Embed Size (px)
Citation preview
www.bba-direct.com
Biochimica et Biophysica Acta 1648 (2003) 174–183
Denaturant-induced equilibrium unfolding of concanavalin A is expressed
by a three-state mechanism and provides an estimate of its protein stability
Anindya Chatterjee, Dipak K. Mandal*
Department of Chemistry, Presidency College, 86/1 College Street, Calcutta 700073, India
Received 18 July 2002; received in revised form 19 February 2003; accepted 21 February 2003
Abstract
The urea and guanidine hydrochloride (GdnHCl)-induced denaturation of tetrameric concanavalin A (ConA) at pH 7.2 has been studied
by using intrinsic fluorescence, 8-anilino-1-naphthalenesulfonate (ANS) binding, far-UV circular dichroism (CD), and size-exclusion
chromatography. The equilibrium denaturation pathway of ConA, as monitored by steady state fluorescence, exhibits a three-state mechanism
involving an intermediate state, which has been characterized as a structured monomer of the protein by ANS binding, far-UV CD and gel
filtration size analysis. The three-state equilibrium is analyzed in terms of two distinct and separate dissociation (native tetramer X structured
monomer) and unfolding (structured monomer X unfolded monomer) reaction steps, with the apparent transition midpoints (Cm), respec-
tively, at 1.4 and 4.5 M in urea, and at 0.8 and 2.4 M in GdnHCl. The results show that the free energy of stabilization of structured monomer
relative to the unfolded state (�DGunf, aq), is 4.4–5.5 kcal mol� 1, and that of native tetramer relative to structured monomer (�DGdis, aq) is
7.2–7.4 kcal mol� 1, giving an overall free energy of stabilization (�DGdis&unf, aq) of 11.6–12.9 kcal mol� 1 (monomer mass) for the native
protein. However, the free energy preference at the level of quaternary tetrameric structure is found to be far greater than that at the tertiary
monomeric level, which reveals that the structural stability of ConA is maintained mostly by subunit association.
D 2003 Elsevier Science B.V. All rights reserved.
Keywords: Concanavalin A; Dissociation; Unfolding; Denaturation curve; Free energy of stabilization
1. Introduction
Lectins, proteins of non-immune origin that bind specif-
ically to carbohydrate epitopes, have been implicated in
various biological recognition processes, such as viral, bac-
terial, mycoplasmal and parasitic infections, targeting of cells
and soluble components, fertilization, cancer metastasis, and
growth and differentiation [1–3]. Once thought to be con-
fined to plant seeds, lectins are now recognized as ubiquitous
in virtually all living systems, ranging from viruses and
bacteria to animals [4]. Plant lectins are the model system
of choice to study the molecular basis of the recognition
phenomena because of their broad distribution and ease of
isolation [5], and their ability to exhibit a wide variety of
1570-9639/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserv
doi:10.1016/S1570-9639(03)00120-1
Abbreviations: ConA, concanavalin A, lectin from jack bean (Cana-
valia ensiformis); GdnHCl, guanidine hydrochloride; ANS, 8-anilino-1-
naphthalenesulfonate; PBS, 0.01 M sodium phosphate buffered with 0.15 M
sodium chloride, pH 7.2; CD, circular dichroism
* Corresponding author. Fax: +91-33-2512-3156.
E-mail address: [email protected] (D.K. Mandal).
carbohydrate specificities despite strong sequence conserva-
tion [6]. Of all the plant lectins studied to date, concanavalin
A (ConA), the lectin from Canavalia ensiformis (jack bean),
is the best known member of the legume lectins because of its
numerous biological applications, which include probing
normal and tumor cell membrane structures and dynamics,
studying glycosylation mutants of transformed cells, and
yielding preparations of polysaccharides, glycopeptides and
glycoproteins from ConA affinity columns [7].
ConA is a Glc/Man-specific lectin, and is a tetramer at
physiological pH, with each monomer (Mr = 26,000) pos-
sessing one saccharide binding site as well as a transition
metal ion site (S1) (typically Mn2 +) and a Ca2 + site (S2) [8].
The monosaccharide-lectin interactions are, however, rela-
tively weak, and the lectin exhibits both high affinity and
exquisite specificity for oligosaccharides of cell surface
glycoproteins and glycolipids by interactions through
extended binding site [9–11]. The three-dimensional struc-
ture of the lectin at 1.75 A resolution has been determined
by X-ray diffraction analysis [12], and has been further
refined at 1.2 A [13]. The lectin monomer is made up
ed.
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183 175
largely of three antiparallel h-sheets: a six-stranded nearly
flat ‘back’ sheet, a seven-stranded concave ‘front’ sheet and
a five-stranded sheet forming a ‘roof’ over the other two.
This ‘jelly roll’ fold architecture of ConA monomer is
essentially conserved in the family of legume lectins, and
the structures are nearly superimposable, irrespective of the
specificity of the lectins [14]. However, these proteins are
oligomeric, forming either dimers or tetramers. The tetra-
meric legume lectins can be described as ‘‘dimers of
dimers’’. For ConA, the lectin dimer is termed the ‘canon-
ical dimer’, which is characterized by a large 12-stranded h-sheet resulting from the antiparallel side-by-side alignment
of the two six-stranded back sheets. Two such canonical
dimers associate with the central parts of their back sheets in
a perpendicular manner to form the tetramer. Recent struc-
tural data have demonstrated that an essentially conserved
monomeric unit can oligomerize in a variety of ways [3],
and the lectins often differ in their quaternary structures,
particularly among tetrameric legume lectins such as ConA,
the peanut agglutinin and soybean agglutinin (SBA). As
‘natural mutants’ of quaternary association, the proteins of
legume lectin family serve as a paradigm for studies
addressing the effect of subunit oligomerization on the
stability, folding and evolution of lectin structures [15].
The quaternary structure of lectin relates to its activity with
a potential for multivalent binding to cells leading to cross-
linking and aggregation of specific glycoprotein and glyco-
lipid receptors, which, in many cases, is associated with
signal transduction effects [16,17]. In this regard, many
legume lectins have been shown to bind and cross-link with
specific branched-chain oligosaccharides to form unique,
homogeneous precipitates with distinct lattice patterns [18].
These lectins have also been shown to form specific cross-
linked complexes with glycoproteins [19,20], and this
unique specificity of interaction of lectins is mediated, in
part, by the effects of the specific quaternary structures to
which these proteins fold and assemble.
Though extensive studies of lectin–carbohydrate inter-
actions are reported in the literature, relatively little is
known about the protein stability and the folding and
assembly reactions of these multimeric lectin systems. The
conformational stability of oligomeric proteins can be
determined from equilibrium unfolding studies using urea
and GdnHCl, the two reagents commonly employed as
protein denaturants. Analysis of the solvent denaturation
curves using these denaturants can provide a measure of the
conformational stability of the protein [21–23]. The dena-
turant-induced unfolding of oligomeric proteins has mostly
been found to be a multiphasic process with the stabilization
of partially folded intermediates [24,25]. Recently, we have
reported a comparative analysis of denaturation of a gal-
actose-specific plant lectin, SBA and its N-dimethyl and
acetyl derivatives [26]. The urea-induced conformational
change of ConA, and the denaturation of dimeric ConA by
urea at acid pH involving extensive aggregation were
reported earlier [27,28]. Very recently, a comparative study
on the conformational stability of dimeric ConA and winged
bean acidic agglutinin has been published [29], which
showed a reversible two-state unfolding with GdnHCl for
both proteins. We have investigated here the characteristics
of the urea and GdnHCl-induced equilibrium denaturation
of tetrameric ConA involving a three-state model with
distinct dissociation and unfolding reaction steps, and pre-
sented a thermodynamic analysis regarding the oligomeric
structural stability of the protein.
2. Materials and methods
2.1. Materials
Jack bean seeds were purchased from Sigma. Sephadex
G-100 and Superdex 75 were obtained from Pharmacia.
Bio-Gel P-100 was obtained from Bio-Rad Laboratories.
Guanidine hydrochloride (GdnHCl) (>99%) and 8-anilino-
1-naphthalenesulfonate (ANS) were purchased from Sigma.
Urea (AR, E. Merck, India) was further crystallized from
hot ethanol to remove possible contamination by cyanate
ions [30]. The concentrations of stock solutions of urea and
GdnHCl were determined by dry weight or by refractive
index measurements as described [21]. All other reagents
used were of analytical grade.
2.2. Protein purification
Native ConA was purified from jack bean seeds accord-
ing to the published procedure [31] using Sephadex G-100
as an affinity matrix. The purity of the preparation was
checked by polyacrylamide gel electrophoresis under non-
denaturing and denaturing conditions [32], and the assay of
activity was done by hemagglutination assay [33] using 3%
suspension of trypsin-treated rabbit erythrocytes. The con-
centration of ConA was determined spectrophotometrically
at 280 nm using A1%, 1 cm = 13.7 at pH 7.2, and expressed in
terms of monomer (Mr = 26,000) [34].
2.3. Spectroscopic measurements
Absorption spectra were recorded on a Hitachi U 3210
UV–VIS spectrophotometer using Sigma cuvette (volume:
2 ml; pathlength: 1 cm).
Fluorescence spectroscopy was performed on a Hitachi
4010 spectrofluorimeter equipped with a constant temper-
ature cell holder. The spectra were measured at 25 jC in PBS
(pH 7.2) using Sigma fluorimeter cuvette (volume: 2 ml;
pathlength: 1 cm). Relative change (%) of emission wave-
length maximum was calculated on the basis of the change of
wavelength maximum between the native and the unfolded
state taken as 100%. Relative change of fluorescence inten-
sity at 336 nm was determined as percent change relative to
the fluorescence intensity of the native protein in absence of
denaturant.
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183176
Far-UV circular dichroism (CD) spectra were measured
on a JASCO J-720 spectropolarimeter purged with N2, and
equipped with a constant temperature cell holder. The buffer
used during measurement was PBS (pH 7.2). The spectra
were measured at 25 jC in 1-mm-pathlength cell using a
scan speed of 20 nm/min with a response time of 2 s, and
averaged over five scans to eliminate signal noise. The data
are represented as the mean residue ellipticity [h], which is
defined as [h] = 100hobs/(lc), where hobs is the observed
ellipticity in degrees, l is the length of the light path in
centimeters, and c is the concentration in moles of residue
per liter. The values obtained were normalized by subtract-
ing the baseline recorded for the buffer having the same
concentration of denaturant under similar conditions.
2.4. Protein denaturation
The denaturation experiments in urea and GdnHCl were
carried out in PBS (pH 7.2) at 25 jC. For each denaturation
experiment, a known amount of PBS was mixed with a fixed
amount of the protein stock solution and varying amounts of
the concentrated denaturant (both in PBS) in a final volume of
2 ml. The mixture was incubated at 25 jC for 18 h to ensure
that the equilibriumwas achieved. The protein concentrations
were in the range of 0.4–2.0 AM. The steady state fluores-
cence measurements were made at 25 jC with excitation at
280 nm, and emission scanned from 300 to 400 nm.
ANS-binding experiments were performed in the pres-
ence of 2 AM protein, 50 AM ANS, and various concen-
trations of denaturant at 25 jC. The fluorescence spectra ofthe samples were measured with an excitation wavelength
of 370 nm, and the fluorescence intensities at 470 nm were
plotted as a function of the denaturant concentration.
To test the reversibility of denaturation, the protein
solution (40 AM), after complete denaturation in 8 M urea
or 6 M GdnHCl in PBS, was diluted with PBS containing
0.1 mM Mn2 + and 0.1 mM Ca2 +, pH 7.2, to residual
denaturant concentrations of V 0.4 M urea or V 0.3 M
GdnHCl. The mixtures were incubated at 25 jC for up to 4
h. The final protein concentrations during renaturation were
0.4–2.0 AM. Reversibility was checked by fluorescence in
absence and presence of the external ANS probe and the far-
UV CD as described above.
2.5. Size-exclusion chromatography
To verify the size corresponding to tetrameric structure
of the lectin at pH 7.2, ConA was loaded onto a Bio-Gel P-
100 column (1.1�100 cm), which was equilibrated with
PBS containing 0.1 mM Mn2 + and 0.1 mM Ca2 + at room
temperature. Fractions of 2 ml were collected at a flow rate
of 12 ml/h and monitored for protein at 280 nm. To
determine the size of the species in presence of denaturant,
size-exclusion chromatography on Superdex 75 was per-
formed. ConA denatured in 2.7 M urea was loaded onto a
Superdex 75 column (1.1�19 cm), which was equilibrated
with PBS containing 2.7 M urea. The flow rate was 6 ml/h
and the fraction size was 0.8 ml. The protein content in the
fractions, after dilution with buffer, was monitored at 280
nm. The chromatography of fully denatured protein in 8 M
urea on Superdex 75 column was also performed. The
Superdex column was precalibrated in presence of urea
with the following marker proteins: bovine serum albumin
(66 kDa), chicken egg ovalbumin (45 kDa) and soybean
trypsin inhibitor (20.1 kDa). As the native ConA binds to
the Superdex column, the elution behavior of native ConA
on Superdex 75 was examined in presence of 0.2 M
glucose in PBS after equilibration of the column with the
same buffer.
3. Results
3.1. Denaturation of ConA as monitored by steady state
fluorescence
Steady state fluorescence is a useful technique for study-
ing the structure and dynamics of proteins [21,35]. The
intrinsic fluorescence of proteins from Trp residues is an
excellent built-in reporter [36]. There are four Trp residues
per subunit of ConA. The wide range of quantum yields
(fluorescence intensities) and emission maxima of Trp resi-
dues in proteins are attributed to differences in the way the
excited indole ring of Trp interacts with its microenvironment
in different proteins. For example, a fluorescence red shift
occurs as the microenvironment surrounding Trp residues
changes from nonpolar to polar [37]. The fluorescence
spectra of ConA in varying concentrations of urea at pH
7.2 are shown in Fig. 1A. At 0 M urea, the protein exhibits
emission maximum at 336F 1 nm, which gradually red-
shifts with increasing concentration of urea, and finally levels
off at 351F1 nm in z 6.6 M urea, indicative of protein
denaturation and Trp exposure to the aqueous environment.
The urea denaturation curve, in terms of relative change of
emission maximum as a function of urea concentration, is
shown in Fig. 2A. The plot reveals two distinct transitions. A
stable intermediate appears in 2.2–3.3 M urea with an
emission maximum around 339 nm corresponding to
f 20% of emissionmaximum red shift obtained for complete
denaturation. In order to study these transitions inmore detail,
we have performed the same experiment in the presence of an
external polarity-sensitive fluorescent probe, ANS, which
binds nonspecifically to hydrophobic surfaces in many pro-
teins, with enhancement of fluorescence intensity together
with a blue shift of emission maximum (520! 470 nm) [38].
Interestingly, as shown in Fig. 2B, a significant increase in
fluorescence intensity at 470 nm has been observed from 0.6
M urea to f 30-fold in 2.1–3.0 M urea, and then the
fluorescence has decreased gradually to almost negligible
intensity at high concentrations (z 6 M) of urea. The ANS
fluorescence data thus correlate well with two distinct tran-
sitions as observed in the denaturation curve (Fig. 2A), and
Fig. 2. (A) Urea denaturation curve of ConA. Relative change of emission
wavelength maximum was calculated on the basis of the shift of wavelength
maximum (from that in the native state) at different concentrations of urea
in PBS. The difference of wavelength maximum between the native state
and the completely unfolded state was taken as 100%. Each data point
represents average of three determinations. The protein concentration was 2
AM. (B) ANS fluorescence intensity at 470 nm at various concentrations of
urea. ANS (50 AM) was present under the same conditions as for (A).
Excitation wavelength, 370 nm; excitation and emission band pass, 5 nm
each; scan rate, 60 nm/min.
Fig. 1. Fluorescence spectra at 25 jC of ConA (2 AM) in (A) 0 M (a), 2.7 M
(b) and 8 M (c) urea in PBS; and (B) 0 M (d), 1.8 M (e) and 3 M (f)
GdnHCl in PBS. The spectra were corrected for the buffers containing
requisite concentrations of denaturant. Excitation wavelength, 280 nm;
excitation and emission band pass, 5 nm each; scan rate, 60 nm/min.
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183 177
suggest the presence of a hydrophobic equilibrium intermedi-
ate at low concentrations of urea.
There is usually a red shift in the emission of a protein
upon denaturation, though the emission quantum yield
(fluorescence intensity) may either increase or decrease.
If the monitoring of the fluorescence intensity at the blue
or red edge of the emission envelope gives a measurable
signal change, that becomes the preferred measurable
signal for estimating the thermodynamic parameters for
denaturation of proteins [39]. The relative change of
fluorescence intensity at 336 nm as a function of urea
concentration is shown in Fig. 3A. The fluorescence
intensity increases to f 45% in 2.1–3.1 M urea, and then
decreases gradually to level off in z 5.7 M urea to about
the same intensity of the native protein in 0 M urea. This
result, coupled with the previous ANS binding data (Fig.
2B), further corroborates the evidences for an intermediate
structure in the denaturation equilibrium of ConA. The
fluorescence intensity data of ConA is, however, in sharp
contrast with that observed for the denaturation of tetra-
meric SBA, when an intermediate state exhibits similar
relative intensity as of native lectin followed by an increase
in fluorescence at higher concentrations of urea [26]. This
may be attributed to the differences in the environment of
Trp residues in the structures of ConA and SBA.
The denaturation of ConA in GdnHCl exhibits similar
characteristics as in urea. Fig. 1B shows the fluorescence
spectra of ConA in varying concentrations of GdnHCl at pH
7.2. The emission maximum gradually shifts from 336F 1
nm in 0 M GdnHCl to 351F1 nm in z 3 M GdnHCl. Since
GdnHCl is a stronger denaturant than urea, the molar
concentration of GdnHCl required for complete denatura-
tion is significantly lower than in urea. The GdnHCl
denaturation process, using emission maximum as a param-
eter, also depicts two distinct transitions (data not shown)
Fig. 4. (A) Elution profile of ConA (3 mg) from Bio-Gel P-100 column
(1.1�100 cm). (B) Gel filtration on Superdex 75 column (1.1�19 cm) of
ConA (0.4 mg) denatured in 2.7 M urea. Experimental conditions have
been described in the text. Inset: Molecular weight calibration curve. The
columns were precalibrated with standard marker proteins (from left to
right): bovine serum albumin (66 kDa), chicken egg ovalbumin (45 kDa)
and soybean trypsin inhibitor (20.1 kDa). The elution position of the
dissociated ConA is marked by arrow in the calibration curve and the
calculated molecular mass is shown by arrow in the elution profile.
Fig. 3. The denaturation equilibrium transitions measured in terms of the
relative change in fluorescence intensity at 336 nm as a function of (A) urea
concentration in PBS; and (B) GdnHCl concentration in PBS. The percent
change was calculated from the change in fluorescence intensity at 336 nm
at different denaturant concentrations relative to the intensity in the native
state. Each data point represents average of three determinations. The
protein concentration was 2 AM.
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183178
with an intermediate corresponding to an emission maxi-
mum around 340 nm. Fig. 3B shows the denaturation curve
in terms of the relative change of fluorescence intensity at
336 nm as a function of GdnHCl concentration. As in urea,
two distinct transitions have been observed involving an
intermediate with increased fluorescence (f 50%) in 1.4–
2.0 M GdnHCl.
3.2. Dissociation of ConA tetramer and the structural
characteristics
Since ConA is a noncovalently associated tetramer at pH
7.2, one possibility is that the intermediate seen in the
fluorescence-monitored denaturation is a monomeric form
produced by subunit dissociation in presence of denaturant
[40]. To clarify this issue, size-exclusion chromatography
was performed. The gel filtration analysis of native ConA at
pH 7.2 on Bio-Gel P-100 column (Fig. 4A) confirms the
tetrameric structure of the protein (the elution volume corre-
sponds to a molecular mass of 102 kDa). Gel filtration size
analysis in presence of denaturant was performed on Super-
dex 75. When ConA denatured in 2.7 M urea (corresponding
to the intermediate structure in the denaturation curves (Figs.
2A and 3A)) was loaded onto the Superdex column, a protein
peak appeared at a position of molecular mass of 25 kDa (Fig.
4B) that corresponds to the lectin monomer. It is interesting to
note that the completely unfolded monomer in 8 M urea was
eluted in the void volume on Superdex column (data not
shown). This may be due to the extensive randomly coiled
conformations of the denatured subunits leading to an appre-
ciable decrease in elution volume compared with that of a
compact globular conformation [41]. These results strongly
suggest that the denaturation of ConA may be described by a
three-state model, in which the tetrameric structure of ConA
has been lost in relatively low concentrations of denaturant,
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183 179
resulting in the formation ofmonomers. Themonomeric form
then undergoes further perturbation to a completely unfolded
state at higher concentrations of denaturant. It is mentioned
that native ConA was completely bound to the Superdex
column in PBS, which served as an affinity matrix. The gel
filtration of native ConA on Superdex 75 was therefore
performed in the presence of 0.2 M glucose in PBS when
the protein was eluted in the void volume because of its
tetrameric structure. In contrast, the monomeric intermediate
in denaturant did not bind to the Superdex matrix.
Next, in order to study the structural characteristics of the
monomer intermediate species, we measured the CD spectra
of ConA in the absence or presence of denaturant. Fig. 5A
shows the far-UV CD spectra of ConA in 0, 2.7 and 8 M
urea. The native protein exhibits an unusual far-UV CD
band centered around 223 nm, which probably arises from
h-structure with atypical far-UV transitions [27]. The spec-
trum in the presence of 8 M urea shows the loss of
secondary structures for the completely unfolded state of
the protein. When the spectrum of ConA in 2.7 M urea is
compared with those of the native and unfolded proteins, it
is clearly different from that in 8 M urea. The dissociated
Fig. 5. Far-UV CD spectra of ConA (2 AM) at 25 jC in PBS in the presence
of (A) 0 M (—), 2.7 M (- - -) and 8 M (– �–) urea; and (B) 0 M (—), 1.6 M
(- - -) and 3 M (– �– ) GdnHCl. The spectra were measured in 1-mm-
pathlength cell using a scan speed of 20 nm/min, and averaged over five
scans. The data are represented in mean residue ellipticities. The spectra
were corrected for the buffers containing requisite concentrations of
denaturant.
monomer shows the characteristics of h structures but the
shift of CD band to 212–215 nm in this case may be
associated with a change in atypical geometry of the h-sheetof the native protein [27]. The far-UV CD spectra of ConA
in presence of GdnHCl, however, show (Fig. 5B) that the
spectrum in 1.6 M GdnHCl closely resembles that of native
protein, and is completely different from that of the
unfolded form in 3 M GdnHCl. These results demonstrate
that the structure of the dissociated monomer retains the
secondary conformation of the native protein more in
GdnHCl than in urea. Since GdnHCl, as an ionic compound,
can render its structure-stabilizing function [42] at low
concentrations (0–2 M), the secondary structure of the
dissociated monomer is better preserved in this denaturant.
The near-UV CD experiments, however, failed to provide
any conclusive evidence in this regard due to the minute
intensities and changes of the near-UV signals (data not
shown). ANS-binding experiments shown in Fig. 2B sup-
port the notion that the intermediate state resembles a
molten globule [43]. However, the dissociation of tetramer
into monomers from the breakage of the antiparallel h-strand between subunits may also facilitate ANS binding to
h-structural conformation due to an increase in accessible
surface area in the dissociated state. This may lead to
increased fluorescence as the h-structure binds ANS
strongly due to its greater hydrophobicity [44]. Thus, in
the present study, we have concluded tentatively that the
ConA monomer in denaturant assumes a structured, parti-
ally folded tertiary conformation.
4. Discussion
We have characterized the urea and GdnHCl-induced
denaturation of tetrameric ConA at pH 7.2, based on the
results obtained by the fluorescence properties (Figs. 2 and
3), the elution volume from the size-exclusion column
(Fig. 4) and the far-UV CD spectra (Fig. 5). The equili-
brium denaturation pathway of ConA may be represented
by the following three-state mechanism:
N4 X 4NX 4U ðScheme 1Þ
where N4 is the native tetrameric state, N is the structured
monomeric state and U is the completely unfolded state of
the protein. These transitions were found to be completely
reversible as monitored by intrinsic fluorescence and ANS
binding, and by the far-UV CD (data not shown). The
apparent midpoint concentrations (Cm) of these transitions
are summarized in Table 1. In urea, the Cm values for
dissociation and unfolding transitions in terms of fluores-
cence intensity, are 1.4 and 4.5 M, respectively. In
GdnHCl, which is a stronger denaturant than urea, the
corresponding Cm values are 0.8 and 2.4 M, respectively.
These observations conform and lend further support to the
‘‘two-fold rule’’ for urea and GdnHCl denaturation of
Fig. 6. (A) DGdis as a function of urea molarity (top scale) for dissociation
reaction (.) and DGunf as a function of urea molarity (bottom scale) for
unfolding (n) reaction. DGdis values were calculated from the data in Fig.
3A using Eqs. (1)– (3), and DGunf from Eqs. (6) and (7) with T= 298 K. A
linear extrapolation of the baselines in the pre- and post-transitional regions
was used to determine the fraction of the dissociated or unfolded monomer
within the transition region by assuming a two-state mechanism of
dissociation or unfolding reaction. Lines for dissociation and unfolding
transitions were drawn according to Eqs. (4) and (8), respectively, using the
least squares analysis of the data. (B) DGdis as a function of GdnHCl
molarity for dissociation reaction (.) and DGunf as a function of GdnHCl
molarity for unfolding (n) reaction. DGdis and DGunf values were calculated
from the data in Fig. 3B and the lines for dissociation and unfolding
transitions were drawn as described in (A) above. See the text for more
details.
Table 1
Denaturant-induced transitions of ConA
Denaturant Transitions (apparent Cm (M)a)
N4
(native
tetramer)
X N
(structured
monomer)
X U
(unfolded
monomer)
Urea 1.4 4.5
GdnHCl 0.8 2.4
a Values at protein concentration of 2 AM from the fluorescence intensity
data (Fig. 3).
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183180
proteins [45,46]. The apparent midpoints were practically
independent of the protein concentration within 0.4–2.0
AM (data not shown).
The nature of the denaturation equilibrium reveals that
the dissociation transition can be regarded as distinct and
separate from the subsequent unfolding transition (Fig. 3).
We have therefore assumed that Scheme 1 can be analyzed
by the following two independent dissociation and unfold-
ing reaction steps:
N4 X 4N ðdissociationÞ ðScheme 2Þ
NXU ðunfoldingÞ ðScheme 3Þ
4.1. Dissociation reaction of the ConA tetramer
For the dissociation reaction (Scheme 2), the equilibrium
constant (Kdis) is defined by:
Kdis ¼ ½N4=½N4 ¼ 4c3ð fNÞ4=ð1� fNÞ ð1Þ
where [N4] and [N] are the concentrations of native tetra-
meric and structured monomeric proteins, respectively, c is
the total concentration of monomeric ConA, and fN is the
fraction of dissociated ConA, i.e. fN=[N]/([N] + 4[N4]).
The free energy of dissociation (DGdis) at a given concen-
tration of denaturant is then obtained from the equation:
DGdis ¼ �RT lnKdis ¼ �RT ln½4c3ðfNÞ4=ð1� fNÞ ð2Þ
fN can be obtained from the fluorescence data as:
fN ¼ ðDIobs � DIN4Þ=ðDIN � DIN4Þ ð3Þ
where DIobs is the observed change in intensity at a particular
concentration of denaturant in the transition region, and DIN4and DIN are the values characteristic of the tetrameric and
monomeric state, respectively.
The fluorescence intensity data in Fig. 3A,B were ana-
lyzed using Eqs. (1)–(3) to determine the free energy for the
dissociation reaction (DGdis) in urea and GdnHCl, respec-
tively. DGdis depends on the denaturant concentration
according to the linear extrapolation method (Fig. 6A,B)
DGdis ¼ DGdis; aq � mdis½D ð4Þ
where mdis is the slope of the plot of free energy of dissoci-
ation versus denaturant concentration [D], and DGdis, aq is the
free energy of dissociation reaction in aqueous solution in
absence of the denaturant.
The thermodynamic parameters for the dissociation reac-
tion are shown in Table 2. For urea as a denaturant, the
DGdis, aq is estimated to be 7.2 kcal mol� 1 (monomer mass),
which is in good agreement with the value of 7.4 kcal
mol� 1 when GdnHCl is used as a denaturant. From Eq. (2),
using the DGdis, aq of 28.8 kcal/mol of tetramer in urea-
induced denaturation, the dissociation constant of ConA in
water (Kdis, aq) is determined to be 7.4� 10� 22 M3 (Table
2). The ability of denaturants to denature a protein is more
Table 2
Thermodynamic parameters for dissociation and unfolding transitions of ConA
Denaturant Kdis, aq
(M3)
DGdis, aq
(kcal mol� 1)amdis
(kcal mol� 1 M� 1)
Kunf, aq DGunf, aq
(kcal mol� 1)amunf
(kcal mol� 1 M� 1)
DGdis&unf, aq
(kcal mol� 1)a
Urea 7.4� 10� 22 7.2F 0.2 0.8 5.9� 10� 4 4.4F 0.2 1.1 11.6
GdnHCl 1.9� 10� 22 7.4F 0.2 1.6 9.2� 10� 5 5.5F 0.2 2.3 12.9
a Expressed in terms of number of moles of monomer; DGdis, aq and DGunf, aq are calculated from analysis of the data in Fig. 3 by linear extrapolation method
(Fig. 6) at T= 298 K. See the text for details.
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183 181
directly reflected in the m values, and the ratio of the mdis
values for the two denaturants, mdis (GdnHCl)/mdis (urea) is
2.0.
4.2. Unfolding reaction of the ConA monomer
For the unfolding reaction (Scheme 3), the equilibrium
constant (Kunf) is defined by:
Kunf ¼ ½U=½N ¼ fU=ð1� fUÞ ð5Þ
where [N] and [U] are the concentrations of structured
monomer and completely unfolded monomer, respectively,
and fU is the fraction of unfolded monomer, i.e. fU=[U]/
([N]+[U]).
The free energy of unfolding (DGunf) at a given concen-
tration of denaturant is obtained from the equation:
DGunf ¼ �RT lnKunf ¼ �RT ln½ fU=ð1� fUÞ ð6Þ
fU can be obtained from the fluorescence data as:
fU ¼ ðDIobs � DINÞ=ðDIU � DINÞ ð7Þ
where DIobs is the observed change in intensity at a particular
concentration of denaturant in the transition region, and DINand DIU are the values characteristic of the structured
monomer and unfolded monomer, respectively.
The free energy for the unfolding reaction (DGunf) in urea
and GdnHCl has been determined by analyzing the data of
fluorescence experiments in Fig. 3A,B, respectively, using
Eqs. (6) and (7). DGunf is found to vary linearly, and the
linear extrapolation analyses are shown in Fig. 6A,B,
according to the equation:
DGunf ¼ DGunf ; aq � munf ½D ð8Þ
where munf is a measure of the dependence of DGunf on
denaturant concentration [D], and DGunf, aq is the free energy
of unfolding reaction in water.
Table 2 summarizes the thermodynamic parameters for the
unfolding reaction. Values of 4.4 kcal mol� 1 (DGunf, aq) and
5.9� 10� 4 (Kunf, aq) have been determined in urea-induced
denaturation, and the values of 5.5 kcal mol� 1 (DGunf, aq) and
9.2� 10� 5 (Kunf, aq) in GdnHCl-induced denaturation. The
slightly higher free energy value obtained for GdnHCl-
induced unfolding may relate to the occurrence of the more
native-like structured ConA monomer in GdnHCl than in
urea as shown from far-UV CD studies. The ratio of the munf
values for the two denaturants, munf (GdnHCl)/munf (urea) is
2.1, being similar to that for the dissociation reaction. The
consistency of the results in both denaturants supports the use
of linear extrapolation method (Eqs. (4) and (8)), which has
been justified on thermodynamic grounds [47] and which has
the advantage that no assumption about the binding of
denaturants to the native, dissociated and unfolded forms of
the protein are needed [21].
4.3. Structural stability of ConA
The overall structural stability of ConA is obtained from
the combined free energy for Scheme 1. The total free
energy for dissociation and subsequent unfolding reactions
in water (DGdis&unf, aq) of ConA tetramer is therefore
determined to be:
DGdis&unf ; aq ¼ DGdis; aq þ 4DGunf ; aq ð9Þ
Table 2 shows the combined free energy values of the
protein in the two denaturants. For denaturation in urea, the
DGdis&unf, aq value is 11.6 kcal mol� 1, which agrees fairly
well with the value of 12.9 kcal mol� 1 obtained from
GdnHCl-induced denaturation. It is notable that all free
energy (DGdis, aq, DGunf, aq and DGdis&unf, aq) values are
normalized in terms of monomer mass (moles of monomer).
Based on this, the free energy of stabilization of structured
monomer (�DGunf, aq) of ConA relative to the unfolded
monomer is obtained as 4.4–5.5 kcal mol� 1, and the
stabilization free energy for association of tertiary subunits
to tetrameric quaternary structure (�DGdis, aq) is estimated to
be 7.2–7.4 kcal mol� 1. Thus, the free energy of stabilization
of the quaternary structure of ConA relative to the unfolded
state (�DGdis&unf, aq) is determined as 11.6–12.9 kcal mol� 1
(monomer mass). However, Eq. (9) provides that free energy
of unfolding of only 4.4 kcal (per mole of monomer) at the
subunit level, leads to overall free energy of 46.4 kcal (per
mole of tetramer) for the quaternary structure of the protein,
with a contribution of free energy of dissociation of 28.8 kcal
(per mole of tetramer) at the quaternary level. Thus, the
structural stability of ConA is maintained mostly by the
formation of the oligomeric structure.
5. Conclusion
We have characterized the structural stability of ConA, a
tetrameric Glc/Man-specific plant lectin, at both subunit
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183182
tertiary and native quaternary levels of structure formation.
The denaturation equilibrium of the protein in the presence
of urea or GdnHCl displays a three-state mechanism involv-
ing a structured monomeric state between native tetrameric
and unfolded monomeric states. Thermodynamic analysis of
the dissociation and the unfolding reactions has provided a
strong free energy preference (thermodynamic driving
force) for the oligomerization of the protein, and the
structural stability of ConA is governed mostly by associ-
ation of subunits. The enhancement of stability in the
multisubunit lectin in an ordered state when compared with
monomeric forms may arise, at least in part, due to the
energetic tendency to exclude water molecules from hydro-
phobic intermolecular surfaces for achieving the comple-
mentarity of the subunit interfaces [48]. The attainment of
specific quaternary structure for a lectin thus provides its
multivalency (multiple binding sites) for its function as
recognition molecules in biological processes. Further work
on the detailed kinetic investigation of the pathway of
association of ConA following the pathway of refolding is
in progress.
Acknowledgements
This work was supported by research grants from the
Department of Science and Technology, Government of
India, and the University Grants Commission, New Delhi,
India. We thank Professor A. Chatterjee, Principal, for his
support, and Professor S. Ghosh, Head of the Department of
Chemistry, for providing necessary facilities and helpful
discussions. We also thank Dr. S. Basak, Saha Institute of
Nuclear Physics, Calcutta for his kind help in CD experi-
ments.
References
[1] I.E. Liener, N. Sharon, I.J. Goldstein (Eds.), The Lectins: Properties,
Functions and Applications in Biology and Medicine, Academic
Press, New York, 1986.
[2] J.M. Rini, Lectin structure, Annu. Rev. Biophys. Biomol. Struct. 24
(1995) 551–557.
[3] H. Lis, N. Sharon, Lectins: carbohydrate-specific proteins that medi-
ate cellular recognition, Chem. Rev. 98 (1998) 637–674.
[4] N. Sharon, H. Lis, Carbohydrates in cell recognition, Sci. Am. 268 (1)
(1993) 82–89.
[5] N. Sharon, Lectin–carbohydrate complexes of plants and animals: an
atomic view, Trends Biochem. Sci. 18 (1993) 221–226.
[6] N. Sharon, H. Lis, Legume lectins—a large family of homologous
proteins, FASEB J. 4 (1990) 3198–3208.
[7] H. Bittiger, H.P. Schnebli (Eds.), Concanavalin A as a Tool, Wiley,
New York, 1976.
[8] L. Bhattacharyya, S.H. Koenig, R.D. Brown III, C.F. Brewer, Inter-
actions of asparagine-linked carbohydrates with concanavalin A. Nu-
clear magnetic relaxation dispersion and circular dichroism studies,
J. Biol. Chem. 266 (1991) 9835–9840.
[9] D.K. Mandal, N. Kishore, C.F. Brewer, Thermodynamics of lectin–
carbohydrate interactions. Titration microcalorimetry measurements
of the binding of N-linked carbohydrates and ovalbumin to concana-
valin A, Biochemistry 33 (1994) 1149–1156.
[10] D.K. Mandal, L. Bhattacharyya, S.H. Koenig, R.D. Brown III,
S. Oscarson, C.F. Brewer, Studies of the binding specificity of
concanavalin A. Nature of the extended binding site for aspara-
gine-linked carbohydrates, Biochemistry 33 (1994) 1157–1162.
[11] D. Gupta, T.K. Dam, S. Oscarson, C.F. Brewer, Thermodynamics of
lectin–carbohydrate interactions. Binding of the core trimannoside of
asparagine-linked carbohydrates and deoxy analogs to concanavalin
A, J. Biol. Chem. 272 (1997) 6388–6392.
[12] K.D. Hardman, R.C. Agarwal, M.J. Freiser, Manganese and calcium
binding sites of concanavalin A, J. Mol. Biol. 157 (1982) 69–86.
[13] S. Parkin, B. Rupp, H. Hope, Atomic resolution structure of conca-
navalin A at 120 K, Acta Crystallogr. D52 (1996) 1161–1168.
[14] R. Loris, T. Hamelryck, J. Bouckaert, L. Wyns, Legume lectin struc-
ture, Biochim. Biophys. Acta 1383 (1998) 9–36.
[15] V.R. Srinivas, G.B. Reddy, N. Ahmad, C.P. Swaminathan, N. Mitra,
A. Surolia, Legume lectin family, the ‘natural mutants of the quater-
nary state’, provide insights into the relationship between protein
stability and oligomerization, Biochim. Biophys. Acta 1527 (2001)
102–111.
[16] G.L. Nicolson, Transmembrane control of the receptors on normal
and tumor cells, Biochim. Biophys. Acta 457 (1976) 57–108.
[17] K.-N. Chung, P. Walter, G.W. Aponte, H.-P. Moore, Molecular sorting
in the secretory pathway, Science 243 (1989) 192–197.
[18] C.F. Brewer, Multivalent lectin-carbohydrate cross-linking interac-
tions, Chemtracts, Biochem. Mol. Biol. 6 (1996) 165–179.
[19] D.K. Mandal, C.F. Brewer, Cross-linking activity of the 14-kilodalton
h-galactoside-specific vertebrate lectin with asialofetuin: comparison
with several galactose-specific plant lectins, Biochemistry 31 (1992)
8465–8472.
[20] D.K. Mandal, C.F. Brewer, Interactions of concanavalin A with gly-
coproteins: formation of homogeneous glycoprotein-lectin cross-
linked complexes in mixed precipitation systems, Biochemistry 31
(1992) 12602–12609.
[21] C.N. Pace, Determination and analysis of urea and guanidine hy-
drochloride denaturation curves, Methods Enzymol. 131 (1986)
266–280.
[22] M. Yao, D.W. Bolen, How valid are denaturant-induced unfolding
free energy measurements? Level of conformance of common as-
sumptions over extended range of ribonuclease A stability, Biochem-
istry 34 (1995) 3771–3781.
[23] T. Higurashi, K. Nosaka, T. Mizobata, J. Nagai, Y. Kawata, Unfolding
and refolding of Escherichia coli chaperonin GroES is expressed by a
three-state model, J. Mol. Biol. 291 (1999) 703–713.
[24] B. Risse, G. Stempfer, R. Rudolph, H. Mollering, R. Jaenicke, Stabil-
ity and reconstitution of puruvate oxidase from Lactobacillus planta-
rum: dissection of the stabilizing effects of coenzyme binding and
subunit interaction, Protein Sci. 1 (1992) 1699–1709.
[25] R. Jaenicke, H. Lilie, Folding and association of oligomeric and mul-
timeric proteins, Adv. Protein Chem. 53 (2000) 329–397.
[26] M. Ghosh, D.K. Mandal, Analysis of equilibrium dissociation and
unfolding in denaturants of soybean agglutinin and two of its deriv-
atives, Int. J. Biol. Macromol. 29 (2001) 273–280.
[27] M.N. Pflumm, S. Beychok, Alkali and urea induced conformation
changes in concanavalin A, Biochemistry 13 (1974) 4982–4987.
[28] H.E. Auer, T. Schilz, Denaturation of concanavalin A by urea at acid
pH, Int. J. Pept. Protein Res. 24 (1984) 569–579.
[29] N. Mitra, V.R. Srinivas, T.N.C. Ramya, N. Ahmad, G.B. Reddy,
A. Surolia, Conformational stability of legume lectins reflect their
different modes of quaternary association: solvent denaturation studies
on concanavalin A and winged bean acidic agglutinin, Biochemistry
41 (2002) 9256–9263.
[30] O. Frohlich, S.C. Jones, Denaturation of a membrane transport protein
by urea: the erythrocyte anion exchanger, J. Membr. Biol. 98 (1987)
33–42.
[31] B.B. Agrawal, I.J. Goldstein, Protein–carbohydrate interaction: VI.
A. Chatterjee, D.K. Mandal / Biochimica et Biophysica Acta 1648 (2003) 174–183 183
Isolation of concanavalin A by specific adsorption on cross-linked
dextran gels, Biochim. Biophys. Acta 147 (1967) 262–271.
[32] U.K. Laemmli, Cleavage of structural proteins during the assembly of
the head of bacteriophage T4, Nature 227 (1970) 680–685.
[33] T. Osawa, I. Matsumoto, Gorse (Ulex europeus) phytohemaggluti-
nins, Methods Enzymol. 28 (1972) 323–327.
[34] I.J. Goldstein, R.D. Poretz, in: I.E. Liener, N. Sharon, I.J. Goldstein
(Eds.), The Lectins: Properties, Functions and Applications in Biol-
ogy and Medicine, Academic Press, New York, 1986, pp. 35–244.
[35] M.R. Eftink, in: J.R. Lakowicz (Ed.), Fluorescence Quenching:
Theory and Applications, Topics in Fluorescence Spectroscopy,
vol. 2, Plenum, New York, 1991, pp. 53–120.
[36] T.E. Creighton, Proteins: Structure and Molecular Properties, Free-
man, New York, 1993.
[37] E.A. Burnstein, N.S. Vedenkina, M.N. Ivkova, Fluorescence and the
location of tryptophan residues in protein molecules, Photochem.
Photobiol. 18 (1973) 263–279.
[38] J. Slavik, Anilinonaphthalene sulphonate as a probe of membrane
composition and function, Biochim. Biophys. Acta 694 (1982) 1–25.
[39] M.R. Eftink, The use of fluorescence methods to monitor unfolding
transitions in proteins, Biophys. J. 66 (1994) 482–501.
[40] R. Jaenicke, Folding and association of proteins, Prog. Biophys. Mol.
Biol. 49 (1987) 117–237.
[41] K.G. Mann, W.W. Fish, Protein polypeptide chain molecular weights
by gel chromatography in guanidium chloride, Methods Enzymol. 26
(1972) 28–42.
[42] Md.S. Akhtar, A. Ahmad, V. Bhakuni, Guanidinium chloride- and
urea-induced unfolding of the dimeric enzyme glucose oxidase, Bio-
chemistry 41 (2002) 3819–3827.
[43] O.B. Ptitsyn, Molten globule and protein folding, Adv. Protein Chem.
47 (1995) 83–229.
[44] G.V. Semisotnov, N.A. Rodionova, O.I. Razgulyaev, V.N. Uversky,
A.F. Gripas, R.I. Gilmanshin, Study of the ‘molten globule’ inter-
mediate state in protein folding by a hydrophobic fluorescent probe,
Biopolymers 31 (1991) 119–128.
[45] J.K. Mayers, C.N. Pace, J.M. Scholtz, Denaturant m values and heat
capacity changes: relation to changes in accessible surface areas of
protein unfolding, Protein Sci. 4 (1995) 2138–2148.
[46] J.S. Smith, J.M. Scholtz, Guanidine hydrochloride unfolding of pep-
tide helices: separation of denaturant and salt effects, Biochemistry 35
(1996) 7292–7297.
[47] J.A. Schellman, R.B. Hawkes, in: R. Jaenicke (Ed.), Protein Fold-
ing, Elsevier/North-Holland Biomedical Press, Amsterdam, 1980,
pp. 331–341.
[48] F.M. Richards, Areas, volumes, packing and protein structure, Annu.
Rev. Biophys. Bioeng. 6 (1977) 151–181.
