Journal of Bioinformatics and Computational Biology
 Imperial College Press
HIGH TORSIONAL ENERGY DISULFIDES: RELATIONSHIP BETWEEN
CROSS-STRAND DISULFIDES AND RIGHT-HANDED STAPLES
NAOMI L. HAWORTH
Computational Biology & Bioinformatics Program, Victor Chang Cardiac Research Institute
Level 6, 384 Victoria Rd, Darlinghurst, New South Wales 2010, Australia
n.haworth@victorchang.unsw.edu.au
School of Chemistry, University of Sydney
New South Wales 2006, Australia
LINA L. FENG
Computational Biology & Bioinformatics Program, Victor Chang Cardiac Research Institute
Level 6, 384 Victoria Rd, Darlinghurst, New South Wales 2010, Australia
l.feng@victorchang.unsw.edu.au
MERRIDEE A. WOUTERS*
Computational Biology & Bioinformatics Program, Victor Chang Cardiac Research Institute
Level 6, 384 Victoria Rd, Darlinghurst, New South Wales 2010, Australia
m.wouters@victorchang.unsw.edu.au
Schools of Biotechnology & Biomolecular Sciences, and Medical Sciences, University of New South Wales
Sydney , New South Wales 2052, Australia
Received (Day Month Year)
Revised (Day Month Year)
Accepted (Day Month Year)
Redox-active disulfides are capable of being oxidized and reduced under physiological conditions.
The enzymatic role of redox-active disulfides in thiol-disulfide reductases is well known, but redoxactive disulfides are also present in non-enzymatic protein structures where they may act as switches
of protein function. Here we examine disulfides linking adjacent β-strands (cross-strand disulfides),
which have been reported to be redox-active. Our previous work has established that these crossstrand disulfides have high torsional energies, a quantity likely to be related to the ease with which
the disulfide is reduced. We examine the relationship between conformations of disulfides and their
location in protein secondary structures. By identifying the overlap between cross-strand disulfides
and various conformations, we wish to address whether the high torsional energy of a cross-strand
disulfide is sufficient to confer redox activity or whether other factors, such as the presence of the
cross-strand disulfide in a strained β-sheet, are required.
Keywords: disulfide; redox signaling; right-handed staple, cross-strand disulfide.
*
All correspondence to m.wouters@victorchang.unsw.edu.au
1
2
N. L. Haworth, L. L. Feng, M. A. Wouters
1. Background
Disulfides are generally viewed as structurally stabilizing elements in proteins. However,
it is well known that some disulfides are redox-active and capable of being reduced under
physiological conditions. The enzymatic role of redox-active disulfides has been well
characterized in thiol-disulfide reductases.1 It is less well known that similar disulfides
are present in other, nonenzymatic, protein structures. The likelihood that a disulfide will
be reduced is a function of its redox potential and its availability to reducing agents.
Disulfide redox potentials measured in thiol-disulfide oxidoreductases range from
-120mV to -270mV.2-4 For disulfides serving structural purposes, the redox potential can
be as low as -470mV.5
We first unearthed examples of a subset of unusual disulfides during a structural data
mining survey of β-sheets6 which revealed disulfides linked across the strands. These
cross-strand disulfides (CSDs) attracted our notice because they occur in a secondary
structure that is already non-covalently linked and thus, at best, seem redundant.
Furthermore, it had previously been asserted that such structures would not exist because
of the incompatible conformational constraints of β-sheets and disulfides.7 Indeed, we
found the disulfides and the β-sheets around them were significantly distorted. Thus,
these particular disulfides do not seem to fit the paradigm of disulfides as structural
stabilizers.
Emerging evidence suggests at least some of these disulfides may act as redox-active
switches of conformational change in proteins. Intriguing evidence for this includes:
demonstration that cleavage of a CSD in CD4 is required for entry of HIV-1 into the
cell;8 the solution of the structure of a novel oxidoreductase with a redox-active CSD; and
the engineering of a CSD into a variant of green fluorescent protein to produce a redoxsensitive sensor which has been used to monitor the redox potential in mitochondria and
cancer cells.9-10
We wish to elucidate computational approaches for determining whether a disulfide is
structurally stabilizing or may be subject to reduction in a protein structure. Redox
potential is a difficult quantity to calculate, being a function of the strain of the disulfide
bond, the electronic environment of the protein and the properties of the solute. Our
previous work established that CSDs have high torsional energies, a significant
contributor to the disulfide bond strain.11 Here we report a further survey of CSDs and
related disulfides in the Protein Data Bank (PDB) to examine the relationship between
conformations of disulfides and their location in protein secondary structures.
2. Methods
In order to carry out the bioinformatics survey of CSDs, we created a custom database of
all disulfides in a unique subset of the PDB (release 110). This subset included only Xray crystal structures with a resolution of 2 Å or better. The database contains
information about each disulfide, including the solvent accessibility, the position of the
two half-cystines in primary, secondary, tertiary and quaternary structure, and the
torsional energy of the disulfide. These data were compiled using various bioinformatics
Redox-active disulfides: Relationship between Cross-Strand Disulfides and Right-Handed Staples
3
tools such as DSSP12 and custom programs, as well as by reference to on-line databases
such as SCOP.13
The torsional energy of each disulfide was predicted via two different methods. The
first used an empirical formula employed in the Amber force field package:14
E(kJ mol-1) = 8.37(1+cos31) + 8.37(1+cos31’) + 4.18(1+cos32)
+ 4.18(1+cos32’) + 14.64(1+cos23) + 2.51(1+cos33).
(1)
where χ1 and χ2 are the dihedral angles of the cysteine residues and χ 3 is the dihedral
angle of the disulfide bond. A factor of 2.05 kJ mol-1 was subtracted from this Amber
energy to ensure that a torsional energy of zero corresponds to the lowest energy disulfide
conformation (ie. all torsional energies are relative). In the second method, the
contributions of 2, 3 and 2’ to the torsional energy of a disulfide were calculated using
a three-dimensional ab initio potential energy surface (PES). Ab initio quantum
chemistry, in particular the MP2/6-31G* level of theory, was used to predict the energetic
cost of torsion around the dihedral angles for a model compound, diethyl disulfide. Grid
points were calculated at increments of ten degrees along each of the coordinates,
corresponding to the three dihedral angles, 2, 3 and 2’. Full details and analysis of this
surface will be reported elsewhere.
The high number of grid points means that this surface is relatively continuous,
making it possible to use a simple linear interpolation in three dimensions to predict the
torsional energy of a disulfide without a significant loss of accuracy. The energetic
contributions of 1 and 1’ were estimated using the first two terms in the Amber Force
Field formula (Eq. (1) ) and included additively.
3. Results and Discussion
3.1 Analysis of the Potential Energy Surface
The potential energy surface, part of which is shown in Figure 1A, reveals several
minima, corresponding to low torsional energy, stable structures (expected to be
structural disulfides) as well as high torsional energy saddle points and maxima.
Disulfides adopting conformations in these high energy regions are less stable and
therefore more likely to be redox-active. For the purposes of this work a high torsional
energy (HiTED) disulfide has a predicted torsional energy exceeding 10.0 kJ mol -1.
Conventionally, the χ2 torsional angle is defined as being gauche when its absolute
value is between 30 and 90 degrees, with gauche+ (g+) being positive and gauche- (g-)
being negative, and trans in the ranges -150° to -180° and 150° to 180°.15 Typically, the
χ3 torsional angle is merely designated right- or left-handed depending on whether the
value is positive or negative, respectively.7 A disulfide is defined as a staple if the χ2, χ3
and χ2’ torsional angles formed a g-+g- motif (right-handed staple) or a g+-g+ motif (lefthanded staple).7,15 Similarly, a disulfide is defined as being a spiral if it has a g ++g+ or g-g- motif (right- and left-handed respectively), while those with a g++g- or g-+g+ motif are
called right-handed hooks and those with a g+-g- or g--g+ motif are left-handed hooks.
N. L. Haworth, L. L. Feng, M. A. Wouters
G-+T
120
30.0
25.0
Population
20.0
15.0
100
80
Hook
60
40
10.0
0
5.0
20
-80
Fig 1A
2' (degrees)
70
40
100
160
130
-140
Hook
-80
0
-170
70
2
(deg.)
40
40
2' (degrees)
100
160
130
-140
-170
-80
-110
-20
-50
120
-110
-160
0.0
-20
Relative energy (kJ/mol)
Staple
140
35.0
-50
4
-20
-80
-140
160
100
40
2
(deg.)
Spiral
Fig 1B
Fig. 1A. A slice through the three dimensional potential energy surface for diethyl disulfide, showing the
change in the relative torsional energy as 2 and 2’ are varied while 3 is held constant at +90 degrees. 1B. The
variation in disulfide population in the PDB as 2 and 2’ are varied while 3 = +90 degrees. Bins are from x-5
to x+5 degrees.
Analysis of the disulfides in recent releases of the PDB has revealed that many adopt
conformations with 2 and 2’ torsional angles outside these conventionally defined
ranges.
In order to account for these structures, we have adopted slightly different dihedral
angle bins based on inspection of the one-dimensional torsional potential energy surfaces
of Görbitz16 and our three-dimensional potential energy surface for diethyl disulfide. This
allows the strain in the torsional angle to be decomposed and investigated more
thoroughly. Thus, G+ (“gauche+”) was defined as being between 10 and 115 degrees for
2, and 2’, with gauche- (G-) being between -20 and -120 degrees. Trans (T) torsional
angles for 2, and 2’ were defined as being between -134 and 128 degrees.
Torsional angles lying between the gauche and trans regions correspond to
comparatively low energy saddle points on the potential energy surface. These regions
are expected to be energetically accessible to disulfides and are therefore labeled skew+
(S+) and skew- (S-). This gives rise to several new conformations: skew-spiral
conformations which sit on the two low energy saddle points neighboring the spirals;
skew-hook conformations which neighbor the hooks; and skew-extended conformations
in the four saddle points surrounding the central TGT region. The remaining disulfide
conformations are simply labeled by their torsional angles.
While these potential energy surface-based definitions are useful for most values of
2, and 2’, they begin to break down in the staple region. Inspection of the diethyl
disulfide PES reveals that, for 3 angles between 70 and 110 degrees, there is no
minimum on the surface corresponding to the staple conformation. This is because the
particular structural constraints of this configuration result in steric interactions between
the Hα atoms of the cysteine residues, which in turn destabilize the system (See Figure
2). Staple-type disulfides therefore belong to a saddle point region on the potential energy
Redox-active disulfides: Relationship between Cross-Strand Disulfides and Right-Handed Staples
5
surface and generally have energies above 10 kJ mol-1. Thus, in almost all cases, they will
be HiTEDs.
Despite the low predicted stability for the staple conformations, significant numbers
of disulfides are found to have the G-+G- motif (right-handed staple). The population
distribution for disulfides with respect to the three dihedral angles shows a peak in the G+G- region of the potential energy surface (Figure 1B). This peak is associated with 2
values between -20 and -140 degrees and 2’ values between -20 and -130 degrees. All
disulfides with 2, and 2’ within these limits were therefore classed as right-handed
staples. The analogous definition was used for left-handed staples.
Fig. 2. Cross-strand disulfide in the right-handed staple conformation. Steric interference between the H atoms
significantly destabilizes staple conformations; in addition, the -sheet is significantly distorted.
3.2. Structural Motifs associated with Cross-strand disulfide types
Experimental evidence suggests some redox-active disulfides are regenerated while
others are probably reduced on a one-off basis. Regenerated disulfides cycle through
reduced and oxidized states as part of their function. Examples of regenerated CSDs
include the engineered GFP and DsbD. Redox-active disulfides which are reduced as
one-off events include disulfides that straddle proteolytic cleavage sites in viruses and
bacterial toxins. For example, upon reduction of the CSD in clostridial neurotoxin A, the
molecule transits to an enzymatically-active molten globule state.17 We thought it likely
that regenerated disulfides and those used once only may be associated with different
secondary structure motifs. For example, we wondered whether one-off CSDs might be
associated with easily dissociated secondary structures. Several other factors may also
impact the secondary structure motif adopted by redox-active disulfides. These include
the latent state of the disulfide, the speed of the transition between the oxidized and
reduced states and the redox pathways transited. For example, reduced cysteine pairs that
are held rigidly in a secondary structure have less conformational space to search during
the regeneration reaction. In addition, formation of some CSDs has been shown to be
dependent on particular enzymes which may require presentation of the thiol groups in
specific structural contexts. For example, oxidation of CSDs in influenza haemaglutinnin
and PapD have been shown to be dependent on erp57 and dsbA respectively. 11,18Crossstrand disulfides were identified through DSSP analysis of the protein structures. We
differentiated five different classes of cross-strand disulfides based on analysis of the
conformation of neighbouring residues and surrounding hydrogen bonds: true CSDs, end
6
N. L. Haworth, L. L. Feng, M. A. Wouters
Fig. 3. Schematic representation of different classes of cross-strand disulfides. In order to be classed as
a true CSD, both cysteine residues and the backbone of all four i1 neighbors must be in  conformation and
neighbours must be appropriately hydrogen-bonded. For end CSDs, surrounding hydrogen bonding must be
maintained but neighbours on one side are permitted to adopt non- conformations. For frayed CSDs, the
hydrogen-bonding condition on one side is also relaxed. In weak frayed CSDs only the pair of residues
immediately neighboring the disulfide were required to fit the hydrogen bonding and  conformation
conditions. This category required visual inspection of the structure to check it was a CSD structure. βbridges are flanked by hydrogen bonds but neighbouring residues are not in -conformation. Amino acid
residues in β-conformation are represented by boxes, while circles are used for those in any other backbone
conformation. The polypeptide chains are represented by thick grey lines and hydrogen bonds are indicated
by arrows. Hydrogen bonding is depicted for antiparallel sheets.
CSDs, frayed CSDs, β-bridges and weak-frayed CSDs. A schematic representation of
each of these classes can be found in Figure 3. The distribution of the cross strand
disulfides amongst the different CSD types can be seen in Figure 4. The true CSDs
comprised 16% of the CSD population, all of which were in the right-handed staple
conformation. The largest CSD subgroup (41%) was the end CSDs; of these, 29 were
parallel CSDs (mostly left-handed hooks and skew-hooks), with the remainder being
almost all right-handed staple disulfides. The frayed CSD group was also quite large
(34%), again being mainly right-handed staples, however there were small populations of
several other conformations, including sixteen left-handed G--Ts, six right-handed G-+Ts
and eight left-handed skew hooks. The weak-frayed CSDs and the -bridges were much
smaller subgroups; of the 76 weak-frayed CSDs, roughly two thirds were right-handed
staples with the remaining third belonging to the right-handed G-+T conformation. The
Redox-active disulfides: Relationship between Cross-Strand Disulfides and Right-Handed Staples
7
26 -bridges were almost all right-handed staples, with the exception of two left-handed
staple structures. There was no distinction between the different CSD types on the basis
of torsional energies, with the averages for the five different groups all being
approximately 14 kJ mol-1. Thus, on the basis of torsional energy alone, no particular
CSD type is more likely to be redox-active.
Table 1 tabulates properties investigated here against experimental evidence
supporting redox activity in a subset of CSDs which we have identified to date. Some
general trends are observable but the numbers are too small to make definite conclusions.
All parallel CSDs identified are of the frayed type and are reduced in the latent state of
their resident protein. Only one of the easily dissociated types was identified in the
experimental dataset but it was associated with a CSD likely to be reduced irreversibly:
the fusion protein of influenza C.
Fig. 4. The distribution of cross-strand disulfides amongst CSD types.
Table 1. CSDs likely to have redox-activity.
Regenerated CSDs
One-off
Protein
CD4a,8
DsbDa,1
GFPa,9-10
PapDa,21
gelsolina
SAPa
CDC25Ba,22
GlmUa
MurDa
Disulfide
130-159
103-109
147-204
207-212
188-201
36-95
426-473
307-324
208-227
CSD type
A-true
A-end
A-frayed
A-true
A-end
A-true
P-frayed
P-frayed
P-frayed
Latent state
oxidized
oxidized
oxidized
oxidized
oxidizedc
oxidized
reduced
reduced
reduced
Conformation
RHS
Str RHS
RHS/Str RHS
Str RHS
RHS
RHS
LHS
RHS
LSH
Gp120a
Flu A HAb
Botulinum Ab
Neurotoxin Bb
Flu C HEFb
296-331
A4-B137
429-453
436-445
6-137
A
A-end
A-end
A-end
A-weak
frayed
oxidized
oxidized
oxidized
oxidized
oxidized
RHS
RHS/Str RHS
-
Note: Classified as redox-active on the basis of
a
crystallographic redox pairs;
b
straddles cleavage site: required for chain dissociation.
c
Oxidized in plasma gelsolin, reduced in cellular gelsolin References indicate additional experimental
evidence for redox-activity. Conformations are only quoted for high resolution structures.
8
N. L. Haworth, L. L. Feng, M. A. Wouters
3.3. Conformations adopted by CSDs
By examining the overlap between CSDs and various disulfide conformations, we wish
to address whether the high torsional energy of a CSD is sufficient to confer redox
activity or whether other factors, such as the presence of the CSD in a strained β-sheet
environment, are required. Harrison & Sternberg suggested a one-to-one correspondence
between the right-handed staple conformation and disulfides straddling antiparallel βstrands in the non-H-bonded or “wide” position.6,19,20 They used the term right-handed
staple synonymously for both the conformation and the secondary structure motif. Our
studies indicate, however, that not all CSDs assume the right-handed staple conformation
and not all right-handed staples are CSDs (Figure 5A).
Our results indicate that around 93% of the 1135 CSDs in our database assume the
right-handed staple conformation.15 It is therefore necessary to differentiate between the
conformation (right-handed staple) and the motif, which we have dubbed the cross-strand
disulfide (CSD). The CSDs in our database are also seen to adopt the lower energy righthanded G-+T conformation (2.4%), the left-handed G+-T conformation (1.2%) and the
higher energy left-handed skew-hook conformation (1.8%). Five structures were seen for
both the right-handed G++T and left-handed hook conformations and three structures
were also found to be in the very high energy left-handed staple conformation. CSDs in
the left-handed skew hook conformation are predominantly found in parallel β-sheets.
These are largely specific to xylanases.
Amongst the right-handed staple disulfides which are not CSDs, there are three major
groups. Firstly, disulfides between two strands which run roughly parallel to each other
but are not in β conformation. These include the disulfides between Cys318 and Cys337
in influenza neuraminidases, Cys902 and Cys987 in golgi α-mannosidases and Cys176
and Cys209 in cellobiohydrolases. In the latter case, it appears that the two strands
involved may have once formed part of a β-sheet but changes in the protein structure
have forced one of the strands to break away. The second class includes disulfides in
extremely weakly bound β-sheets, where the hydrogen bond strength is below the DSSP
cutoffs, and also disulfides in sheets with very large β-bulges, for example, the disulfide
between Cys6 and Cys12 in the methanol dehydrogenase light subunit. The final, and
most interesting, class of right-handed staple disulfides which are not CSDs occur in
plant peroxidase enzymes, a subset of heme-dependent class III plant peroxidases (SCOP
b.b.bdc.b.b.group 4).13 This group includes the well known, commercially available,
enzyme, horseradish peroxidase C. A vicinal i,i+5 disulfide between Cys44 to Cys49
occurs in the final loop of an α-helix, effectively terminating the helical structure. This
same structural motif is also observed in glutathione reductase and related enzymes
(where they adopt a left-handed staple conformation) and in a group of non-heme
dependent peroxidases: the Prx Q-like enzymes.21 It is possible that the right-handed
staple, Cys44-Cys49 disulfide in plant peroxidases may constitute a second redox centre
in these proteins.
Redox-active disulfides: Relationship between Cross-Strand Disulfides and Right-Handed Staples
9
Right-handed staples which are not CSDs are on average slightly lower in energy
than other right-handed staple disulfides (12.8 ± 3.3 kJ mol-1 vs 13.7 ± 2.6 kJ mol-1
respectively).
Fig. 5A. Relationship between right-handed staples and CSDs. 5B. The torsional energy distributions of CSDs
in the three dominant conformations (anti-parallel right-handed staples, anti-parallel right-handed G-+Ts and
parallel left-handed hooks) in comparison with the overall disulfide distribution. X-axis bins are labeled with
the upper bound.
The torsional energy distribution of CSDs in various conformations is shown in
Figure 5B. The largest population of cross-strand disulfides belonged to the anti-parallel
right-handed staple conformation, comprising 85% of all CSDs. Naturally, therefore, the
anti-parallel right-handed staple and cross strand disulfide distributions are very similar,
both being very narrow, with averages around 14 kJ mol -1 (13.7 ± 2.1 kJ mol-1 for antiparallel right-handed staples and 14.0 ± 3.7 kJ mol-1 for CSDs). Almost all structures in
both these groups correspond to high-torsional-energy disulfides. A small number of
CSDs are not HiTEDs; these are almost all disulfides in the G-+T conformation. This
G-+T population had 28 members, some of which are, in fact, found to be high-torsionalenergy disulfides; the energy distribution is fairly wide, with an average of 10.9 ± 6.4 kJ
mol-1. Parallel CSDs in the left-handed hook conformation are predominantly at the high
10
N. L. Haworth, L. L. Feng, M. A. Wouters
torsional energy end of the spectrum (27 examples with an average energy of 20.1 ± 4.5
kJ mol-1). This distribution is bimodal, with the upper peak corresponding, in the most
part, to the xylanase structures and the lower peak being produced by disulfides in the
ADP forming ligase, MurD. The distribution suggests that on the basis of torsional
energy alone, the left-handed hook conformation within a parallel CSD might be even
more redox-active than the more common right-handed staple.
No preference for particular high torsional conformations are discernable in Table 1
but these may become apparent when the experimental dataset expands and other
variables are taken into account.
3.4. Conformational lability of CSDs
Comparison of the structures of homologs and ensembles of NMR structures (eg.
reference 22) has suggested there is conformational lability between left-handed and
right-handed staple disulfides, although the lower energy right-handed staple
conformation is generally more highly populated for homologous structures.
We have therefore further investigated the NMR structures within the PDB which
contain cross-strand disulfides. We found 93 cross-strand disulfides in 79 different
protein structures, giving a total of 2004 different NMR models. 22 of the cross-strand
disulfides adopted only one conformation, with 21 of these being always right-handed
staples and one having only models in the left-handed staple conformation. 26 of the
CSDs alternated between right- and left-handed staples and nine alternated between righthanded staples and left-handed hooks. 12 of the cross-strand disulfides were split
between all three of these conformations (right- and left-handed staples and right-handed
hooks). The remaining CSDs showed significant numbers of other less common
conformations. There seemed to be no correlation between the protein type and the
conformations adopted, with, for example, the disulfide between Cys37 and Cys47 in
actinoxanthin-type domains showing up three of the four classes.
Overall, 53% of the NMR models predicted the CSD to be in the right-handed staple
conformation, 16% predicted a left-handed staple and 8% predicted a left-handed hook.
4% of the structures were in the right-handed G-+T conformation and 5% were lefthanded disulfides in a particularly high energy conformation, with one of 2 or 2’ close
to 0 degrees. There were also a small number of models which predicted the right-handed
hook, left-handed G+-T, left-handed skew hook and left-handed G--T conformations (2%
each).
We are currently investigating factors which might influence the adoption of one of
these conformations over the others.
4. Conclusions
In conclusion, both the right-handed staple conformation and the cross-strand disulfide
motif were found to give rise to high-torsional-energy disulfides with the potential to be
redox active. While there was a high degree of correlation between these two sets (93%),
examples were seen of both right-handed staples which were not CSDs and CSDs which
Redox-active disulfides: Relationship between Cross-Strand Disulfides and Right-Handed Staples
11
did not adopt the right-handed staple conformation. Amongst these structures, redox
activity has only been reported for disulfides which are both right-handed staples and
CSDs, however some right-handed staples which are not CSDs are seen in secondary
structure motifs which have been associated with redox activity. These data suggest that
any conformation associated with high torsional energies is potentially redox-active.
A feature of the redox-active disulfides we have encountered thus far is that they
occur within particular secondary structure motifs, such as CSDs in -sheets; or close in
primary structure, such as i,i+3 disulfides in thioredoxins, or the i,i+5 disulfide in
glutathione reductase. In many cases, hydrogen bonds seem to be involved in stabilizing
the disulfide in its high energy conformation. In the case of CSDs, the constraining
hydrogen bonds are part of the regular -sheet structure.
Other variables likely to be important for disulfide bond redox activity include
solvent accessibility and the electronic environment of the disulfide.
Solvent accessibility is required to enable access by the reductant whether it be
reactive oxygen species, glutathione or an exogeneous protein molecule. In some cases
such as the engineered redox sensitive GFP, the disulfide always remains solvent
accessible and such cases are straightforward to study. However in other cases the
disulfide is normally protected from solvent and a conformational change is required to
expose the disulfide for reduction. Examples include the CSD in influenza
haemagglutinin, which is exposed during a pH-dependent conformational change;11 the
CSD in CD4 which has been proposed to be solvent exposed by the process of domain
swapping; and DsbD where a cap structure normally protects the disulfide. 1 Systematic
approaches for dealing with these occult solvent accessible sites are necessary.
The electronic environment around the CSD is also likely to influence redox
processes. For example, during the oxidation process, electronic environments that lower
the pKa of thiols enhance their reactivity. The helix-dipole effect and the effects of
spatially conserved amides and basic amino acids are known to contribute to reactivity of
the CSD in CDC25B.23 We are investigating the electronic environment surrounding
CSDs using a computational chemistry approach.
Acknowledgments
The authors would like to express our thanks to the Australian Partnership for Advanced
Computing for access to the COMPAQ AlphaSever SC system.
References
1.
2.
3.
C. W. Goulding, M. R. Sawaya, A. Parseghian, V. Lim, D. Eisenberg and D. Missiakas,
“Thiol-disulfide exchange in an immunoglobulin-like fold: structure of the N-terminal
domain of DsbD,” Biochemistry. 41:6920-7 (2002).
M. Wunderlich and R. Glockshuber, “Redox properties of protein disulfide isomerase
(DsbA) from Escherichia coli,” Protein Sci. 2:717-26 (1993).
M. Huber-Wunderlich and R. Glockshuber, “A single dipeptide sequence modulates the
redox properties of a whole enzyme family,” Fold Des. 3:161-71 (1998).
12
N. L. Haworth, L. L. Feng, M. A. Wouters
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
T. Y. Lin and P. S. Kim, “Urea dependence of thiol-disulfide equilibria in thioredoxin:
confirmation of the linkage relationship and a sensitive assay for structure,” Biochemistry.
28:5282-7 (1989).
H. F. Gilbert, “Molecular and cellular aspects of thiol-disulfide exchange,” Adv Enzymol
Relat Areas Mol Biol. 63:69-172 (1990).
M. A. Wouters and P. M. Curmi, “An analysis of side chain interactions and pair
correlations within antiparallel β-sheets: the differences between backbone hydrogenbonded and non-hydrogen-bonded residue pairs,” Proteins. 22:119-31 (1995).
J. S. Richardson, “The anatomy and taxonomy of protein structure,” Adv Protein Chem.
34:167-339 (1981).
L. J. Matthias, P. T. W. Yam, X-M. Jiang, N. Vandergraaff, P. Li, P. Poumbourios, N.
Donoghue and P. J. Hogg, “Disulphide exchange in domain 2 of CD4 is required for entry
of the Human Immunodeficiency Virus Type 1,” Nature Immunol. 3:727-732 (2002).
G. T. Hanson, R. Aggeler, D. Oglesbe, M. Canon, R. A. Capaldi, R. Y. Tsien and S. J.
Remington, “Investigating mitochondrial redox potential with redox-sensitive green
fluorescent protein indicators,” J Biol Chem. 279(13):13044-53 (2004).
R. Rossignol, R. Gilkerson, R. Aggeler, K. Yamagata, S. J. Remington and R. A. Capaldi,
“Energy substrate modulates mitochondrial structure and oxidative capacity in cancer
cells,” Cancer Res. 64:985-93 (2004).
M. A. Wouters, K. K. Lau and P. J. Hogg, “Cross-strand disulphides in cell entry proteins:
poised to act,” Bioessays. 26:73-9 (2004).
W. Kabsch and C. Sander, “Dictionary of protein secondary structure: pattern recognition of
hydrogen-bonded and geometrical features,” Biopolymers. 22:2577-637 (1983).
A. G. Murzin, S. E. Brenner, T. Hubbard and C. Chothia, “SCOP: a structural classification
of proteins database for the investigation of sequences and structures,” J Mol Biol. 247:53640 (1995).
D. A. Pearlman, D. A. Case, J. W. Caldwell, W. R. Ross, T.E. Cheatham, III, S. Debolt, D.
Ferguson, G. Seibel and P. Kollman, “AMBER, a computer program for applying molecular
mechanics, normal mode analysis, molecular dynamics and free energy calculations to
elucidate the structures and energies of molecules,” Comp. Phys. Commun. 91:1-41 (1995).
P. M. Harrison and M. J. Sternberg, “The disulphide β-cross: from cystine geometry and
clustering to classification of small disulphide-rich protein folds,” J Mol Biol. 264:603-23
(1996).
C. H. Görbitz, “Conformational properties of disulphide bridges. 2. Rotational potentials of
diethyl disulphide,” J. Phys. Org. Chem. 7: 259-267 (1994).
S. Cai, B. R. Singh, “Role of the disulfide cleavage induced molten globule state of type a
botulinum neurotoxin in its endopeptidase activity,” Biochemistry 40:15327-33 (2001).
F. Jacob-Dubuisson, J. Pinkner, Z. Xu, R. Striker, A. Padmanhaban, S. J. Hultgren. “PapD
chaperone function in pilus biogenesis depends on oxidant and chaperone-like activities of
DsbA”. Proc Natl Acad Sci U S A. 91: 11552-6 (1994).
F. R. Salemme, “Structural properties of protein beta-sheets,”Prog Biophys Mol Biol. 42(23): 95-133 (1983).
J. S. Richardson and D. C. Richardson, “Prediction of protein structure and the principles of
protein conformation”. In Prediction of Protein Structure and the Principles of Protein
conformation, ed. G. D. Fasman, p1-99, Plenum Press, New York 1989.
K. J. Dietz, “Plant peroxiredoxins,” Annu Rev Plant Biol. 54: 93-107 (2003).
J. P. Powers, A. Rozek and R. E. Hancock, “Structure-activity relationships for the βhairpin cationic antimicrobial peptide polyphemusin I,” Biochim Biophys Acta. 1698: 23950 (2004).
Redox-active disulfides: Relationship between Cross-Strand Disulfides and Right-Handed Staples
13
23. G. Buhrman, B. Parker, J. Sohn, J. Rudolph, C. Mattos, “Structural Mechanism of
Oxidative Regulation of the Phosphatase Cdc25B via an Intramolecular Disulfide Bond,”
Biochemistry 44: 7602 (2005).
Naomi Haworth received her PhD in Theoretical Chemistry
from the University of Sydney, Australia, in 2003. She was a
postdoctoral research fellow at the University of Sydney from
2003 to 2004 where she investigated the thermochemistry of
small calcium containing molecules. In 2004 she moved to the
Computational Biology & Bioinformatics Program at the Victor
Chang Cardiac Research Institute, Sydney, Australia, where she
has been using both theoretical chemistry and bioinformatics
techniques to investigate the stability of disulfide bonds.
Lina Feng received her B. Engineering (computer) degree and M.
Biomedical Engineering degree from University of New South
Wales, Australia in 2004. She is currently in the Computational
Biology & Bioinformatics Program at the Victor Chang Cardiac
Research Institute, Sydney, Australia.
Merridee Wouters is a Senior Research Scientist and Group
Leader in the Computational Biology & Bioinformatics Program at
the Victor Chang Cardiac Research Institute, Sydney, Australia.
She obtained her PhD from the School of Physics, University of
New South Wales (1996) based on a study of amino acid pair
correlations in β-sheet. She is particularly interested in the study of
proteins as machines.