Gd(III) complexes for electron-electron dipolar spectroscopy: Effects of deuteration, pH and zero field splitting

Article abstract of DOI:10.1016/j.jmr.2015.08.009

Spectral parameters of Gd(III) complexes are intimately linked to the performance of the Gd(III)-nitroxide or Gd(III)-Gd(III) double electron-electron resonance (DEER or PELDOR) techniques, as well as to that of relaxation induced dipolar modulation enhancement (RIDME) spectroscopy with Gd(III) ions. These techniques are of interest for applications in structural biology, since they can selectively detect site-to-site distances in biomolecules or biomolecular complexes in the nanometer range. Here we report relaxation properties, echo detected EPR spectra, as well as the magnitude of the echo reduction effect in Gd(III)-nitroxide DEER for a series of Gadolinium(III) complexes with chelating agents derived from tetraazacyclododecane. We observed that solvent deuteration does not only lengthen the relaxation times of Gd(III) centers but also weakens the DEER echo reduction effect. Both of these phenomena lead to an improved signal-to-noise ratios or, alternatively, longer accessible distance range in pulse EPR measurements. The presented data enrich the knowledge on paramagnetic Gd(III) chelate complexes in frozen solutions, and can help optimize the experimental conditions for most types of the pulse measurements of the electron-electron dipolar interactions.

Full text of DOI:10.1016/j.jmr.2015.08.009

Journal of Magnetic Resonance 259 (2015) 163–173
Contents lists available at ScienceDirect
Journal of Magnetic Resonance
journal homepage: www.elsevier.com/locate/jmr
Gd(III) complexes for electron–electron dipolar spectroscopy: Effects
of deuteration, pH and zero field splitting
Luca Garbuio a, Kaspar Zimmermann b, Daniel Häussinger b, Maxim Yulikov a,
⇑
a Laboratory of Physical Chemistry, ETH Zurich, Zurich, Switzerland
b Department of Chemistry, University of Basel, Basel, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Spectral parameters of Gd(III) complexes are intimately linked to the performance of the Gd(III)–
nitroxide or Gd(III)–Gd(III) double electron–electron resonance (DEER or PELDOR) techniques, as well
as to that of relaxation induced dipolar modulation enhancement (RIDME) spectroscopy with Gd(III) ions.
These techniques are of interest for applications in structural biology, since they can selectively detect
site-to-site distances in biomolecules or biomolecular complexes in the nanometer range. Here we report
relaxation properties, echo detected EPR spectra, as well as the magnitude of the echo reduction effect in
Gd(III)–nitroxide DEER for a series of Gadolinium(III) complexes with chelating agents derived from
tetraazacyclododecane. We observed that solvent deuteration does not only lengthen the relaxation
times of Gd(III) centers but also weakens the DEER echo reduction effect. Both of these phenomena lead
to an improved signal-to-noise ratios or, alternatively, longer accessible distance range in pulse EPR
measurements. The presented data enrich the knowledge on paramagnetic Gd(III) chelate complexes
in frozen solutions, and can help optimize the experimental conditions for most types of the pulse
measurements of the electron–electron dipolar interactions.
Received 26 May 2015
Revised 6 August 2015
Available online 20 August 2015
Keywords:
Lanthanide
Gadolinium(III)
Pulse EPR
DEER
RIDME
Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction
solutions, in samples of low concentrations, down to a few lM,
and with nearly any arbitrary molecular weight of the spin-
labeled biomacromolecules [18–21]. In cases of the stand-alone
application of PDS techniques coarse-grained structural models
are obtained instead of the molecular structures at atomic resolu-
tion [22–24]. However, such PDS-based structural models can be
capable of answering multiple biological questions. Elaborate
signal assignment, as in NMR spectroscopy, is not necessary since
the PDS signals originate uniquely from the spin labels.
A possible further advance in PDS methodology can be based on
experiments with new spin labels with spectral characteristics dif-
ferent from the ones the nitroxide radicals. In the last decade it was
shown that Gd(III) chelate complexes are suitable for nanometer-
range distance measurements, especially at high fields and high
frequencies [25]. In particular, an advantageous factor is the nar-
rowing of the central resonance line in the Gd(III) EPR spectrum
at increasing magnetic fields, which allows for better sensitivity
and increased excitation bandwidth for a given power and dura-
tion of a microwave pulses. Contrary to nitroxide radicals, when
using Gd(III) as spin labels, one can also benefit from the lack of
orientation selection effects [26]. Importantly, the spectroscopic
separation of Gd(III) centers from nitroxide radicals is possible,
which allows for orthogonal distance measurement schemes
[27–29]. Noteworthy, Gd(III) is relatively stable against reduction,
A series of pulse techniques in electron paramagnetic resonance
(EPR) that can be generally named as electron–electron pulse
dipolar spectroscopy (PDS) [1–9] in combination with site directed
spin labeling (SDSL) [10–13] is of great relevance in structural
investigations of bio-macromolecules as such an approach fur-
nishes distance information in the 1.5–12 nm range [14,15]. Stable
nitroxide radicals are currently a common choice in SDSL-based
studies, and the nitroxide–nitroxide double electron–electron
resonance (DEER), a pulse EPR technique based on the use of two
distinct excitation frequencies, is the most popular distance
measurement technique [16,17]. Apart from DEER, the relaxation
induced dipolar modulation enhancement (RIDME) [5–7] is a
single-frequency technique that also detects static dipolar interac-
tion, which can be then analyzed to get distances between spin
labels. Other single-frequency experiments include the double
quantum coherence (DQC) [8] and the SIFTER [9] techniques.
The PDS techniques can complement (and in some cases
substitute) X-ray crystallography or NMR spectroscopy by provid-
ing long-range distance distributions in non-crystalline frozen
⇑
Corresponding author.
E-mail address: maxim.yulikov@phys.chem.ethz.ch (M. Yulikov).
http://dx.doi.org/10.1016/j.jmr.2015.08.009
1090-7807/Ó 2015 Elsevier Inc. All rights reserved.

164
L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
making it favorable for in-cell studies, in which the cellular
reducing environment is able to degrade nitroxide-based spin
labels [30].
Experimental observations seem to suggest an effect of the ZFS
magnitude on the DEER echo reduction effect in Gd(III)–nitroxide
DEER experiments [28,42]. This phenomenon, i.e. the intensity
reduction of the refocused echo upon application of the pump
pulse, is one of the main factors that affects the sensitivity of the
Gd(III)–nitroxide DEER method at X and Q band. The detection in
this case is performed on the Gd(III) centers and the pump pulse
acts on the nitroxide species. In nitroxide–nitroxide DEER
experiments the DEER echo reduction is described by the Bloch–
Siegert effect [43,44]. The off-resonant pump pulse produces
dynamic phase shifts of spin packets that are forming the
refocused echo, thus resulting in a partial echo defocusing. Since
in Gd(III)–nitroxide DEER low-spin (S = 1/2) nitroxide radicals
require significantly higher microwave power than Gd(III) centers,
the known Bloch–Siegert effect is stronger in this case compared to
the nitroxide–nitroxide or Gd(III)–Gd(III) DEER experiments [45].
So far, most of the literature addressed the use of Gd(III) as con-
trast agents in magnetic resonance imaging (MRI) [31–40]. One
important consequence of those activities was a development of
bio-compatible Gd(III) tags. In the human body, the aqua ion [Gd
(H2O)9]3+ binds to endogenous chemical species leading to high
in-vivo toxicity. For this reason, several cyclic chelating agents
possessing high thermodynamic stability in complex with Gd(III)
as well as kinetic inertness against ion dissociation have being
developed [32].
Since Gd(III)-complexes have attracted growing attention for
their use in PDS, a study of their spectroscopic properties regarding
the dipolar spectroscopy applications is meaningful. The previous
MRI-related studies concentrated on the number and properties
of water protons in the first and second hydration sphere, which
is not of direct interest for PDS. However, these MRI-related publi-
cations discussed to quite some extent the zero-field splitting (ZFS)
in Gd(III) complexes, which is critically important for the PDS stud-
ies. In MRI, ZFS of Gd(III) centers was considered very important as
it affects the Gd(III) electronic spin relaxation as well as the accu-
racy of the Gd(III)–water proton distance measurements [34]. Gd
(III) echo-detected (ED) EPR spectra recorded at high fields were
analyzed with respect of ZFS determination [35] and a stochastic
superposition model was utilized to extract the characteristic ZFS
parameters, D and E and their probability distributions, in glassy
matrixes. In the spectra simulation, D was varied according to a
Gaussian distribution (or a bimodal distribution consisting of two
Gaussians with positive and negative mean D value) and the E/D
distribution could be approximated by a second order polynomial
[35]. EPR experiments at 240 GHz allowed Benmelouka et al. [36]
to determine the magnitude and sign of the ZFS parameters in fro-
zen solutions for several Gd(III)-complexes. A correlation between
complex symmetry and sign of D was established [36].
In solution, it was proposed that a time-dependent or transient
ZFS [37] acts in addition to a time-independent or static ZFS [38].
Whereas the former arises from the perturbation of the ligands
field due to instantaneous distortions or vibrations, the latter,
due to the static crystal field, is modulated by the Brownian motion
and rotation of the complex and it was found dependent on the
nature of the ligands [39]. The study of the ZFS interaction in these
complexes is complicated by the contemporaneous presence of
structural isomers and different sets of ligand-field parameters.
Multiple-frequency continuous-wave (CW) EPR investigations
were performed in order to evaluate the contribution of the static
and transient ZFS in the temperature range 276–350 K [39]. It was
suggested that inter-conversion between coordination geometries
does not affect the magnitude of the transient ZFS which should
be dominated by faster processes such as vibrations and as a
consequence, coordination isomers are indistinguishable at any
EPR field strength [40].
However, at least at
Q band, the DEER echo reduction in
Gd(III)–nitroxide case depends on the strength of ZFS interaction
of Gd(III) centers and thus cannot be exclusively caused by the
Bloch–Siegert effect [43,44].
The PDS experiments with Gd(III) centers are performed at
cryogenic temperatures. Thus, relaxation properties of Gd(III)
approximately in the range 5–30 K are of particular importance
for these EPR techniques. Raitsimring et al. [46] recently proposed
that the mechanism concurring to the Gd(III) phase memory time
is transition dependent. At W band the contribution to the relax-
ation of the |ꢀ1/2i M |+1/2i transition was suggested to be domi-
nated by nuclear spin diffusion, whereas other transitions were
analyzed according to a ZFS-driven relaxation mechanism. The
ZFS-driven relaxation at the central transition might contribute
more strongly at Q band (35 GHz) than at W band (95 GHz) as this
contribution would scale down with the ZFS-induced anisotropy of
the central transition, which decreases proportional to D2/
((gb)2B0). Since the central transition is influential in PDS with Gd
(III), the determination of the interplay between different relax-
ation mechanisms is of interest. In particular, the nuclear spin dif-
fusion drives electron transverse relaxation trough stochastic
fluctuations of hyperfine fields, thus, the level of deuteration of
the sample can influence the relaxation properties of Gd(III). As
deuterons possess a smaller magnetic dipole moment than pro-
tons, deuteration slows down the phase memory time of the
observed species and it is thus used to increase the longest mea-
surable length of the DEER trace [15,47]. This way one can extend
the upper distance limit reachable by PDS and improve sensitivity.
The longitudinal relaxation time of Gd(III) centers (T1) is also
important since it is directly connected to the minimal possible
shot repetition time (i.e. the time in DEER between an experiment
and the next for a given pump position). A shorter T1 allows thus to
decrease the experimental time needed to obtain a certain signal-
to-noise ratio (S/N).
The purpose of this paper is to inspect the spectroscopic prop-
erties of Gd(III) chelate complexes at conditions that are relevant
for PDS experiments. We particularly focus on DOTA (1,4,7,10-tetra
azacyclododecane-1,4,7,10-tetraacetic acid) because it is the most
common cyclic complexing agent for Gd(III) and [Gd(DOTA)]ꢀ is
frequently employed in DEER measurements [48]. Upon chemical
modification of the basic structure of DOTA, several derivatives
can be obtained. Representatives of this family are DOTAM (1,4,7,
10-tetrakis(carbamoylmethyl)-1,4,7,10-tetraazacyclododecane),
M8DOTA [49] and DOTP (1,4,7,10-tetraazacyclododecane-1,4,7,10-
tetra(methylene phosphonic acid)) (Fig. 1) that we tested to obtain
a systematic comparison with DOTA. We report a series of mea-
surements at Q band, which is one of the regularly used microwave
bands for measurements with Gd(III) and we concentrate on the
temperature of 10 K, used in several previous studies [29,50,51].
First, the determination of the ZFS parameters is performed for
Here we are interested in discussing the effect of ZFS term on
the performance of PDS experiments. As just mentioned, ZFS
affects heavily the appearance of the Gd(III) ED EPR spectrum.
The spectrum is characterized by a narrow peak, arising from the
central transition |ꢀ1/2i M |+1/2i, and by a broad background
connected to the other six remaining transitions [34,41]. The mag-
nitude of the ZFS influences the width of the central transition,
which in turn affects sensitivity and modulation depth in DEER
experiments since the excitation bandwidth of the microwave
pulses is limited. In PDS measurements the detection frequency
(or the pump frequency in DEER) is typically set at the maximum
absorption of Gd(III) ED EPR spectrum where the central transition
|ꢀ1/2i M |+1/2i dominates. A weak ZFS is thus desirable to allow
for a higher fraction of spins to be pumped or observed.

L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
165
O
O
O
O
O
O
HO
P
P
P
P
OH
HO
N
N
N
N
OH
H2N
N
N
N
N
NH2
N
N
N
N
HO
OH
HO
OH
HO
O
OH
O
H2N
O
NH2
O
HO
OH
O
O
DOTA
DOTAM
DOTP
H
N
H
N
O
S
O
N
N
S
O
O
S
S
N
N
N
N
N
N
N
N
HO
HO
OH
OH
O
O
O
OH
O
OH
M8DOTA(8S)SPy
M8DOTA(4S4R)SPy
Fig. 1. Chemical structures of the investigated chelating agents.
the above listed Gd(III)-complexes. Second, an experimental pic-
ture of the impact of matrix deuteration on relaxation times T2,
T1 at the maximum intensity point of the Gd(III) spectrum is given.
Next, the values of the Gd(III)–nitroxide DEER echo reduction are
reported and a correlation with the ZFS magnitude is proposed.
After this we provide reference data on the dependence of the
relaxation parameters (T1, T2) on pH in a biologically-pertinent
range. And finally we compare longitudinal and transverse relax-
ation of different complexes in a D2O/glycerol-d8 glassy matrix at
10 K.
2.2. Gd(III)-complex formation
The molecular structures of the studied ligands and correspond-
ing abbreviated names are shown in Fig. 1. [Gd(M8DOTA(8S)SPy)]
and [Gd(M8DOTA(4R4S)SPy)] were prepared in aqueous buffered
solution (100 mM ammonium acetate) according to Ref. [49]. Due
to the extremely high affinity of the M8DOTA-ligands to gadolinium
ions, the resulting [Gd(M8DOTA)] complexes could be purified by
HPLC followed by a Sep-pak-C18 column (Waters) for desalting.
Aqueous solutions of the [Gd(M8DOTA)] complexes were mixed
with the corresponding buffer, the pH was adjusted, the solution
was lyophilized and the final concentration was obtained by dis-
solving in the desired amount of water (H2O or D2O) and glycerol-
2. Materials and methods
d8 All other gadolinium complexes were prepared as follows. 300
of a 128 mM buffer solution of GdCl3ꢁ6H2O (38.4 mol, 1 eq.) were
added to 1 ml of a 50 mM buffer solution of chelating agent
(50 mol, 1.3 eq.). The solution was stirred for 24 h. The presence
of non-chelated Gd(III) ions was detected from the color of the solu-
tion by eye using xylenol orange as indicator according to the pro-
ll
2.1. Chemicals
l
The following acronyms will be used: DOTA, 1,4,7,10-tetraazacy
clododecane-1,4,7,10-tetraacetic acid; DOTP, 1,4,7,10-tetraazacyclo
dodecane-1,4,7,10-tetra(methylene phosphonic acid); DOTAM, 1,4
,7,10-tetrakis(carbamoylmethyl)-1,4,7,10-tetraazacyclododecane;
MES, 2-(N-morpholino)ethanesulfonic acid; Xylenol Orange, 3,30-
Bis[N,N-bis(carboxymethyl)aminomethyl]-o-cresolsulfonphthalein
disodium salt. M8DOTA(8S)SPy, (2S,20S,200S)-2,20,200-((2S,5S,8S,11S)-
2,5,8,11-tetramethyl-10-((S)-1-oxo-1-((2-(pyridin-2yldi-sul-fanyl)-
ethyl)amino) propan-2-yl)-1,4,7,10-tetraazacyclo-dodecane-1,4,7
-triyl)tripropanoic acid; M8DOTA(4R4S)SPy, (2R,20R,200R)-2,20,
200-((2S,5S,8S,11S)-2,5,8,11-tetramethyl-10-((R)-1-oxo-1-((2-(pyri-
din-2yldi-sul-fanyl)-ethyl)amino) propan-2-yl)-1,4,7,10-tetraazacy
clo-dodecane-1,4,7-triyl)tripropanoic acid; when simultaneously
referring to both M8DOTA(8S)SPy and M8DOTA(4R4S)Spy ligands,
an abbreviation M8DOTA will be used. Unless stated otherwise, all
commercially available chemicals were used as received. M8DOTA
ligands were synthesized according to Ref. [49]. DOTAꢁ6H2O, DOTP
and DOTAM were purchased from Macrocyclics, Dallas. GdCl3ꢁ6H2O
(99.99%), MES (99.5%), sodium acetate (>99%), D2O (99.9%, d2) were
purchased from Sigma–Aldrich. Glycerol (C3H8O3, 99%) was pur-
chased from Honeywell. Perdeuterated glycerol-d8 (C3D8O3, 98%)
was purchased from Isotec. Xylenol Orange was purchased from
TCI. The buffer was prepared as follow: 1 g of MES was dissolved
in 20 ml of H2O or D2O. The pH of the corresponding solution
(250 mM) was adjusted to 5.8 with 0.1 M NaOH aqueous solution
(H2O or D2O). At this pH, precipitation of Gd(III)–hydroxides during
sample preparation is prevented. MES was used as it does not
coordinate to Gd(III).
l
cedure described earlier [52]. 100 ll of the Gd(III)-complex solution
were diluted with 1.2 ml of a 50 mM acetate buffer solution at
pH = 5.8 and few drops of a 0.016 mM xylenol orange solution were
added. An orange color of the resulting solution was indicative of
the negligible presence of Gd(III) aquo-ions. Next, 10
(III)-complex solution were diluted to 490 l of buffer solution. To
25 l of this solution were added 25 l of glycerol (or glycerol-d8)
in order to obtain a final concentration of ꢂ300 M Gd(III)-
ll of the Gd
l
l
l
l
complex in 1/1 volume-to-volume buffer/glycerol mixture. The
addition of glycerol is necessary for EPR samples to form a glassy
matrix and thus a homogeneous distribution of Gd(III) species upon
shock freezing. EPR quartz tubes of outer diameter of 2.95 mm were
filled with the desired sample and finally shock frozen by immer-
sion into liquid nitrogen. Samples were stored at ꢀ80° C. In the
course of this paper, the binary solvent system buffer/glycerol will
be simply referred to as glassy matrix and the contained isotopo-
logues (H2O vs. D2O and glycerol vs. glycerol-d8) will be specified
on a case-by-case basis.
2.3. Relaxation time traces
All EPR measurements reported here were performed at 10 K
with a home-built high-power Q-band spectrometer [53] equipped
with a rectangular cavity that allows for oversized samples [54].

166
L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
Table 1
home-written program based on EasySpin function ‘pepper’ [55].
Spectra were computed by using a bimodal distribution consisting
of two Gaussian distributions of the D parameter centered at the
Magnitudes of hDi and
r parameters used in simulations of the ED EPR spectra for
different Gd(III) complexes. The spectra of Gd(III) centers were simulated by
assuming a bi-Gaussian distribution of the D parameter centered at the mean values
mean value hDi and ꢀhDi, with the same standard deviation
r in
hDi with the standard deviation
distribution (see text). As an exception, the ED EPR spectrum for DOTP was simulated
by assuming two Gaussian distributions with different |hDi| and
r and a second order polynomial for the E/D
both cases (with an exception for DOTP that is explained in the
Results section). A second order polynomial form of the E/D distri-
bution: P(E/D) / (E/D) ꢀ 2(E/D)2 was used [28,55]. The simulation
parameters are listed in Table 1.
r
.
Chelating agent
hDi (MHz)
r
DOTAM
DOTA
M8DOTA(4R4S)SPy
M8DOTA(8S)SPy
DOTP
560
685
890
1260
hDi/3.5
hDi/3
hDi/4
hDi/4
D1/3
2.5. Echo reduction effect
D1 = ꢀ750 (60%)
Echo reduction experiments (Fig. 2) were performed by running
D2 = +1950 (40%)
D2/10
a static four-pulse DEER scheme [3,4] with
p/2 = 12 ns, p = 12 ns
pulses and a +x/ꢀx phase cycle on the first pulse. The time position
of the pump pulse was set to overlap with the primary echo and
was kept fixed during the echo intensity measurement (i.e. the first
inter-pulse delay was set to 400 ns, and the pump pulse position
The sample temperature was stabilized with a He-flow cryostat (ER
4118CF, Oxford Instruments). The T2 time traces were obtained by
a Hahn echo sequence with the pulse durations (
) = 32 ns. T1 time traces were recorded with an inversion recov-
ery experiment with an inversion -pulse of 32 ns and detection
sequence of ( /2) = 52 ns and ( ) = 104 ns. Relaxation time traces
p/2) = 16 ns and
was at 800 ns after the p/2 pulse). It has been checked experimen-
(p
tally that the reduction of the Gd(III) DEER echo is not dependent
on the presence of nitroxide species [28], therefore the experiment
was set up for the samples with only Gd(III) chelate complexes pre-
sent. At the pump frequency, to set the proper microwave power
p
p
p
were recorded at the intensity maximum of Gd(III) ED EPR
spectrum.
for the inversion pulse, a 12 ns
p pulse was set at the maximum
intensity of the Gd(III) spectrum, then the microwave power was
increased by 12 dB by opening a precise attenuator (this was the
experimentally estimated difference in the power between Gd
(III) and nitroxide species), and the detection field was reduced
by 11 mT in order to put the spectral position of the pump pulse
at the place that would correspond to the maximum of the nitrox-
ide spectrum at given settings. Detection frequency was set
300 MHz lower than the pumping frequency, which corresponded
to the maximum intensity of the Gd(III) spectrum. Such pulse
settings were typically employed in previously reported Q-band
Gd(III)–nitroxide DEER experiments in our group [27]. At this
2.4. Echo detected EPR spectra
ED EPR spectra of Gd(III) complexes were recorded with the
Hahn echo detection sequence at different pulse lengths and at dif-
ferent interpulse delays d1 (i.e. the time between the first and the
second pulse in the Hahn echo sequence). We used ‘hard’
(p/2 = 12 ns, p = 12 ns) and ‘soft’ (p/2 = 16 ns, p = 32 ns) pulses to
check for any differences due to the change of the excitation
bandwidth. Simulations of the ED EPR spectra were carried out
according to the previously described procedure [28] with a
obs ~10 mT~300MHz
pump
1
1
pump
a
b
detection
0.8
0.8
0.6
0.4
0.2
0
0.6
0.4
0.2
0
1280
1285
1290
1295
1300
1305
1240
1260
1280
1300
1320
1340
B [mT]
B [mT]
c
d
2
det
d1 d1
d2
d2
t
pump
50
100
150
200
d3
d3
t [ns]
Fig. 2. (a) ED EPR spectrum of [Gd(DOTA)]ꢀ (blue) and nitroxide (red) at 10 K at Q band. The frequency offset between pump and detection frequency is approximately
300 MHz. (b) ED EPR spectrum of [Gd(DOTA)]ꢀ in a larger magnetic field range. (c) Pulse sequence in the Gd(III)–nitroxide DEER experiment. The detection pulses (blue) are
set on intensity maximum of Gd(III) ED EPR spectrum and the pump frequency (red) at the maximum of nitroxide ED EPR spectrum. h is the nominal flip angle of the pump
pulse. (d) Echo reduction effect in [Gd(M8DOTA(8S)SPy)]: the shape of the refocused echo in the DEER experiment with (blue) and without (red) the pump pulse h =
p applied.
(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
167
frequency offset, for the resonator used, 12 ns was the shortest sys-
tematically achievable duration of detection pulses. In the follow-
ing, the echo reduction is expressed in percent ratio between the
intensity of the echo with and without applying the inversion pulse
on the pump frequency.
determined with certainty. The simulation presented in Fig. S1
was performed with a negative hD1i value. For the symmetric bimo-
dal distributions of D values (DOTA, DOTAM, M8DOTA) the esti-
mated error is between 5% and 10% for hDi and between 10% and
20% for
r. For DOTP we estimate a larger error between 10% and
20% for hD1i and hD2i and an error between 30% and 40% for r1
and 2 values.
r
3. Results
3.1. Determination of the ZFS magnitude of Gd(III)-complexes
3.2. Effect of matrix deuteration on the relaxation of [Gd(DOTA)]ꢀ
First, we determined the strength of the ZFS interaction in the
discussed series of macrocyclic complexes (Fig. 1). The experimental
ED EPR spectra of [Gd(DOTAM)]3+, [Gd(DOTA)]ꢀ, two stereoisomers
of [Gd(M8DOTA)] and [Gd(DOTP)]5ꢀ and corresponding simulations
are shown in Fig. S1. The parameters used in these simulations are
listed in Table 1. The hDi value for DOTA is consistent with our pre-
vious calculations [42] and with the prior results of other authors
[35,36], thus confirming the validity of the ZFS determination
approach. The strength of obtained characteristic |hDi| values for
these macrocyclic complexes follows the trend DOTAM < DOTA <
M8DOTA(4R4S)SPy < M8DOTA(8S)SPy < DOTP (Table 1). Note how-
ever, that for DOTP a different type of the bimodal distribution had
to be used in simulation with one Gaussian component character-
In order to better understand the effect of deuteration on relax-
ation properties of Gd(III) centers, we performed a series of exper-
iments with [Gd(DOTA)]ꢀ. Fig.
3 shows the transverse and
longitudinal relaxation time traces in linear and semi-logarithmic
representation obtained for [Gd(DOTA)]ꢀ in glassy matrix water/
glycerol (1:1 v/v). The longitudinal relaxation is very weakly
changing upon deuteration of the solvent. The same outcome
was reported on nitroxide-labeled LHCIIb [56] where T1 of the
nitroxide radicals were virtually unaffected by buffer deuteration.
In the latter case the small or even insignificant change in longitu-
dinal relaxation was rationalized by evaluating that T1 is mainly
driven by fluctuations of the hyperfine field induced by libration
motion of nitroxide moiety [57]. Here, similarly, the thermal fluc-
tuations of orbit-lattice coupling in Gd(III) centers should be
responsible for the main contribution to the longitudinal relax-
ation at cryogenic temperatures, so that the hyperfine interactions
with the magnetic nuclei of solvent molecules exert only a very
small impact on T1.
In contrast to longitudinal relaxation, the transverse relaxation
traces manifest a remarkable sensitivity to proton/deuterium rela-
tive concentration (Fig. 3), again comparable to the effect observed
on nitroxide spin labels [15,47]. This has an important conse-
quence for the DEER technique because a longer T2 of the detected
species (Gd(III) ions in the Gd(III)–nitroxide DEER) allows detection
of longer distances and better sensitivity. In the case at hand, a
ized by |hD1i| = 750 MHz and
r
1 = D1/3, and the other component
characterized by hD2i = +1950 MHz and
r2 = D2/10. The relative
weight of the first component of 60% and second component of
40% was used in the simulation shown in Fig. S1. Despite the two-
component simulation, the discrepancy between the simulation
and the experimental ED EPR spectrum is still somewhat worse
for [Gd(DOTP)]5ꢀ sample, as compared to other four (Fig. S1).
Importantly, only the second component has a hDi value that is
larger in absolute magnitude than the one for M8DOTA(8S). The sign
of the second component determines the type of the asymmetry of
the ED EPR spectrum of [Gd(DOTP)]5ꢀ complex and could be
assigned to be positive. The sign of the hD1i value is less critical to
the overall shape of the spectrum and thus it could not be
length of the DEER trace of 12–13
ls should be possible, thus
0
-2
[Gd(DOTA)] –
a
b
0.8
0.6
-4
WG
WdG
dWG
dWdG
0.4
-6
0.2
-8
0
-10
0
5
10
15
0
2
4
6
8
10
12
t [ s]
t [ s]
0
-2
-4
-6
1
0.5
0
c
d
-0.5
0
2
4
6
8
0
1
2
3
4
5
6
t [ms]
t [ms]
Fig. 3. Solvent dependency of the transverse (a and b) and longitudinal (c and d) relaxation time traces of [Gd(DOTA)]ꢀ in linear (a and c) and semilogarithmic (b and d)
representation. Measurements were recorded at 10 K at Q band at pH 5.8 and in different solvents: red (H2O/glycerol), blue (H2O/glycerol-d8), green (D2O/glycerol) and black
(D2O/glycerol-d8). Each time trace was normalized to its maximum. The codes in the figure legend stand for ‘d’ deuterated, ‘W’ water, and ‘G’ glycerol. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version of this article.)

168
L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
allowing for detection of distances up to ꢂ8 nm. If only one of the
two solvents in the mixture is deuterated, the combination H2O/
glycerol-d8 reveals a slightly more beneficial transverse relaxation
than the combination D2O/glycerol in the intermediate delay times
range (Fig. 3). This might be an indication of some differences in
solvation of the Gd(III) complexes by glycerol and by water mole-
cules. Fig. 3 shows that in both H2O/glycerol-d8 and D2O/glycerol
1
0.8
0.6
0.4
0.2
0
–
a
[Gd(DOTA)]
400 ns
3000 ns
5000 ns
*
matrices at the times longer than 4 ls, the longest feasible trans-
verse evolution times, the two time traces are very similar. In these
two cases the longest detectable distance and the signal-to-noise
ratio in a DEER experiment with Gd(III) centers would be thus
nearly the same.
1200
1220
1240
1260
1260
3
1280
1280
4
1300
B [mT]
The transverse relaxation time trace recorded in a fully proto-
nated matrix presents a characteristic curvature at the initial range
of times, more visible in the semi-logarithmic plot, suggesting a
significant impact of the nuclear spin diffusion mechanism to
transverse relaxation. When the D2O/glycerol-d8 matrix is used,
the trace assumes a ‘‘mono-exponential” character, almost linear
in the semi-logarithmic plot (Fig. 3).
1
0.8
0.6
0.4
0.2
0
b
It was suggested that at W band the relaxation of the central
transition of [Gd(DOTA)]ꢀ is dominated by the nuclear spin diffu-
sion mechanism [46]. At Q band we clearly observe a change of
the dominating relaxation mechanism. In the fully protonated
matrix, transverse relaxation is dominated by the nuclear spin dif-
fusion, whereas it seems to be dominated by the ZFS-driven, or,
perhaps, ligand hyperfine-driven mechanism in the fully deuter-
ated matrix. Importantly, at W band the transverse relaxation at
the central transition was also almost mono-exponential (linear
in the semi-logarithmic plot) [46]. Note also a slower relaxation
reported at W band [46] as compared to our Q-band data. This
might be related to the reduction of ZFS-driven relaxation contri-
bution for the central transition of Gd(III) at increased detection
frequency. These observations suggest that the assumption of the
dominating impact of nuclear spin diffusion on the transverse
relaxation of the central transition of [Gd(DOTA)]ꢀ in deuterated
matrix [46] might be an overestimation. For Gd(III) complexes with
ZFS stronger than that of [Gd(DOTA)]ꢀ the transverse relaxation of
the central transition in fully deuterated matrix might still be
ZFS-driven at W band.
As discussed by Raitsimring et al., the ZFS-driven contribution
to the relaxation of the central transition of Gd(III) should always
be weaker than for the other transitions [46]. Nonetheless, if the
main mechanism of the transverse relaxation of the central transi-
tion is not the same as for the other transitions, both faster and
slower rates might be possible. An example of such a case is
demonstrated in Fig. 4. The value R is calculated as a ratio between
the Hahn echo intensity at the maximum of Gd(III) spectrum and at
the position of the ‘kink’ (indicated by an arrow and an asterisk),
where the contribution of the central |ꢀ1/2i M |+1/2i transition
of Gd(III) is close to zero, but all other transitions do contribute
to the intensity of the electron spin echo. Note that here we don’t
make a detailed decomposition into the individual transitions,
since the qualitative picture to be drawn is clear enough without
such a data processing: first, the central transition initially con-
tributes at least 80% of the echo intensity at the maximum of the
[Gd(DOTA)]ꢀ spectrum; second, transverse relaxation of all the
other transitions is ZFS-driven (at least in a deuterated matrix
[46]) so that for qualitative conclusions they can be considered
together, indicating thus an average ZFS-driven relaxation rate.
Fig. 4a demonstrates that in a fully protonated solvent mixture
the value R is continuously decreasing with the increase of the
*
1200
1220
1240
1300
B [mT]
1
0.9
0.8
0.7
0.6
0.5
0.4
c
H2O / glycerol-d8
D2O / glycerol
0
1
2
5
d1 [s ]
Fig. 4. Dependency of the ED Q-band EPR spectra of [Gd(DOTA)]ꢀ on the solvent
deuteration. Spectra were recorded at 10 K in Hahn echo experiment with
a
different interpulse delays d1 and with lengths of both m.w. pulses of 12 ns. (a) H2O/
glycerol; (b) D2O/glycerol-d8; red: d1 = 400 ns, blue: d1 = 3000 ns, green:
d1 = 5000 ns. Spectra were normalized in order to reach the same intensity for
each spectrum of the outer transitions of Gd(III) at the point marked with the brown
arrow and asterisk (see text for details). (c) Dependency of the relative intensity of
the central transition of [Gd(DOTA)]ꢀ on the Hahn echo delay time. Spectra were
recorded at 10 K with different interpulse delays and both pulse lengths of 12 ns.
Purple circles: H2O/glycerol-d8; dark blue squares: D2O/glycerol. The value R stays
for the ratio between the intensity at the maximum of the Gd(III) ED EPR spectrum
and the intensity close to the ‘kink’ of the central transition, as indicated by brown
arrow and asterisk in (a and b). (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
transition is dominated by the spectral diffusion, and that this
contribution is stronger for the central transition than for the other
ones. Moreover, the spectral diffusion-driven relaxation of the
central transition is faster than the ZFS-driven relaxation of other
transitions. For the fully deuterated solvent mixture (Fig. 4b) the
picture is opposite and the contribution of the central transition
interpulse delay time
s
, and, from the shape of the ED EPR spec-
of the Gd(III) grows with the increase of s. Qualitatively analogous
trum at = 5 s, one can conclude that the coherence formed on
s
l
behavior was observed for DOTAM, M8DOTA and DOTP complexes.
A transition between the two regimes can be observed in the
glassy matrix having only one component (water or glycerol)
deuterated (Fig. 4c). In both D2O/glycerol and H2O/glycerol-d8
the central transition of Gd(III) has completely decayed while the
other transitions still provide a detectable echo signal. This implies
that in a protonated matrix the transverse relaxation of the central

L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
169
matrices, experiments reveal that the transverse relaxation of the
central transition is slower than for the other transitions only for
long mixing times suffer from bigger uncertainties due to the dif-
ficulty in tuning the echo phase for lower signals. In spite of this
variability, the overall trend of the DEER echo reduction effect is
always maintained so that, for example, DOTA shows always a
weaker reduction effect in fully deuterated matrices in comparison
with protonated matrices. These observations ultimately suggest
that hyperfine interactions of nuclei surrounding the complex do
play a role in the echo reduction effect.
A series of echo reduction measurements on DOTAM, M8DOTA
and DOTP also indicates a dependency of this effect on the magni-
tude of ZFS in a fully deuterated matrix. Namely, the comparison of
values in Tables 1 and 2 shows that the echo reduction is generally
weaker for Gd(III)-complexes with stronger ZFS. Following the rule,
DOTAM that has the weakest ZFS in the series at the same time has
the strongest echo reduction: only about 5% of the DEER echo
intensity is left if the pump pulse is applied. For Gd(III)–nitroxide
distance measurements with such complexes the use of modified
DEER sequence with reduced flip angle of the pump pulse would
be recommended [29].
short interpulse delays. For delay times longer than ꢂ2.5
ls, the
central transition of Gd(III) begins to decay faster than other tran-
sitions. This second decay regime is more pronounced in the D2O/
glycerol matrix (compared to H2O/glycerol-d8) thus suggesting
heterogeneous complex solvation, with glycerol residing on
average closer to the macrocycle than water. These different relax-
ation regimes of the Gd(III) central transition need to be taken into
account whenever a detailed analysis of the Gd(III) spin dynamics
is to be performed. In particular, DEER experiments with shaped
broad band microwave pulses were recently proposed that might
require correction for such transition-dependent relaxation in their
setup and analysis [58–61]. Note also that the ED EPR spectra in
the investigated four different matrices overlap perfectly at short
interpulse delays of d1 = 400 ns regardless of the pulse length
(see SI, Fig. S2).
3.3. DEER echo reduction
3.4. Effect of pH on relaxation times and ED EPR spectra of [Gd
(DOTA)]ꢀ
The Bloch–Siegert mechanism of the reduction of refocused
echo upon the application of pump pulse describes the
frequency-dependent phase shift for detected spins due to the
interference with the pump pulse [43,44]. This effect appears in
cases of partial overlap of the excitation profiles of the pump and
detection pulses in the DEER experiment. In addition to the
Bloch–Siegert mechanism, relevant for any spin, the second pro-
posed echo reduction mechanism [28] is related to the spin
dynamics of e.g. the high-spin Gd(III) system. It was suggested that
upon application of the pump pulse, single-quantum coherences of
Gd(III) could be transferred to non-detectable double-quantum
and higher coherences, which leads to an additional reduction of
the echo amplitude. Simulations of the echo reduction effect
[28,42] that took into account the Bloch–Siegert shift and the
magnetization loss to multiple-quantum coherences of the Gd(III)
high-spin, were able to partially reproduce the intensity drop of
the echo in [Gd(DTPA)]2ꢀ complexes (DTPA = diethylene triamine
pentaacetic acid) additionally suggesting that other interactions
should be considered for a full understanding of the phenomenon.
It is thus of interest to study the impact of deuteration on the echo
reduction in Gd(III) centers. Here we report echo reduction data for
different interpulse delays at 10 K at Q band. Measurements
carried out on [Gd(DOTA)]ꢀ (first four lines in Table 2) indicate
not only that deuteration plays a role but also that this effect is vir-
tually independent on the mixing time. Note that the values
reported in Table 2 must be taken only as indicative numbers.
The outcome of the DEER echo reduction measurement is sensitive
to the shape of the low-frequency wing of the microwave res-
onator dip. Absolute values of echo intensity vary thus between
different measurement sessions. In addition, values obtained for
Since interaction with surrounding protons is the dominating
mechanism of transverse relaxation of the central transition of
Gd(III) in a fully protonated matrix, we tested the effect of pH on
the relaxation and ED EPR spectrum of [Gd(DOTA)]ꢀ, based on
the assumption that solvation of Gd(III) complexes should depend
on pH. Fig. 5 shows the transverse and longitudinal relaxation time
traces of [Gd(DOTA)]ꢀ in linear and semi-logarithmic plots. One
can observe in Fig. 5 that while no significant change in longitudi-
nal relaxation takes place upon using different pH, from 5.8 to 9,
the shape of the transverse relaxation time trace reveals small
but measurable pH dependency.
At acidic pH (pH = 5.8), the transverse relaxation shows largest
non-linearity in the semi-logarithmic plot and can thus be consid-
ered to contain strongest contribution from spectral diffusion. At
neutral (pH = 7) and basic (pH = 9) conditions the transverse relax-
ation traces are more linear in the semi-logarithmic coordinates,
however not yet very close to
behavior.
a mono-exponential (linear)
The ED EPR spectrum of [Gd(DOTA)]ꢀ did not reveal any pH
dependence. Concerning the relatively slow kinetics of the [Gd
(DOTA)]ꢀ complex formation [62,63] no changes in the ED EPR
spectrum could be observed for incubation times up to 12 h.
3.5. Comparison of relaxation times of different complexes
Finally, a comparison of relaxation properties of the whole ser-
ies of macrocyclic complexes was performed in D2O/glycerol-d8.
Corresponding transverse and longitudinal relaxation traces are
shown in Fig. 6(a and b) and (c and d) in linear and semi-
logarithmic representation. Differences in longitudinal relaxation
within this series are basically negligible. This supports the
assumption that the longitudinal relaxation of these complexes is
predominantly due to the orbit lattice coupling and is thus very lit-
tle affected by the strength of ZFS interaction or by the hyperfine
interaction to surrounding magnetic nuclei.
The series of transverse relaxation measurements reveals more
differences between different complexes. There is no apparent cor-
relation between the strength of the ZFS interaction and the rate of
Gd(III) transverse relaxation (Fig. 6(c and d)). In particular, the two
stereoisomers of [Gd(M8DOTA)] have quite different ZFS (most
probably, due to the different strength of the symmetry distortion)
but, nevertheless, transverse relaxation is virtually the same in
both cases (Fig. 7). The [Gd(M8DOTA)] and [Gd(DOTAM)]3+
Table 2
Echo reduction effect for the investigated samples at different interpulse delays d2.
Echo reduction is expressed in % ratio between the intensity of the echo with and
without applied pump pulse.
Matrix
Chelating agent
Interpulse delay time
1.2
ls
3
l
s
4
9
11
11
18
n.d.
n.d.
5
ls
H2O/glycerol
DOTA
9
12
13
17
19
15
5
6
H2O/glycerol-d8
D2O/glycerol
DOTA
DOTA
11
14
18
20
14
6
D2O/glycerol-d8
D2O/glycerol-d8
D2O/glycerol-d8
D2O/glycerol-d8
D2O/glycerol-d8
D2O/glycerol-d8
DOTA
M8DOTA(8S)SPy
M8DOTA(4S4R)SPy
DOTAM
DOTP
35
24
28
24
28
n.d.
H2O

170
L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
0
[Gd(DOTA)] –
a
b
0.8
0.6
0.4
0.2
0
-2
-4
pH 5.8
pH 7
pH 9
0
1
2
3
4
5
6
7
0
1
2
3
4
5
t [ s]
t [ s]
0
-2
-4
-6
1
0.5
0
c
d
-0.5
0
2
4
6
8
0
1
2
3
4
5
6
t [ms]
t [ms]
–
Fig. 5. pH dependency of the transverse (a and b) and longitudinal (c and d) relaxation times traces of [Gd(DOTA)]ꢀ in linear (a and c) and semilogarithmic (b and d)
representation. Measurements were recorded at 10 K at Q band at different pH: red (pH 5.8), blue (pH 7) and green (pH 9). Buffer solution was prepared in H2O and sample
was diluted with glycerol. Each time trace was normalized to its maximum. (For interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article.)
0
DOTA
DOTAM
DOTP
M8DOTA(4R4S)SPy
H2O
a
b
0.8
0.6
0.4
0.2
0
-2
-4
-6
-8
-10
0
5
10
15
t [ s]
20
25
0
5
10
t [ s]
15
20
0
-2
-4
-6
1
0.5
0
c
d
-0.5
0
2
4
6
8
0
1
2
3
4
5
6
t [ms]
t [ms]
Fig. 6. Transverse (a and b) and longitudinal (c and d) relaxation times traces of Gd(III)-complexes in the linear (a and c) and semilogarithmic (b and d) representation.
Measurements were recorded at 10 K at pH 5.8 and in a fully deuterated matrix (D2O/glycerol-d8) for: [Gd(DOTA)]ꢀ (red), [Gd(DOTP)]5ꢀ (blue), [Gd(DOTAM)]3+ (green), [Gd
(M8DOTA(4R4S)SPy)] (light blue), and [Gd(III)]aq (black). Each time trace was normalized to its maximum. (For interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)
complexes exhibit faster relaxation than other complexes, whereas
the transverse relaxation of the water coordinated Gd(III)
(abbreviated as [Gd(III)]aq), [Gd(DOTA)]ꢀ and [Gd(DOTP)]5ꢀ is
approximately the same with slight advantage for [Gd(DOTP)]5ꢀ
at long times.
The particularly fast transverse relaxation of [Gd(M8DOTA)]
should be, apparently, attributed to the presence of the eight
methyl groups with 24 protons. At 10 K the rotational motion of
these groups is frozen, but the proton tunneling is a known impor-
tant source of relaxation [64]. Importantly, at long mixing times

L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
171
including the arrangement of solvation molecules. Furthermore,
one water molecule would have a possibility of directly coordinat-
ing to Gd(III) in such a complex thus providing possibility of further
components in the relaxation trace. The transverse relaxation of the
[Gd(M8DOTA)] complexes is identical in MES buffer (200 mM) and
phosphate buffer (20 mM), indicating rather small effect of 31P
nuclei distributed at millimolar concentration in the glassy matrix
on the relaxation of Gd(III). Only a very small difference in trans-
verse relaxation was observed for [Gd(M8DOTA(8S)SPy)] with con-
a
[Gd(M8DOTA)]
0.8
0.6
0.4
0.2
0
8S
4R4S
centrations of 50 lM and 300 lM (Fig. 7b), thus indicating almost
no electron spin diffusion contribution.
0
5
10
15
20
25
t [ s]
4. Discussion and conclusions
The qualitative picture presented in this work demonstrates
that despite being weak as compared to the ZFS interaction the
hyperfine couplings to surrounding nuclei play an important,
sometimes critical role for the spectroscopic properties of Gd(III)
chelate complexes. This primarily concerns the central transition
of Gd(III), which is most important for PDS measurements, and
for which the contribution of ZFS to the line width is strongly
reduced compared to other transitions. Fortunately for the bio-
chemical applications of Gd(III)-based spin labels, our results
demonstrate that Gd(III) relaxation properties are rather weakly
changing with pH or buffer type, thus making the labels broadly
applicable and reliable. The effect of solvent deuteration on trans-
verse relaxation of Gd(III) is nearly as strong as it is for organic rad-
icals [15,47]. Importantly, the main mechanism responsible for the
transverse relaxation of the central transition of Gd(III) seems to
switch between protonated and deuterated matrices, being mostly
driven by nuclear spin diffusion in the former case. In the latter
case it is difficult to discriminate any mechanism, based on the
qualitative considerations we reported here. However, strong
change of the transverse relaxation of [Gd(M8DOTA)] as compared
to other complexes, as well as nearly no change for the two [Gd
(M8DOTA)] stereoisomers with rather different ZFS suggest that
hyperfine interaction with magnetic nuclei of the chelator mole-
cule might play a significant role in the transverse relaxation of
Gd(III) centers. Furthermore, the example of DOTP demonstrates
that the directly coordinated as well as the second solvation sphere
molecules can influence relaxation properties and ZFS of Gd(III)
centers. The echo reduction in Gd(III)–nitroxide DEER measure-
ments also demonstrates a clear dependence on hyperfine interac-
tions with solvent molecules.
The relative rate of transverse relaxation for the central transi-
tion of Gd(III) is important whenever details of spin dynamics of
such paramagnetic centers are considered. In such studies one
has to take into account the different regimes shown in Fig. 4. Fur-
thermore, as it is apparent from Fig. 6, the difference between
transverse and longitudinal relaxation rates for Gd(III) chelate
complexes in fully deuterated matrices can be less than two orders
of magnitude. In such cases, e.g. in very long distance measure-
ments with Gd(III)–Gd(III) DEER, the longitudinal relaxation might
need to be considered for proper data analysis.
Regarding the optimization of PDS experiments, one conclusion
of this work is that magnetic nuclei should be avoided whenever
possible when designing the new chelating molecules. Our study,
thus, suggests that the Gd(III) tags optimized for MRI that allow
for one or more directly coordinated water molecules might be
sub-optimal for the PDS measurements, and that tags that leave
no possibility of directly coordinated solvent molecules might per-
form better in terms of transverse relaxation of Gd(III). Concerning
the Gd(III)–nitroxide DEER experiments, our study confirms previ-
ous theoretical predictions that DEER echo reduction in such
experiment is ZFS dependent. The increase of the DEER echo
b
0.8
0.6
0.4
0.2
0
(8S) 300 uM MES
(8S) 300 uM PO4
(8S) 50 uM MES
0
5
10
15
20
25
t [ s]
Fig. 7. Comparison of transverse relaxation properties for stereoisomers [Gd
(M8DOTA(8S)SPy)] and [Gd(M8DOTA(4R4S)SPy)] (a), and comparison of the
transverse relaxation properties of [Gd(M8DOTA(8S)SPy)] in MES and phosphate
buffer at the concentrations of the chelate complexes of 50
Magenta: 300 [Gd(M8DOTA(8S)SPy)] in MES buffer; black: 300
(M8DOTA(8S)SPy)] in phosphate buffer; green: 50 M [Gd(M8DOTA(8S)SPy)] in
MES buffer; light blue: 300 M [Gd(M8DOTA(4R4S)SPy)] in MES buffer. Measure-
l
M and 300
lM (b).
lM
lM [Gd
l
l
ments were performed at 10 K. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
the transverse relaxation of [Gd(M8DOTA)] is not so much slower
than for other complexes, and the maximum achievable length of
the e.g. DEER trace would be only ꢂ2
ls shorter. The complexes
of M8DOTA with lanthanide ions are particularly favorable in para-
magnetic NMR experiments [49]. Therefore, despite somewhat
worse relaxation properties in pulse EPR, these complexes might
appear particularly well performing for the combined NMR/EPR
approaches.
One can further speculate that somewhat faster relaxation of
the [Gd(DOTAM)]3+ is connected to the presence of the extra four
14N nuclei in addition to the nitrogen quartet in the tetraazacy-
clododecane ring. However, generalizing this to a statement that
magnetic nuclei in the chelator molecule lead to faster transverse
relaxation might not be possible, since, for instance, the presence
of four phosphate groups in DOTP does not increase the relaxation,
despite higher magnetic moment of 31P as compared to the 14N (in
DOTAM). It might be thus possible that the change of transverse
relaxation is not always directly induced by the magnetic nuclei
within the chelator molecule, but might also be due to the solvent
molecules arrangement around the particular chelate complex.
Unlike DOTAM and DOTA, it is known that [Gd(DOTP)]5ꢀ forms a
strong framework of hydrogen bonds in its second solvation sphere
and that it lacks a H2O directly bound to the metal center [65].
These two features might be beneficial in the design of chelators
with particularly slow Gd(III) relaxation times.
The transverse relaxation traces of [Gd(M8DOTA)] possess two
kinks and thus consist of at least three distinct components
(Fig. 7). It is reasonable to attribute different relaxation components
to the few possible conformations of [Gd(M8DOTA)] [49], possibly

172
L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
[18] O. Duss, E. Michel, M. Yulikov, M. Schubert, G. Jeschke, F.H.-T. Allain, Structural
basis of the non-coding RNA RsmZ acting as a protein sponge, Nature 509
(2014) 588–592.
[19] R. Ward, M. Zoltner, L. Beer, H. El Mkami, I.R. Henderson, T. Palmer, D.G.
Norman, The orientation of a tandem POTRA domain pair, of the beta-barrel
assembly protein BamA, determined by PELDOR spectroscopy, Structure 17
(2009) 1187–1194.
[20] I. Smirnova, V. Kasho, J.-Y. Choe, C. Altenbach, W.L. Hubbell, H.R. Kaback, Sugar
binding induces an outward facing conformation of LacY, Proc. Natl. Acad. Sci.
USA 104 (2007) 16504–16509.
[21] X. Han, J.H. Bushweller, D.S. Cafiso, L.K. Tamm, Membrane structure and
fusion-triggering conformational change of the fusion domain from influenza
hemagglutinin, Nat. Struct. Biol. 8 (2001) 715–720.
reduction with decreasing the strength of ZFS interaction counter-
acts the sensitivity improvement due to narrowing of the central
transition of Gd(III). Thus, in the Gd(III)–nitroxide DEER the use
of Gd(III) complexes with smaller ZFS is expected to bring signifi-
cantly less improvement in the signal-to-noise ratio, as compared
to Gd(III)–Gd(III) DEER. However, sample deuteration appears to
bring an additional advantage for this experiment as DEER echo
reduction is measurably weaker for deuterated samples.
Overall, we believe that this work provides a useful overview of
the spectroscopic properties of Gd(III) centers based on tetraazacy-
clododecane derivatives, with respect to their applications in PDS.
[22] S. Bleicken, G. Jeschke, C. Stegmueller, R. Salvador-Gallego, A.J. García-Sáez, E.
Bordignon, Structural model of active Bax at the membrane, Mol. Cell 56
(2014) 496–505.
Acknowledgments
[23] I. Hanelt, D. Wunnicke, E. Bordignon, H.J. Steinhoff, D.J. Slotboom,
Conformational heterogeneity of the aspartate transporter Glt(Ph), Nat.
Struct. Biol. 20 (2013) 210–214.
[24] E.R. Georgieva, P. Borbat, C. Ginter, J.H. Freed, O. Boudker, Conformational
ensemble of the sodium-coupled aspartate transporter, Nat. Struct. Biol. 20
(2013) 215–221.
[25] A.M. Raitsimring, C. Gunanathan, A. Potapov, I. Efremenko, J.M.L. Martin, D.
Milstein, D. Goldfarb, Gd3+ complexes as potential spin labels for high field
pulsed EPR distance measurements, J. Am. Chem. Soc. 129 (2007)
14138–14139.
We wish to thank Dr. Roche Walliser (Uni Basel) for help in the
synthesis of M8DOTA(4R4S)Spy. The Swiss National Science
Foundation is acknowledged for a Grant No. SNF 200021_130263
(to D.H.) and
a Grant No. SNF 200020_14441 (L.G., M.Y.).
Prof. Dr. Gunnar Jeschke, the head of the EPR group at the Labora-
tory of Physical Chemistry at ETH Zurich, is acknowledged for the
support of this project, numerous discussions, and critical reading
of the manuscript.
[26] I. Kaminker, I. Tkach, N. Manukovsky, T. Huber, H. Yagi, G. Otting, M. Bennati,
D. Goldfarb, W-band orientation selective DEER measurements on a Gd3+
/
nitroxide mixed-labeled protein dimer with a dual mode cavity, J. Magn.
Reson. 227 (2013) 66–71.
[27] P. Lueders, G. Jeschke, M. Yulikov, Double electron–electron resonance
measured between Gd3+ ions and nitroxide radicals, J. Phys. Chem. Lett. 2
(2011) 604–609.
Appendix A. Supplementary material
[28] M. Yulikov, P. Lueders, M.F. Warsi, V. Chechik, J. Jeschke, Distance
measurements in Au nanoparticles functionalized with nitroxide radicals
and Gd3+–DTPA chelate complexes, Phys. Chem. Chem. Phys. 14 (2012)
10732–10746.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jmr.2015.08.009.
[29] L. Garbuio, E. Bordignon, E.K. Brooks, W.L. Hubbell, G. Jeschke, M. Yulikov,
Orthogonal spin labeling and Gd(III)–nitroxide distance measurements on
bacteriophage T4-lysozyme, J. Phys. Chem. B 117 (2013) 3145–3153.
[30] A. Martorana, G. Bellapadrona, A. Feintuch, E. Di Gregorio, S. Aime, D. Goldfarb,
Probing protein conformation in cells by EPR distance measurements using
Gd3+ spin labeling, J. Am. Chem. Soc. 136 (2014) 13458–13465.
[31] C.D. Barry, J.A. Glasel, R.J.P. Williams, A.V. Xavier, Quantitative determination
of conformations of flexible molecules in solution using lanthanide ions as
nuclear magnetic resonance probes: application to adenosine-50-
monophosphate, J. Mol. Biol. 84 (1974) 471–490.
[32] P. Caravan, J.J. Ellison, T.J. McMurry, R.B. Lauffer, Gadolinium(III) chelates as
MRI contrast agents: structure, dynamics, and applications, Chem. Rev. 99
(1999) 2293–2352.
[33] E. Boros, E.M. Gale, P. Caravan, MR imaging probes: design and applications,
Dalton Trans. 44 (2015) 4804–4818.
[34] A.M. Raitsimring, A.V. Astashkin, P. Caravan, High-frequency EPR and ENDOR
characterization of MRI contrast agent. Biological magnetic resonance, in: L.
Berliner, G. Hanson (Eds.), High Resolution EPR: Applications to
Metalloenzymes and Metals in Medicine, vol. 28, Springer, New York, 2009,
pp. 581–621.
[35] A.M. Raitsimring, A.V. Astashkin, O.G. Poluektov, P. Caravan, High-field pulsed
EPR and ENDOR of Gd3+ complexes in glassy solutions, Appl. Magn. Reson. 28
(2005) 281–295.
References
[1] A.D. Milov, K.M. Salikhov, M.D. Shirov, Use of the double resonance in electron
spin echo method for the study of paramagnetic center spacial distribution in
solids, Fiz. Tverd. Tela (Leningrad) 23 (1981) 975–982.
[2] A.D. Milov, A.B. Pomarev, Y.D. Tsvetkov, Electron electron double resonance in
electron-spin echo-model biradical systems and the sensitized photolysis of
decalin, Chem. Phys. Lett. 110 (1984) 67–72.
[3] R.E. Martin, M. Pannier, F. Diederich, V. Gramlich, M. Hubrich, H.W. Spiess,
Determination 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. Engl. 37 (1998) 2834–2837.
[4] M. Pannier, S. Veit, A. Godt, G. Jeschke, H.W. Spiess, Dead-time free
measurements of dipole–dipole interactions between electron spins, J. Magn.
Reson. 142 (2000) 331–340.
[5] L.V. Kulik, S.A. Dzuba, I.A. Grigoryev, Y.D. Tsvetkov, Electron dipole–dipole
interaction in ESEEM of nitroxide biradicals, Chem. Phys. Lett. 343 (2001)
315–324.
[6] S. Milikisyants, F. Scarpelli, M.G. Finiguerra, M. Ubbink, M.A. Huber, A pulsed
EPR method to determine distances between paramagnetic centers with
strong spectral anisotropy and radicals: the dead-time free RIDME sequence, J.
Magn. Reson. 201 (2009) 48–56.
[7] S. Razzaghi, M. Qi, A.I. Nalepa, A. Godt, G. Jeschke, A. Savitsky, M. Yulikov,
[36] M. Benmelouka, J. Van Tol, A. Borel, M. Port, L. Helm, L.C. Brunel, A.E.
RIDME spectroscopy with Gd(III) centers, J. Phys. Chem. Lett.
3970–3975.
[8] P.P. Borbat, J.H. Freed, Multiple-quantum ESR and distance measurements,
Chem. Phys. Lett. 313 (1999) 145–154.
[9] G. Jeschke, M. Pannier, A. Godt, H.W. Spiess, Dipolar spectroscopy and spin
alignment in electron paramagnetic resonance, Chem. Phys. Lett. 331 (2000)
243–252.
[10] C. Altenbach, T. Marti, H.G. Khorana, W.L. Hubbell, Transmembrane protein
structure: spin labeling of bacteriorhodopsin mutants, Science 248 (1990)
1088–1092.
[11] J.P. Klare, H.-J. Steinhoff, Spin labeling EPR, Photosynth. Res. 102 (2009)
377–390.
[12] E. Bordignon, Site-directed spin labeling of membrane proteins, Curr. Top.
Chem. 321 (2012) 121–157.
[13] W.L. Hubbell, D.S. Cafiso, C. Altenbach, Identifying conformational changes
with site-directed spin labeling, Nat. Struct. Biol. 7 (2000) 735–739.
[14] G. Jeschke, Y. Polyhach, Distance measurements on spin-labelled
biomacromolecules by pulsed electron paramagnetic resonance, Phys. Chem.
Chem. Phys. 9 (2007) 1895–1910.
[15] R. Ward, A. Bowman, E. Sozudogru, H. El-Mkami, T. Owen-Hughes, D.G.
Norman, EPR distance measurements in deuterated proteins, J. Magn. Reson.
207 (2010) 164–167.
5 (2014)
Merbach,
A high-frequency EPR study of frozen solutions of Gd-III
complexes: straightforward determination of the zero-field splitting
parameters and simulation of the NMRD profiles, J. Am. Chem. Soc. 128
(2006) 7807–7816.
[37] A. Hudson, J.W.E. Lewis, Electron spin relaxation of S-8 ions in solution, Trans.
Faraday Soc. 66 (1971) 1297–1301.
[38] S. Rast, P.H. Fries, E. Belorizky, Static zero field splitting effects on the
electronic relaxation of paramagnetic metal ion complexes in solution, J.
Chem. Phys. 113 (2000) 8724–8735.
[39] M. Benmelouka, A. Borel, L. Moriggi, L. Helm, A.E. Merbach, Design of Gd(III)-
based magnetic resonance imaging contrast agents: Static and transient zero-
field splitting contributions to the electronic relaxation and their impact on
relaxivity, J. Phys. Chem. B 111 (2007) 832–840.
[40] A. Borel, J.F. Bean, R.B. Clarkson, L. Helm, L. Moriggi, A.D. Sherry, M. Woods,
Towards the rational design of MRI contrast agents: electron spin relaxation is
largely unaffected by the coordination geometry of Gadolinium(III)–DOTA-
type complexes, Chem. Eur. J. 14 (2008) 2658–2667.
[41] A.V. Astashkin, A.M. Raitsimring, Electron spin echo envelope modulation
theory for high electron spin systems in weak crystal field, J. Chem. Phys. 117
(2002) 6121–6132.
[42] P. Lueders, H. Jager, M.A. Hemminga, G. Jeschke, M. Yulikov, Distance
measurements on orthogonally spin-labeled membrane spanning WALP23
polypeptides, J. Phys. Chem. B 117 (2013) 2061–2068.
[16] O. Schiemann, T.F. Prisner, Long-range distance determination in
biomacromolecules by EPR spectroscopy, Q. Rev. Biophys. 40 (2007) 1–53.
[17] G. Jeschke, DEER distance measurements on proteins, Ann. Rev. Phys. Chem. 63
(2012) 1–28.
[43] F. Bloch, A. Siegert, Magnetic resonance for nonrotating fields, Phys. Rev. 57
(1940) 522–527.

L. Garbuio et al. / Journal of Magnetic Resonance 259 (2015) 163–173
173
[44] M.K. Bowman, A.G. Maryasov, Dynamic phase shifts in nanoscale distance
measurements by double electron electron resonance (DEER), J. Magn. Reson.
185 (2007) 270–282.
35 GHz pulse ENDOR probehead accommodating large sample sizes:
performances and applications, J. Magn. Reson. 200 (2009) 81–87.
[55] S. Stoll, A. Schweiger, EasySpin, a comprehensive software package for spectral
simulation and analysis in EPR, J. Magn. Reson. 178 (2006) 42–55.
[56] A. Volkov, C. Dockter, T. Bund, H. Paulsen, G. Jeschke, Pulsed EPR determination
of water accessibility to spin-labeled amino acid residues in LHCIIb, Biophys. J.
96 (2009) 1124–1141.
[57] D. Leporini, V. Schadler, U. Wiesner, H.W. Spiess, G. Jeschke, Electron spin
relaxation due to small-angle motion: theory for the canonical orientations
and application to hierarchic cage dynamics in ionomers, J. Chem. Phys. 119
(2003) 11829–11846.
[45] I. Kaminker, H. Yagi, T. Huber, A. Feintuch, G. Otting, D. Goldfarb, Spectroscopic
selection of distance measurements in a protein dimer with mixed nitroxide
and Gd3+ spin labels, Phys. Chem. Chem. Phys. 14 (2012) 4355–4358.
[46] A. Raitsimring, A. Dalaloyan, A. Collauto, A. Feintuch, T. Meade, D. Goldfarb,
Zero field splitting fluctuations induced phase relaxation of Gd3+ in frozen
solutions at cryogenic temperatures, J. Magn. Reson. 248 (2014) 71–80.
[47] M. Huber, M. Lindgren, P. Hammarstrom, L.-G. Martensson, U. Carlsson, G.R.
Eaton, S.S. Eaton, Phase memory relaxation times of spin labels in human
carbonic anhydrase II: Pulsed EPR to determine spin label location, Biophys.
Chem. 94 (2001) 245–256.
[58] P. Schoeps, P.E. Spindler, A. Marko, T.F. Prisner, Broadband spin echoes and
broadband SIFTER in EPR, J. Magn. Reson. 250 (2015) 55–62.
[48] D. Goldfarb, Gd3+ spin labeling for distance measurements by pulse EPR
spectroscopy, Phys. Chem. Chem. Phys. 16 (2014) 9685–9699.
[49] D. Häussinger, J. Huang, S. Grzesiek, DOTA-M8: an extremely rigid, high-
affinity lanthanide chelating tag for PCS NMR spectroscopy, J. Am. Chem. Soc.
131 (2009) 14761–14767.
[59] A. Doll, S. Pribitzer, R. Tschaggelar, G. Jeschke, Adiabatic and fast passage ultra-
wideband inversion in pulsed EPR, J. Magn. Reson. 230 (2013) 27–39.
[60] P.E. Spindler, S.J. Glaser, T.E. Skinner, T.F. Prisner, Broadband inversion PELDOR
spectroscopy with partially adiabatic shaped pulses, Angew. Chem., Int. Ed. 52
(2013) 3425–3429.
[50] Y. Song, T.J. Meade, A.V. Astashkin, E.L. Klein, J.H. Enemark, A. Raitsimring,
Pulsed dipolar spectroscopy distance measurements in biomacromolecules
labeled with Gd(III) markers, J. Magn. Reson. 210 (2011) 59–68.
[51] B. Joseph, V.M. Korkhov, M. Yulikov, G. Jeschke, E. Bordignon, Conformational
cycle of the vitamin B-12 ABC Importer in liposomes detected by double
electron–electron resonance (DEER), J. Biol. Chem. 289 (2014) 3176–3185.
[52] A. Barge, G. Cravotto, E. Gianolio, F. Fedeli, How to determine free Gd and free
[61] A. Doll, G. Jeschke, Fourier-transform electron spin resonance with bandwidth-
compensated chirp pulses, J. Magn. Reson. 246 (2014) 18–26.
[62] A. Bianchi, L. Calabi, F. Corana, S. Fontana, P. Losi, A. Maiocchi, L. Paleari, B.
Valtancoli, Thermodynamic and structural properties of Gd(III) complexes
with polyamino-polycarboxylic ligands: basic compounds for the
development of MRI contrast agents, Coord. Chem. Rev. 204 (2000) 309–393.
[63] A.D. Sherry, P. Caravan, R.E. Lenkinski, Primer on gadolinium chemistry, J.
Magn. Reson. Imaging 30 (2009) 1240–1248.
[64] A. Zecevic, G.R. Eaton, S.S. Eaton, M. Lindgren, Dephasing of electron spin
echoes for nitroxyl radicals in glassy solvents by non-methyl and methyl
protons, Mol. Phys. 95 (1998) 1255–1263.
ligand in solution of Gd chelates.
Imaging 8 (2006) 184–188.
A technical note, Contrast Media Mol.
[53] I. Gromov, J. Shane, R. Forrer, R. Rakhmatoullin, Y. Rozentzwaig, A. Schweiger,
Q-band pulse EPR/ENDOR spectrometer and the implementation of
A
advanced one- and two-dimensional pulse EPR methodology, J. Magn. Reson.
149 (2001) 196–203.
[54] R. Tschaggelar, B. Kasumaj, M.G. Santangelo, J. Forrer, H. Leger, F. Dube, F.
Diederich, J. Harmer, I. Schumann, I. Garcia-Rubio, G. Jeschke, Cryogenic
[65] S. Aime, M. Botta, D. Parker, J.A.G. Williams, Extent of hydration of octadentate
lanthanide complexes incorporating phosphinate donors: solution
relaxometry and luminescence studies, J. Chem. Soc., Dalton Trans. (1996)
17–23.