Communications
DOI: 10.1002/anie.200905206
Protein Structures
Direct Visualization of Disulfide Bonds through Diselenide Proxies
Using 77Se NMR Spectroscopy**
Mehdi Mobli, Aline Dantas de Araffljo, Lynette K. Lambert, Gregory K. Pierens,
Monique J. Windley, Graham M. Nicholson, Paul F. Alewood, and Glenn F. King*
Disulfide bridges are a common posttranslational modifica-
tion in proteins that result from the formation of a covalent
bond between the thiol groups of two cysteine residues.
Disulfide bridges are important for directing and stabilizing
the three-dimensional (3D) structure of proteins[1] and they
are often functionally critical.[2,3] Incorrect pairing of disulfide
bonds typically leads to a nonnative 3D fold accompanied by
a loss of protein function. The importance of this structural
motif is highlighted by the fact that 24% of the circa 53000
protein structures deposited in the Protein Data Bank (PDB)
contain at least one disulfide bond.
Disulfide bonds are highly prevalent in three important
classes of secreted proteins: peptide hormones, growth
factors, and peptide toxins excreted in animal venoms.[4]
NMR spectroscopy is an ideal technique for determining
the 3D structure of these small proteins; of the 4285 structures
available in the PDB for proteins smaller than 8 kDa, 71%
were determined using NMR methods. However, this raises a
conundrum—NMR spectroscopy is a poor method for
establishing disulfide-bond connectivities owing to the
absence of a sulfur isotope with favorable NMR properties.
In some cases, dipolar interactions between the b-methylene
protons of covalently linked cysteine residues can be used to
infer disulfide bridges,[5] but these connectivities are often
ambiguous when multiple disulfide bridges are present.[6,7]
Thus, disulfide bonds often constitute a “blind spot” in NMR
structural analyses.
is a genetically encoded amino acid residue that is identical to
cysteine except for the replacement of sulfur with sele-
nium.[8,9] The human selenoproteome comprises 25 seleno-
proteins.[10] Sec residues can be introduced into proteins by
chemical synthesis[11–14] or recombinant expression,[15,16] and
replacement of disulfide bonds with diselenide bonds has
been demonstrated previously not to alter protein structure or
function.[13,14,17]
With a natural abundance of 7.6%, 77Se is an NMR-active
s = 1/2 nucleus that provides a route for direct determination
of disulfide bridge connectivities using NMR scalar couplings.
We demonstrate this approach using a 37-residue spider toxin
(k-ACTX-Hv1c) that contains four disulfide bonds, including
a rare and functionally critical vicinal disulfide bridge
between the adjacent cysteine residues Cys13 and Cys14.[2]
Since recombinant expression would lead to replacement of
all Cys residues with Sec, for the sake of simplicity we used
solid-phase peptide synthesis to produce a toxin in which only
Cys13 and Cys14 were replaced with Sec residues (see the
Supporting Information).
We determined the 3D solution structure of the
[Sec13,Sec14] toxin using homonuclear NMR methods (see
the Supporting Information) and found it to be equivalent to
that of the native toxin; the two structures can be super-
imposed over the well-ordered region (residues 1–34) with a
backbone root-mean-square deviation (rmsd) of 0.91 ꢀ
(Figure 1a). k-ACTX-Hv1c is a high-affinity blocker of
calcium-activated potassium (KCa) channels in insects, and
the vicinal disulfide bond is a key component of the channel
binding site.[2,18,19] Figure 1b shows that the [Sec13,Sec14]
variant is equipotent with the native toxin in blocking KCa
currents in cockroach neurons (see the Supporting Informa-
tion). Thus, replacing the functionally critical vicinal disulfide
bridge in k-ACTX-Hv1c with a vicinal diselenide bond does
not affect either the folding or function of the protein.
Figure 2 shows that the diselenide connectivities in the
[Sec13,Sec14] toxin (which serve as a proxy for the disulfide
connectivities) can be unequivocally determined from scalar
couplings present in a 2D 1H–77Se heteronuclear multiple
bond correlation (HMBC) experiment. The two intense cross-
peaks in Figure 2b result from large intraresidue two-bond
couplings (ca. 35 Hz) between each pair of Sec methylene
protons and the 77Se nucleus of the same residue. The two less
intense but clearly visible cross-peaks arise from interresidue
three-bond couplings across the diselenide bond between
methylene protons and 77Se. This latter coupling, although
small (ca. 2 Hz), allows unequivocal assignment of the
diselenide bond connectivity.
One possible solution to this “disulfide blind spot” would
be to replace the NMR-inactive 32S nucleus with a nucleus of
similar chemical, but more favorable magnetic, properties.
Such a replacement occurs naturally, albeit infrequently.
Selenocysteine (Sec), often referred to as the 21st amino acid,
[*] Dr. M. Mobli, Dr. A. D. de Araffljo, Prof. P. F. Alewood, Prof. G. F. King
Institute for Molecular Bioscience, The University of Queensland
St. Lucia QLD 4072 (Australia)
Fax: (+61)7-3346-2101
E-mail: glenn.king@imb.uq.edu.au
Dr. L. K. Lambert, Dr. G. K. Pierens
Centre for Magnetic Resonance, The University of Queensland
St. Lucia QLD 4072 (Australia)
M. J. Windley, Prof. G. M. Nicholson
Department of Medical and Molecular Biosciences
University of Technology, Sydney (Australia)
[**] This work was supported by Discovery Grants DP0774245 and
DP0878450 from the Australian Research Council and the
Queensland Smart State Research Facilities Fund.
Supporting information for this article is available on the WWW
9312
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 9312 –9314