Tyrosine-targeted spin labeling and EPR spectroscopy: An alternative strategy for studying structural transitions in proteins

Article abstract of DOI:10.1002/anie.201102539

Keeping tabs on tyrosine: A three-component Mannich-type reaction extends the scope of site-directed spin labeling by selectively labeling the unique tyrosine residue of CP12 protein (see picture), as was confirmed by mass spectrometry. EPR spectroscopy of the labeled protein showed a very high mobility of the probe, which remained very mobile after complex formation with GAPDH. Copyright

Full text of DOI:10.1002/anie.201102539

Communications
DOI: 10.1002/anie.201102539
Protein Labeling
Tyrosine-Targeted Spin Labeling and EPR Spectroscopy: An
Alternative Strategy for Studying Structural Transitions in Proteins**
Magali Lorenzi, Carine Puppo, Rꢀgine Lebrun, Sabrina Lignon, Valꢀrie Roubaud,
Marlꢁne Martinho, Elisabetta Mileo, Paul Tordo, Sylvain R. A. Marque,* Brigitte Gontero,
Bruno Guigliarelli, and Valꢀrie Belle*
Site-directed spin labeling (SDSL) combined with electron
paramagnetic resonance (EPR) spectroscopy is a powerful
technique for studying structural transitions in proteins,[1]
especially flexible or disordered proteins.[2] The conventional
use of this technique is based on insertion of a paramagnetic
label (nitroxide) at a cysteine residue, most often introduced
by site-directed mutagenesis. However, although cysteine
residues are rare in proteins, they frequently have functional
roles and are involved in structural elements such as disulfide
bridges or in the binding of metal cofactors.[3]
Such a difficulty was recently encountered in the study of
a small and flexible chloroplast protein CP12 from the green
alga Chlamydomonas reinhardtii by spin-labeling EPR spec-
troscopy. In this organism, CP12 contains four cysteine
residues involved in two disulfide bridges in its oxidized
state. Although the introduction of spin labels at the two
cysteine residues of the C-terminal disulfide bridge enabled
us to identify a new role of the partner protein glyceraldehyde
3-phosphate dehydrogenase (GAPDH), it precluded the
direct study of the complex formation GAPDH/CP12.[4] To
overcome this difficulty, grafting of the nitroxide probe to
residues other than cysteines is required. One strategy using a
genetically encoded unnatural amino acid has been recently
proposed.[5] The incorporation of such unnatural amino acids
relies, however, on a rather complex strategy involving an
orthogonal tRNA/aminoacyl-tRNA synthetase pair specific
for the unnatural amino acid. As such a strategy is difficult to
set up, we propose an alternative method consisting of
selectively targeting residues other than cysteines with a
nitroxide probe.
Bioconjugation of small molecules to protein residues is
very challenging, and several reactions have recently been
proposed to target specific residues selectively.[6] In particular,
efforts have been paid to modify the aromatic amino acid side
chains of tryptophan[7] and tyrosine.[8] Among them, a three-
component Mannich-type reaction has been developed that
allows the modification of tyrosine under mild, biocompat-
ible, and metal-free conditions.[8a] Inspired by these recent
studies, we present the selective grafting of a nitroxide probe
to tyrosine by using the Mannich-type reaction on CP12, a
protein bearing only one natural tyrosine residue. This unique
tyrosine residue, located at position 78 in the sequence of a
total of 80 amino acids, makes this protein an ideal candidate
for demonstrating the feasibility of tyrosine-targeted spin
labeling (Figure 1).
[*] M. Lorenzi, C. Puppo, Dr. M. Martinho, Dr. B. Gontero,
Prof. B. Guigliarelli, Dr. V. Belle
Bioꢀnergꢀtique et Ingꢀnierie des Protꢀines UPR 9036, CNRS
and
Aix-Marseille Universitꢀ
Institut de Microbiologie de la Mꢀditꢀrranꢀe
31 chemin J. Aiguier, 13402 Marseille Cedex 20 (France)
E-mail: belle@ifr88.cnrs-mrs.fr
Dr. R. Lebrun, S. Lignon
Plateforme Protꢀomique, IFR88
IBiSA Marseille-Protꢀomique, CNRS
Institut de Microbiologie de la Mꢀditꢀrranꢀe
13402 Marseille Cedex 20 (France)
The three-component Mannich-type reaction was per-
Dr. V. Roubaud, Dr. E. Mileo, Prof. P. Tordo, Prof. S. R. A. Marque
Laboratoire Chimie Provence, LCP-UMR 6264
Universitꢀ de Provence
formed as described by McFarland et al.[8a] on a mixture of
case 521, Avenue Escadrille Normandie-Niemen
13397 Marseille Cedex 20 (France)
E-mail: sylvain.marque@univ-provence.fr
[**] This work was supported by the Agence Nationale de la Recherche
ANR SPINFOLD no. 09-BLAN-0100, the Centre National de la
Recherche Scientifique (CNRS) and Aix-Marseille University. M.L. is
grateful to the French Minister of Research for her PhD fellowship.
We thank M. Joint and M. Belghazi for help in the MS and MS/MS
data collection and analyses on the Ultraflex MALDI-ToFToF, IFR
Jean-Roche, MaP. We thank C. Chendo and V. Monnier from the
Spectropole of Aix-Marseille University for the ESI-MS study. We
thank E. Etienne for the simulation of the EPR spectra and A.
Cornish Bowden for revising the English manuscript.
Supporting information for this article, including experimental
procedures, experimental conditions for EPR and circular dichroism
spectroscopy, and mass spectrometry analyses, is available on the
WWW under http://dx.doi.org/10.1002/anie.201102539.
Figure 1. Modeled structure of CP12 in its oxidized state, indicating
the position of the tyrosine residue and the two disulfide bridges
(from Ref. [9]).
9108
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9108 –9111

130 mm CP12 in 10 mm phosphate buffer at pH 6.4 with
formaldehyde and 4-amino-2,2,5,5-tetramethyl-3-imidazo-
line-1-yloxy nitroxide in 1250-fold and 10-fold molar excess,
respectively (Scheme 1). The mixture was stirred for 16 h at
room temperature in air. The excess of unreacted spin label
was removed by gel filtration (PD10 column, GE Health-
care). The elution fractions were analyzed both by measuring
the protein concentration (using the Bio-Rad reagent protein
assay)[10] and by calculating the spin concentration from the
double integral of EPR spectroscopy signals, yielding
70(10)% labeling.
Figure 2. MALDI-ToF mass spectra obtained after tryptic digestion of
the reduced and alkylated samples of unlabeled (gray) and labeled
CP12 (black). Focus is made in the range of 3000–3320 Da and in the
inset in the range of 3282–3310 Da to show the different ions
corresponding to the C-terminal peptide of 27 amino acids. IAA
corresponds to iodoacetamide and means that the cysteine residue is
carbamidomethylated.
Scheme 1. Three-component Mannich-type reaction on tyrosine resi-
dues with the two possible products A and B.
To check that the experimental conditions of the Man-
nich-type reaction and the presence of the grafted label did
not alter the structural and biochemical characteristics of the
CP12, control experiments were performed. The global
structure of the labeled protein was checked by circular
dichroism studies. The CD spectra of the modified CP12
exhibited similar features as the unmodified one in its
oxidized state: it revealed the presence of some a-helical
secondary structure elements (Figure 3a), and the use of
To confirm the site selectivity of the reaction on CP12, the
cysteine residues of the labeled protein and a control sample
of unlabeled CP12 were reduced by dithiothreitol and
alkylated by iodoacetamide (mass increment of 57 Da per
cysteine residue), then digested with trypsin. Analysis of the
samples by MALDI-ToF MS indicated the presence of
different peptides (see the Supporting Information), one of
which bore the expected mass increment for the grafted spin
probe (180 Da) corresponding to product B (Scheme 1). This
peptide corresponded to the C-terminal 27 amino acids
containing the tyrosine residue, a sequence confirmed by MS/
MS data obtained after fragmentation of the intense ion
precursor at m/z 2999.4 (see the Supporting Information).
Reduction and alkylation reactions were not complete, and
the same C-terminal peptide bearing the spin label was
observed as mono- and dicarbamidomethylated forms of
cysteines C66 and C75 (Figure 2).
The possibility that the spin probe could be grafted to
cysteine residues was rejected, as they do not participate in
the reaction except as reduced disulfides,[8a] a requirement
that was not satisfied under our experimental conditions
during the labeling reaction. Moreover, the detection of the
labeled dicarbamidomethylated form of CP12 reinforced this
finding. It is known that after a long reaction time, the
Mannich-type reaction can modify tryptophan residues.[8a]
Modification of the unique tryptophan residue at position
35 in CP12 was not detected, however, under our experimen-
tal conditions. From these analyses we concluded that the spin
probe was grafted on the unique tyrosine residue of CP12. A
mass increment of 168 Da corresponding to the product A
(Scheme 1) was not found on the peptide mass fingerprint of
the labeled CP12. This observation was confirmed by electro-
spray MS analyses of an esterified tyrosine free amino acid
modified by the nitroxide probe, which showed that B was the
only product of the reaction (see the Supporting Informa-
tion).
Figure 3. a) CD spectra of the oxidized CP12 and the labeled CP12.
b) Western blot analyses of the in vitro reconstitution of the GAPDH/
CP12 complex (CP12* means labeled CP12).
trifluoroethanol as secondary-element stabilizer induced the
same increase of the a-helical content (see the Supporting
Information).[4] Moreover, this result indicated that the
disulfide bridges are present in the labeled CP12, as it is
known that the reduction of these bridges leads to a CD
spectrum typical of a disordered protein.[11] The ability of
modified CP12 to form a stable complex with GAPDH was
checked by an in vitro reconstitution test (Figure 3b).[12]
Altogether these experiments showed that, even using a
high concentration of formaldehyde and a long incubation
time at room temperature, modified CP12 kept its biochem-
ical and structural characteristics.
Angew. Chem. Int. Ed. 2011, 50, 9108 –9111
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 9109

Communications
When grafted onto proteins, nitroxide spin labels give rise
from 0.125 to 0.225 ns (Table 1 and Figure 4). The rather high
mobility of the grafted probe located at the C-terminal end of
the protein confirmed the non-structuration and thus high
flexibility of this segment, as proposed by Gardebien et al. on
the basis of structural modeling (Figure 1).[9] Mobility of the
nitroxide probe and flexibility of its binding site are reflected
in the EPR spectral shape, and the type of signal obtained
here is not unusual for a standard MTSL (methanethiosulfo-
nate spin label) spin probe either grafted on a small peptide or
in a disordered region of a protein.[2b,15,16]
to EPR spectra whose shape reflects the spin-label mobility.
The EPR signal of the labeled CP12 was typical of a nitroxide
probe having a very high mobility (Figure 4). In this high-
Introduction of the partner protein GAPDH in an
equimolar ratio (protein concentration 40 mm) induced a
weak but significant modification of the EPR spectral shape;
the h(À1)/h(0) ratio decreased from 0.67 in the absence of the
partner protein to 0.58 in its presence the partner protein, and
tc increased from 0.125 to 0.180 ns (Table 1 and Figure 4).
Owing to the high molecular mass of the GAPDH/CP12
complex (greater than 150 kDa), the mobility of the ensemble
does not contribute to the EPR spectral shape. Hence the
observed decrease of the ratio can be attributed to the
cancelation of this contribution rather than a local change in
the environment of the probe, in accordance with the similar
decrease of mobility observed with sucrose (Figure 4). This
result indicates that the C-terminal region where the label is
grafted remained highly flexible in the complex and that the
tyrosine residue at position 78 is not directly involved in the
interaction site between the two partners. This observation is
consistent with the proposed interaction region involving the
central a helix of CP12 and not the C-terminal end of the
protein.[17]
Figure 4. Amplitude-normalized EPR spectra recorded at 296(1) K:
a) free label, b) CP12*, c) CP12* in the presence of 30% sucrose,
d) CP12* complexed to the partner protein GAPDH, and e) inset of the
high-field line (tall to small, (a), (b), (d), (c))
mobility regime, the most sensitive semi-quantitative param-
eter to describe this spectral change is the ratio of the peak-to-
peak amplitude of the high-field line h(À1) over the central
line h(0).[1c,13] A significant decrease of this parameter was,
however, observed, going from 0.83 for the free nitroxide to
0.67 for the grafted nitroxide (Table 1 and Figure 4). Simu-
lation of the EPR spectra showed an increase of the rotational
This work exemplifies that a nitroxide probe can be
grafted to tyrosine residues by using the Mannich-type
electrophilic aromatic substitution reaction. Application of
this approach to other proteins is currently under investiga-
tion. Beyond the demonstration of feasibility of grafting a
nitroxide radical to residues other than cysteine, this approach
extends the capabilities of SDSL EPR spectroscopy and
opens new perspectives for the study of proteins bearing
functional cysteines.
Table 1: Ratio h(À1)/h(0) and correlation times tc for EPR signals
displayed in Figure 4.
Received: April 12, 2011
Revised: June 3, 2011
Published online: August 24, 2011
Free
label
CP12*[a]
CP12*+
CP12*GAPDH[c]
sucrose[b]
h(À1)/h(0)[d]
tc [ns][e]
0.83
0.060
0.67
0.125
0.54
0.225
0.58
0.180
Keywords: bioconjugation · EPR spectroscopy ·
Mannich reaction · proteins · radicals
.
[a] CP12* for labeled CP12. [b] CP12* in the presence of 30% sucrose.
[c] CP12* complexed to the partner protein GAPDH. [d] See Figure 4a,
error Æ0.02. [e] See the Supporting Information, errors Æ0.015 ns.
[1] a) W. L. Hubbell, D. S. Cafiso, C. Altenbach, Nat. Struct. Biol.
2000, 7, 735 – 739; b) J. P. Klare, H. J. Steinhoff, Photosynth. Res.
2009, 102, 377 – 390; c) V. Belle, S. Rouger, S. Costanzo, S.
Longhi, A. Fournel in Assessing Structures and Conformations of
Intrinsically Disordered Proteins (Ed.: V. N. Uversky), Wiley,
Hoboken, 2009.
[2] a) B. Morin, J. M. Bourhis, V. Belle, M. Woudstra, F. Carriere, B.
Guigliarelli, A. Fournel, S. Longhi, J. Phys. Chem. B 2006, 110,
20596 – 20608; b) N. L. Pirman, E. Milshteyn, L. Galiano, J. C.
Hewlett, G. E. Fanucci, Protein Sci. 2011, 20, 150 – 159.
[3] C. Kumsta, U. Jakob, Biochemistry 2009, 48, 4666 – 4676.
correlation time tc from 0.060 to 0.125 ns (see Table 1 and the
Supporting Information). For such small proteins (8.5 kDa),
the contribution of the protein rotation to the EPR spectral
line shape is not negligible. To eliminate this contribution, we
used 30% sucrose as a viscosity agent that increased the
protein rotational correlation time without exerting any
influence on the mobility of the nitroxide itself.[14] As
expected, a significant decrease of the h(À1)/h(0) ratio was
observed, from 0.67 to 0.54, corresponding to an increase of tc
9110
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9108 –9111

[4] J. Erales, M. Lorenzi, R. Lebrun, A. Fournel, E. Etienne, C.
Courcelle, B. Guigliarelli, B. Gontero, V. Belle, Biochemistry
2009, 48, 6034 – 6040.
[5] M. R. Fleissner, E. M. Brustad, T. Kalai, C. Altenbach, D.
Cascio, F. B. Peters, K. Hideg, S. Peuker, P. G. Schultz, W. L.
Hubbell, Proc. Natl. Acad. Sci. USA 2009, 106, 21637 – 21642.
[6] M. D. Best, Biochemistry 2009, 48, 6571 – 6584.
[7] J. M. Antos, M. B. Francis, J. Am. Chem. Soc. 2004, 126, 10256 –
10257.
[11] E. Graciet, P. Gans, N. Wedel, S. Lebreton, J. M. Camadro, B.
Gontero, Biochemistry 2003, 42, 8163 – 8170.
[12] J. Erales, L. Avilan, S. Lebreton, B. Gontero, FEBS J. 2008, 275,
1248 – 1259.
[13] K. B. Qu, J. L. Vaughn, A. Sienkiewicz, C. P. Scholes, J. S. Fetrow,
Biochemistry 1997, 36, 2884 – 2897.
[14] a) H. S. Mchaourab, M. A. Lietzow, K. Hideg, W. L. Hubbell,
Biochemistry 1996, 35, 7692 – 7704; b) V. P. Timofeev, V. I.
Tsetlin, Biophys. Struct. Mech. 1983, 10, 93 – 108.
[8] a) J. M. McFarland, N. S. Joshi, M. B. Francis, J. Am. Chem. Soc.
2008, 130, 7639 – 7644; b) H. Ban, J. Gavrilyuk, C. F. Barbas, J.
Am. Chem. Soc. 2010, 132, 1523 – 1525.
[9] F. Gardebien, R. R. Thangudu, B. Gontero, B. Offmann, J. Mol.
Graphics Modell. 2006, 25, 186 – 195.
[15] S. Popp, L. Packschies, N. Radzwill, K. P. Vogel, H. A. Steinhoff,
J. Reinstein, J. Mol. Biol. 2005, 347, 1039 – 1052.
[16] V. Belle, S. Rouger, S. Costanzo, E. Liquiꢀre, J. Strancar, B.
Guigliarelli, A. Fournel, S. Longhi, Proteins 2008, 73, 973 – 988.
[17] S. Lebreton, S. Andreescu, E. Graciet, B. Gontero, FEBS J. 2006,
273, 3358 – 3369.
[10] M. M. Bradford, Anal. Biochem. 1976, 72, 248 – 254.
Angew. Chem. Int. Ed. 2011, 50, 9108 –9111
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9111