahr S, Petukhov K, Mosser V, Wernsdorfer W (2008) Phys Rev B 77:064404
131. Bal M, Friedman JR, Chen W, Tuominen MT, Beedle CC, Rumberger EM, Hendrickson DN
(2008) EPL 82:17005
132. de Loubens G, Kent AD, Krymov V, Gerfen GJ, Beedle CC, Hendrickson DN (2008) J Appl
Phys 103:07B910
133. Rieger PH (2007) Electron spin resonance-analysis and interpretation. RSC, Cambridge
134. Bertini I, Martini G, Luchinat C (1994) In: Poole J, Farach HA (eds) Handbook of electron
spin resonance. AIP, New York
135. Eaton SS, Eaton GR (2000) Biol Magn Reson 19:29
136. Tsukerblat BS, Botsan IG, Belinskii MI, Fainzilberg VE (1985) Mol Phys 54:813
137. Meier F, Levy J, Loss D (2003) Phys Rev Lett 90:047901
138. Meier F, Levy J, Loss D (2003) Phys Rev B 68:134417
139. Troiani F, Affronte M (2011) Chem Soc Rev 40:3119
140. Trif M, Troiani F, Stepanenko D, Loss D (2010) Phys Rev B 82:045429
141. Troiani F, Affronte M, Carretta S, Santini P, Amoretti G (2005) Phys Rev Lett 94:190501
142. Candini A, Lorusso G, Troiani F, Ghirri A, Carretta S, Santini P, Amoretti G, Muryn C, Tuna
F, Timco G, McInnes EJL, Winpenny REP, Wernsdorfer W, Affronte M (2010) Phys Rev
Lett 104:037203
143. Trif M, Troiani F, Stepanenko D, Loss D (2008) Phys Rev Lett 101:217201
144. Schweiger A, Jeschke G (2001) Principles of pulse electron paramagnetic resonance. Oxford
University Press, Oxford
145. Sato K, Nakazawa S, Rahimi R, Ise T, Nishida S, Yoshino T, Mori N, Toyota K, Shiomi D,
Yakiyama Y, Morita Y, Kitagawa M, Nakasuji K, Nakahara M, Hara H, Carl P, H ofer P,
Takui T (2009) J Mater Chem 19:3739
146. Nakazawa S, Sato K, Shiomi D, Yano M, Kinoshita T, Franco MLTMB, Lazana MCRLR,
Shohoji MCBL, Itoh K, Takui T (2011) Phys Chem Chem Phys 13:1424
147. Ruby RH, Benoit H, Jeffries CD (1962) Phys Rev 127:51
148. Hoffmann SK, Lijewski S, Goslar J, Ulanov VA (2010) J Magn Reson 202:14
149. Schlegel C, van Slageren J, Timco G, Winpenny REP, Dressel M (2011) Phys Rev B
83:134407
150. Mitrikas G, Sanakis Y, Raptopoulou CP, Kordas G, Papavassiliou G (2008) Phys Chem
Chem Phys 10:743
151. Bertaina S, Gambarelli S, Mitra T, Tsukerblat B, M uller A, Barbara B (2010) Nature
466:1006
152. Schlegel C, van Slageren J, Manoli M, Brechin EK, Dressel M (2008) Phys Rev Lett
101:147203
153. Takahashi S, van Tol J, Beedle CC, Hendrickson DN, Brunel L-C, Sherwin MS (2009) Phys
Rev Lett 102:087603
154. Kulik LV, Lubitz W, Messinger J (2005) Biochemistry 44:9368
155. Beardwood P, Gibson JF, Bertrand P, Gayda JP (1983) Biochim Biophys Acta 742:426
156. Lorigan GA, Britt RD (1994) Biochemistry 33:12072
157. Britt RD, Zimmermann JL, Sauer K, Klein MP (1989) J Am Chem Soc 111:3522
158. Dilg AWE, Mincione G, Achterhold K, Iakovleva O, Mentler M, Luchinat C, Bertini I,
Parak FG (1999) J Biol Inorg Chem 4:727
159. Dilg AWE, Capozzi F, Mentler M, Iakovleva O, Luchinat C, Bertini I, Parak FG (2001) J Biol
Inorg Chem 6:232
160. Guigliarelli B, More C, Fournel A, Asso M, Hatchikian EC, Williams R, Cammack R,
Bertrand P (1995) Biochemistry 34:4781
161. Bertrand P, Gayda JP, Rao KK (1982) J Chem Phys 76:4715
162. Harris EA, Yngvesson KS (1968) J Phys C 1:990
New Directions in Electron Paramagnetic Resonance Spectroscopy 233
201. Bertaina S, Gambarelli S, Tkachuk A, Kurkin IN, Malkin B, Stepanov A, Barbara B (2007)
Nat Nanotech 2:39
202. Rakhmatullin R, Kurkin I, Mamin G, Orlinskii S, Gafurov M, Baibekov E, Malkin B,
Gambarelli S, Bertaina S, Barbara B (2009) Phys Rev B 79:172408
203. Mims WB, Nassau K, McGee JD (1961) Phys Rev 123:2059
204. Mims WB (1968) Phys Rev 168:370
Index
235
236 Index
H Nanoshelters, 79
Heterogeneous catalysis, 1 Neomycin-responsive riboswitch, 176
Hexene, 16 NhaA, 146
HFEPR, 227 Nickel(II)-ethylenediamine-diacetic acid
Homogeneous catalysis, 1 (NiEDDA), 103, 134, 137
Human immunodeficiency virus (HIV), 177 Nitroxides, 92, 163, 166
Hydrolytic kinetic resolution (HKR), 14 Non-covalent interactions, 67
Hyperfine spectroscopy, 159, 165, 169, 191 Nucleic acids, 159
Hyperfine sublevel correlation (HYSCORE), 4, secondary structure, 185
14, 47, 73, 170, 184 tertiary structure, 187
Nucleobases, 162
Nucleotide binding domains (NBDs), 135
I
Interspin distances, 141
Intrinsically disordered proteins, 91 O
Iodoacetamido-tetramethyl-1-pyrrolidinyloxy Octene, 16
radical labels (IAP), 123 Oligodeoxynucleotide, 2-fluorohypoxanthine,
Iron oxygenases, non-heme, 8 163
Oxidation, 12
selective, 19
K state, 12
KcsA, 128
P
L Parkinsons disease (PD), 105
Light harvesting complex (LHCIIb), 138 PEO-PPO-PEO, 84
Lightoxygenvoltage (LOV) domains, Phase-memory times, 222
ENDOR, 50 Phenolate radicals, 14
Linear alpha olefins (LAOs), 16 Phenoxyl radical, 14
Lipid A flippase, 135 Photocatalysts, titanium dioxide-based, 27
Photolyase, 41, 45, 47
Phototropin, 41
M Photovoltaics, 85
Magnetic nanoparticles, 226 Plugged hexagonal templated silica
Membrane binding, 106 (PHTS), 26
Membrane proteins, 121 Poly(N-isopropylacrylamide) (PNIPAAM), 77
spin labeling, 123 Poly(N-isopropylmethacrylamide)
Mesoglobules, 79 (PNIPMAM), 79
Metal ion binding sites, 184 Polymer electronics, 85
Metallacycloheptane, 17 Polymers, 67
Metalloenzymes, artificial, 8 Potassium channel KcsA, 128
Metal organic framework (MOF), 29 Prion protein H1, 104
Methylbenzylamine, 7 Proteins, intrinsically disordered, 91, 103
Micelles, 107 SDSL, 92
Molecular nanomagnet (MNM), 199, 200 PROXYL, 72
MTSSL (1-oxyl-tetra-methylpyrroline-3- Pulsed electron-electron double resonance
methyl)-methanthiosulfonate, 93, 123 (PELDOR), 4, 159, 173
Pulse-EPR, 183
N
Na+/H+ antiporter, 146 Q
Nanomagnetism, 218 Quantum coherence, 199
Index 237
T Z
Tat protein, 177 Zeolites, 24
TEMPO, 72, 78, 81 Zero-field splitting (ZFS), 199, 200, 205