- Email: [email protected]
Thermodynamic bases of the stability of protein structure
Thermochimica Acta, 163 (1990)33-46 ElsevierSciencePublishersB.V.,Amsterdam-PrintedinTheNetherIands
THERMODYNAMIC
BASES
OF THE STABILITY
OF PROTEIN
33
STRUCTURE
P‘L. PRIVALOV Institute of Protein Research. Academy 142292 Pushchino, Moscow Region, USSR One can judge disruption; studyinK
the stability
as a change
The native
structure
ssriables
Kowever,
a protein
structure
a variable, system
because
determining
the state
dependence includes
between
of a system
structure
and the reverse
Experimentally, determined absorbed
only by direct upon heatinq
calorimetric material,
= .J;
Since
in the case
intramolecular heat capacity
thermal
measurements
of this
states
heat
temperature
can be
energy
i.e. only capacity
us to
protein
structure,
of enthalpy
of the thermal
of the excess
by
of this
range:
(1)
of protein
is interested
between
in the
it is necessary
in a solution
the heat effect absorbed
one
process,
of the protein
0040~~31/~/$03.5~
functional
of the native
dependence
as
parameters
dT
of the interaction
enerEy
The
ran.ge and permits
of formation
in the considered
disruption
such a solution
temperature
of
of a
on the macroscopic
of the studied.material,
measurements
tACp(T)>,
system.
of disruption
process
basis
temperature
variables
thermodynamic
the temperature
of
or enthalpy,
and extensive
information
the process
by using
and energy,
two basic
in the considered
in detail
conditions,
on the energetic
of a macroscopic
these
all thermodynamic
analyze
effect
temperature
are the conjurtate intensive
be
by changing
the external
only
by
is a should
pH, and concentration
be obtained
its
of the system.
can be disrupted
the information
can
molecule
state
specifying
such as temperat.ure, pressure, denaturants.
by studying
of its structure
of the macroscopic
protein
intensive
of the USSR,
can be determined
a protein
the disruption
system,
different
of a protein
Since
its denaturation.
regarded
of any structure
thus the stability
macroscopic
of Sciences
dilute
to measure enough
proteins
is negligible.
that should
be measured,
by the protein
molecule
that But
the in
the
upon heating,
01999 Eis~vierSciencePubiishers~.V.
the
is
34
also
especiaily
small,
against
absorption
by the solvent,
TheXWfOre*
studies
development which
of protein
of a special
is now known
the background
which
dominates
thermodynamics
supersensitive
as scanning
of intensive
in dilute required
calorimetric
microcalorimetry
heat
solutions. the technique,
(reviewed
in ref.
11.
I
I
O20
f
4a
30
w
50
70
,
t
30
99
T (“C) Fia,
1.
Lysazpme
Temperature in
dependence
solutions
The native temperature,
structure which
intense
analysis
heat
of
heat
and denatured.
states
does
unstable.
single
extremely
Since discrete
a
not usually Thus,
of
exceed
space
these
states.
elobular
native system
of enthalpy,
between
that
shows
that
protein
are a
(ref. 21. has only
by two surfaces thermodynamic these
they
represents
protein system
two
process,
intermediate
the
(N) and denatured
to extensive
The transition
thermodynamic
shows
all
which
(e.g+
and then disrupts
denaturation
globular
macroscopic
small
states,
correspondins!
by the differences
5%,
a small
conditions
profile
The population
of
critical
certain
Detsi2ed
the
this macroscopic
phase
1).
in
single-domain
can describe
{Fig.
capacity
heat
(ref. 2).
of denaturantsl,
absorption
cooperative
macroscopic
up to a
predominate
native
pW values
on the environmental
presence
states
highly
is stable
absorption
the
macroscopic
of the partial
various
depends
ph, ionic strength, uith
with
states
two
(D), one in the functions
is determined
of
35
A;HfT,pH+J
= HD~T,~H,~il
- HN(T,pH,ai),
12)
= SDcT,pH,ail
- SN(T,pH,ai),
(3)
entropy,
AzSfT,pH,aiJ
and Gibbs
energy,
A~GIT,~H,EL~I = GD(T,pHqai) - GN(T,pH,ai)
(4)
- TA$(T.pH,ail
= A;H(T,pH,ail
The midpoint
of the transition
=
is determined
is used as a variable,
transition
between
the native
determined
esperimentally
peak
1).
(Fig.
transition
A!$TGI
where
=
of the heat
sianificant
absorption
7, the entropy
of
16)
proteins
heat
(Fig.
are equal
specific
capacity
It.
at which
concentrations
of the native
and A$&O.
features
increment
of protein that always
denaturation accompanies
IT)
ACp/T = (?AS)/(irrI,
(8)
that the enthalpy
and denatured
is a this
As
= (6AH)/(aTt,
it follows
of
can be
I$H(TGWTG
One of the most
p
states
is
and denatured
AC
from equation
TG is the temperature
process
then the enthalpp
and denatured
from the area
As follows
(6)
=o
- TA$(T,pH,ai)
If the temperature
by the condition
states
should
and entropy strongly
A;,,,,
=
+ J-; +CpdT. G‘
A;SfTf
= $S(Tf
+ _f';~(~~p)/T~dT G
difference
depend
of the native
on temperature:
19)
(10)
36 ?'!:edeaendence be observed protein
directly
temperature
states
the activity point
obtained
a universal
function
of
for the heat It follows
by heating,
are similar
In all cases
can
temperatures,
corrected
bindine: (ref. 3) (Fid. 2). of protein
on temperature
pH or denaturant
is always
is properly
of denaturants
of view.
denaturation
by varying
The enthalpy
if its value
of lisand
denatured
caused
enthalpy
at different
by studying,
denaturation
concentration.
effect
of the denaturation
that
the
pH variation,
or
from the thermodynamic
the heat capacity
increments
of
are indistinguishable.
I
L
I
8
60
SO
I
20
40 ?(Q
Fig.
2.
Enthalpies
of lysozyme
denaturation
obtained
by various
methods and under different conditions plotted against the temperature of denaturation. Circles indicate solutions without guanidinium chloride fGuHC11: open, denaturation by temperature at fixed pH; filled, denaturation by pii at fixed temperatures. Triangles indicate solutions with GuHCl: filled, denaturation by temperature at fixed concentration of GuHCl; open, denaturatian by GuHCl at fixed temperatures (ref. 3). It appeared increment Fig.
l),
in earlier
of denaturation
studies
4) showed
enthalpy
carried
protein
increases
(Fig. 3).
is likely
to be a linear
the range
from zero
temperature
to 80°C,
However,
the native
(ref.
and that
as temperature
of the native
of temperature
in which
(see dependence
range
protein
in parallel
the heat capacity
function
linear
(Fig. 2).
of the native
do not change
capacity
on temperature
on temperature
capacity
While
the heat
by the observed
out over a broad
that the heat
of the denatured
that
does not depend
and this was confirmed
of the denaturation recent
experiments
state
(in any case state
can be
in
37 practically studiedf, the heat capacity of the denatured state is a nonlinear function, k%hiGh asymptotically approaches some constant level at high temperature.
Linear extrapolation of the
beat capacity of the native state above 8Q'C indicates that A$R
is 1iheJ.y to decrease to zero at about 140'.
Fig. 3, Temperature dependence of the partial specific heat capacity of pancreatic ribonucfease (RNaseI, hen egg white lysozyms fbysl, sperm whale myosfobin (Mb), and catslase from Xhermua thermophil us (CTT) . The flattened curves represent RNase and Lys with disrupted disulfide cross-links and apomyoglobin whose polppcptide chains have a random coil conformation uithout noticeable residual structure (ref. 4), The main consequence of the decrease in the beat capacity increment of denaturation with temperature increase is that the e~~ba~~y
and
entropy
of
protein denaturation are increasina:
functions of temperature that as~~tot~~~l~~ Levels at about 140°C iPia. 4).
approach definite
These levels are likely to be
universal for the specific values of the enthafpy and entropy of all compact globular proteins irefs, 4, 5).
If one neglected the temperature dependence of AbC R RI then the specific enthalpy and
entropy of all globular proteins would CQ~B to the same values, However, in this case one could hardly suggest any physical meaning for these universal values, as the
but at about IlOOC.
enthalpy and entropy functions increase continuously above 11O'C. Since in reality an infinite increase of these functions is improb~b~e~ the assumption that AbC N F decreases with temperature increetse is more justified than the assumption that it is ~n~~p~n~~n~ of temperature.
38
T
(Kf 350
so0
Temperature
Fig. 4. RNase
and Mb when
or constant (ref. 5). Another
their
peculiarity
temperature assume
TR,
to be decreasinu
follows
temperature
$H=O.
from equations
9 and
line)
increases
10 is
become
zero and then change
A$%=GI must
Indeed,
always
for simplicity
exceed let us
which is not very ADC R P' in the vicinity of TS and TR, does
approximation
to the temperature
that
In this case
not depend% on temperature.
(solid
of
decrease upon temperature
TS, at which
at which
functions
line) as the temperature
functions
Temperature
in the first
sensitive
that
and A$
and at some
sign.
of the Aili and A$
is assumed
A$P
(dotted-and-dashed
that both the A$ decrease
dependence
400
from equations
9 and
10
we have: A+(T)
= A;H(TG)
A;SIT)
= fAzH(TGkl/TG
Thus
A@Tl
- (T-T
G
I Ilf
,ADC Np'
+ ep
(12)
In T/TG.
for AiG(T),
= A;H(Tl - TA;SiT)
= UTG-T)tTGlA$TGl
+ jT ADC (T)dT - T.fT A$Zp(Ttdtln RP
TG
TG
T)
39 U ({TG-TJ/TGIAiH(TGl + [T-TGl$Cp
-
TA$
In (TO/T,.
$13)
Then D D TH p1TG - iANH(TGWANCpr TS r TG/IiA~WtTG)i/t~&pTG,
2 + 1t = TG/(ZTG-TW),
and for the difference of these values D D Tg: - TH = USjH(TGW(ANCpTCt
2 TG,
(16)
which is always positive (ref. 6). The temperature shift of the enthalpp and entropy functions is very important for stabilization of the native Protein structure,
500
250 0
._.W.--.
-2Jo -500
Fib?.
5.
0
5a
Temperature dependence of the AgG function of RNase and
?lb when A?C is assumed to be decreasing (solid linef or constant ."P (dotted-and-dashed line) as the temperature increases (ref. 51. Stability of pratein is usually expressed in the Gibbs energy values, since A$Z is the work required for disruption of the native protein structure. estremum of A$
As can be seen from Fig. 5, the
iequation 13) is not very sensitive to the ADC HP dependence on temperature. The maximum A:0 is reached when
40
i.e. at most
temperature
stable
the native
Ts.
and denatured
only by the enthalpy .4t temperatures
should
state
temperature observed,
cold
that while
proceeds
should
with
decrease,
that
below
of
(ref. 3). decrease
and,
of the native
zero stability
of the
temperatures,
at high
of protein
is
T&,
the breakdown feature
proceeds
of the enthalpy
a release
since
the stability
also cause
heat denaturation increase
is
difference
should
the heat denaturation
The distinguishing
of protein
states A$
at two difference
TG, at which
structure.
TS,
therefore,
and a low temperature
at which
thus with
of these
and below decrease
is reached
state
the entropy
is zero, and it is stabilized
difference
One can expect,
native
the native where
states
above
correspondingly, state.
Thus,
at the temperature
of heat,
of the native
of these with
two processes
heat
absorption
and entropy,
i.e. with
T s and TN both
cold
enthalpy
these
functions
is
and
denaturation and entropy change
their
sign. Cold denaturation themodynamic expect
a priori
structure increase
that
could
crucial
test
of cold
proposed
phenomena
predicted
values
below
release
point
upon cooling solutions
various
experimental
under
phenomena
direct
However,
of protein
below
It has been
different
was
of
task,
proteins
denaturation
as
a
the since
were
far
with
heat
the
by supercoolins
O°C and studying
including shown
solvent
regarded
thermodynamic
observation
achieved
techniques,
in an
experimental
solutions.
of protein
was recently
native
experimental
of aqueous
the
one can hardly
i.e.
Therefore
of the protein
was a complicated
aqueous
proteins
denaturation
demonstration
microcalorimetry.
decrease,
of Th for all the studied
the freezing
The direct
from
since
of the ordered
in an entropy
above.
follows
paradoxal,
of the system.
for the correctness
presented
which
appears
the breakdown
result
in the order
demonstration
theory
of pro&ins,
formalism,
for a number
conditions
them by
scanning of globular
lmyoFflobin (ref. 61,
41
apomyonlobin (ref. 7) and staphylococcal
nuclease
(ref. 811 that
upon coolinn, the compact protein structure unfolds, releasing heat (Fig. 61, This process is highly reversible; if the cooled protein solution is heated, the protein folds back, absorbinu the beat.
Ali
the studied
cases
IFirc.
71
are satisfactorily
described by the thermodynamic equations presented above, which confirms that the denaturation of a single-domain protein can be regarded
in
good approximation as a transition between the two
macroscopic states, rihich differ significantly in their heat capacity.,
-IO
IO
Jo
50
70
T SC, Fig, 6, The heat effect observed upon eoolina and heating of apomyo~fobin solution (ref. 7).
subsequent
One of the most general characteristics of protein denaturation is the heat capacity increment, which is the same upon heat, cold, acid, or guanidinium chloride (GuHClf denaturation and is specific for a rtiven protein.
As for the
entbalpy and entropy of denaturation, they can be positive, negative, and even zero, depending on temperature.
However, it is most surprisina that at hiah temperatures the specific values
of these functions are similar for all globular proteins studied, What do the asymptotic values of the denaturation enthalpy and entropy mean and why are they apparently universal for very different proteins? Why should the denaturation enthalpy and entropy depend so much on temperature and consequently have negative values at low temperature?
42
20
0
40
60
80
T (t) Fig, 7. The temperature dependence of the partial heat capacities of metrn~o~~ob~n fMb1 (ref. 61, aponyoalobin fa%b! (ref. 7!, and staphylococcal nucfease (Nase) (ref. 8) in solutions with different pH values. Since neaa$ive
entropy
is a measure of disorder in R system9 the
entropy of denaturation might only mean that at low
temperature the native state is less ordered than the denatured one, notwithstanding the much higher order in the arrangement. of the polypeptide chain in the native protein.
This paradox can be
resolved if one takes into account that the partial entropy of 8 protein molecule in solution is determined not only by its conformation, but also by the state of the solvent in the vicinity of the protein. Dissolution of a nonpolar molecule in water leads to a decrease in the entropy of the system owing to ordering of the water molecule3, while a decrease in this order with :emperature aives the safution
excess
beat capacity.
It is assumed that
unfolding of the compact protein structure with the exposure of internal nonpolar droups ordering
of water
to water
molecules:
should also lead to the
the extent
of this ordering
should
43 decrease with increasing temperature.
An extra energy
expenditure upon the gradual "meltintf" of water ordered by the exposure of nonpolar groups of the unfolded protein is indeed the only reasonable explanation for most of the denaturation heat capacity increments observed upon a temperature increase (ref. 6). On the other hand, if the observed heat capacity increment of denaturat~o~ is caused mainly by melting water ordered by exposed nonpolar groups of protein, one would expect that the influence of these groups on the surrounding water should vanish at some temperature TO. I propose that this should be the temperature at which the specific enthalpies and entropies of denaturation of various proteins become equal, which is about 14GoC according to recent calorimetric studies.
This suggestion was confirmed by
the finding that at about the same temperature the entropy of transfer of various nonpolar substances from the liquid phase to water becomes zero, i.e. these substances no longer affect the order of the water {ref. 61. If one neglects the dependence of the denaturat~on heat capacity increment on temperature, then T G decreases to about llO°C (Pi!?. 4).
This modification does not notably affect the
AiG function fFig, 3).
Therefore, considering protein stability,
in the first approximation one can neglect the temperature dependence of hiCp
and present the A$
and AiS functions in the
folloKinft way: = A+TGI
- IT,-TtA% N p'
1191
A;S(T) = A;S(TGI - fin tTG/T)lA$2p
Here A~H~T~~ and $SITOk
are temperature -independent parts of the enthalpg and entropy of protein denaturatlon that do not include
the effects of water ordering by the protein's nonpolar Ltroups. Then, for the Gibbs energy of stabilization of the native protein structure we have:
44 A;G(TI
= A;HIT) - TA;StT)
In this expression, only the first term is positive, i.e. the enthslpy of protein unfolding in the absence of water salvation effects from the protein's non-polar groups.
This term
represents the temperature-independent contribution of van der Waals and hydrogen bonds in the stabilization of the protein's compact structure.
The second term, which is negative and
increases with temperature, represents the disordering action of dissipative forces.
The third term merely expresses the
contribution of water soivation by nonpolar groups in the stabilization of the protein's native structure.
The most
remarkable feature of this term is that it is negative and that its value decreases to zero a? the temperature increases to TO (Fig. 81.
Therefore at all temperatures below TO, water
salvation by protein nonpolar Ffroups leads to a decrease in the stability of the native compact state; stability is maintained only by the enthalpic interactions, i.e. by van der Waals and hydrogen bonding. This conclusion disagrees with the widespread opinion that water salvation by nonpolar groups is responsible for the hydrophobicity of these groups and for the stability of the compact state of a protein molecule.
Hydrophobic interactions
are usually taken to include, in fact, not onfy the effect of the hydration of nonpolar groups, but also the van der Waals interaction between these groups, which is far from neafisible, contrary to what was previously supposed (for details see ref. 6).
It is important that these two effects are of opposite sign
and of a different range; the van der Waals interaction is short-range, while the hydration effect is long-range. Therefore, the hydrophobic interaction should be attractive at short distances and repulsive at long distances (exceedins the size of a water molecule).
This might be one
of the reasons for
the extreme cooperativity of a tightly packed native domain. Et is remarkable that for a31 proteins studied the value of the A$2 function determining the stability of the native state does not exceed 50 kJ,mof-'(ref. 3).
Since the cooperative
domain usually includes about 100 amino acid residues, it appears
that the contribution of each of the residues in stab~~i~at~o~ of tke native structure does not exceed 500 S per mole sf residue,
FiR”
8.
Contributions of the dissipative forces TAS(TOI and the
wetel: salvation effect ~A~~~~~~T~-T~/TI' Lhe
native
This
state
is five
value
motion at ordered
of
~hbulas
times
protein
less
temperature.
room
native structwe
than Lt
to
fsef.
that
follows
of that
the
stabilization
of
6).
the
energy
protein
of kharmaX ha3
an
onfy because protein is a c~a~erat~~?e
system whose oom~o~e~ts can change their state o&y Zn other words, the stability
cooperatively. deterntined
the system,
by
an
inte~rai
of
m-w+
a
srstem
contribution of all the com~~~e~ts of
The stability af a c~~p~rat~ve domain exceeds bp
almost 20 times the energy of thermal motion; this st;abilitp is quite sufficient ta ensure the existence of the ordered structure of the domain,
Et is this requirement for stability
that seema to determine the lawer limit of the size of the eo~~~~~t~~~e unit:
The must include
ressfdues to be stable amu&
at
least
50 amincl acid
at pb~~s~c~c~~ca~ temperature,
is
46
dissipative
action
cooperation
of these
A% present
of thermal
we know
of the intrazaoiecular
motion,
fittfe
about
interactions
cooperativity
when
all the elements
into a single
unit
seems
tight
and unique
eooperativity looks
interlacing groups
packing
of cooperation Extreme
of the system
to be achieved
of all the short-
only
can provide
and long-range
chain
that
are integrated
in molecules
In other
words,
of the aperiodic
such a structure
in the polypeptide
the mechanism in proteins.
of groups.
is a peculiarity
as if only
but in the effective
interactions.
with
extreme
structure.
It
the complex interactians
is necessary
between
for their
cooperation.
REFEHENCES 1
2
3 4
5
6
7
3
P.L. Privalov and V.V. Plotnikov, Three generations of scanning microcaforimeters; Thermochim, Acta, 139 (1988) 267-277. P.L. Privalov and N.N. Khechinashvili, Thermodynamic approach to the problem of stabilization of globular protein structure, J, Mol. Biol., 86 tl9741 666-684. P.L. Privafov, Stability of proteins. Small globular Advances in Protein Chemistry, 33 (1979) 167-241. proteins, P.L. Privalov, E.I. Tiktopulo, S.Yu. Venyaminov, Yu.V. Griko, G.I, Makhatadze and N.N. Rhechinashvili, Heat capacitp'and conformation of proteins in the denatured state, J. Mol. Biol., 205 (1989) 737-750. P.L. Prisalov and S.J. Gill, _IStability of protein structure and hydrophobic interaction, Advances in Protein Chemistry, 39 (1988) 191-234, P.L. Privalov, Yu.V. Griko, S.Yu. Venyaminov and V,P, Kutyshenko, Cold denaturation of myoglobin, J. Mol. Biol., 190 (1986) 487-498. Yu.V. Griko, P,L, Prisalov, S.Yu. Venyaatinov and S.P. Kutyshenko, Thermodynamic study of the apozayoglobin structure, J. ?lol. Biol,, 202 119881 X27-138. Yu.V, Griko, P.L. Privalov, J.M. Sturtevant and S.Yu. Venyaminov, Cold denaturation of staphylococcal nuclease, Proc, Natl. Acad. Sci. USA, 85 (1988) 3343-3347.
THERMODYNAMIC
BASES
OF THE STABILITY
OF PROTEIN
33
STRUCTURE
P‘L. PRIVALOV Institute of Protein Research. Academy 142292 Pushchino, Moscow Region, USSR One can judge disruption; studyinK
the stability
as a change
The native
structure
ssriables
Kowever,
a protein
structure
a variable, system
because
determining
the state
dependence includes
between
of a system
structure
and the reverse
Experimentally, determined absorbed
only by direct upon heatinq
calorimetric material,
= .J;
Since
in the case
intramolecular heat capacity
thermal
measurements
of this
states
heat
temperature
can be
energy
i.e. only capacity
us to
protein
structure,
of enthalpy
of the thermal
of the excess
by
of this
range:
(1)
of protein
is interested
between
in the
it is necessary
in a solution
the heat effect absorbed
one
process,
of the protein
0040~~31/~/$03.5~
functional
of the native
dependence
as
parameters
dT
of the interaction
enerEy
The
ran.ge and permits
of formation
in the considered
disruption
such a solution
temperature
of
of a
on the macroscopic
of the studied.material,
measurements
tACp(T)>,
system.
of disruption
process
basis
temperature
variables
thermodynamic
the temperature
of
or enthalpy,
and extensive
information
the process
by using
and energy,
two basic
in the considered
in detail
conditions,
on the energetic
of a macroscopic
these
all thermodynamic
analyze
effect
temperature
are the conjurtate intensive
be
by changing
the external
only
by
is a should
pH, and concentration
be obtained
its
of the system.
can be disrupted
the information
can
molecule
state
specifying
such as temperat.ure, pressure, denaturants.
by studying
of its structure
of the macroscopic
protein
intensive
of the USSR,
can be determined
a protein
the disruption
system,
different
of a protein
Since
its denaturation.
regarded
of any structure
thus the stability
macroscopic
of Sciences
dilute
to measure enough
proteins
is negligible.
that should
be measured,
by the protein
molecule
that But
the in
the
upon heating,
01999 Eis~vierSciencePubiishers~.V.
the
is
34
also
especiaily
small,
against
absorption
by the solvent,
TheXWfOre*
studies
development which
of protein
of a special
is now known
the background
which
dominates
thermodynamics
supersensitive
as scanning
of intensive
in dilute required
calorimetric
microcalorimetry
heat
solutions. the technique,
(reviewed
in ref.
11.
I
I
O20
f
4a
30
w
50
70
,
t
30
99
T (“C) Fia,
1.
Lysazpme
Temperature in
dependence
solutions
The native temperature,
structure which
intense
analysis
heat
of
heat
and denatured.
states
does
unstable.
single
extremely
Since discrete
a
not usually Thus,
of
exceed
space
these
states.
elobular
native system
of enthalpy,
between
that
shows
that
protein
are a
(ref. 21. has only
by two surfaces thermodynamic these
they
represents
protein system
two
process,
intermediate
the
(N) and denatured
to extensive
The transition
thermodynamic
shows
all
which
(e.g+
and then disrupts
denaturation
globular
macroscopic
small
states,
correspondins!
by the differences
5%,
a small
conditions
profile
The population
of
critical
certain
Detsi2ed
the
this macroscopic
phase
1).
in
single-domain
can describe
{Fig.
capacity
heat
(ref. 2).
of denaturantsl,
absorption
cooperative
macroscopic
up to a
predominate
native
pW values
on the environmental
presence
states
highly
is stable
absorption
the
macroscopic
of the partial
various
depends
ph, ionic strength, uith
with
states
two
(D), one in the functions
is determined
of
35
A;HfT,pH+J
= HD~T,~H,~il
- HN(T,pH,ai),
12)
= SDcT,pH,ail
- SN(T,pH,ai),
(3)
entropy,
AzSfT,pH,aiJ
and Gibbs
energy,
A~GIT,~H,EL~I = GD(T,pHqai) - GN(T,pH,ai)
(4)
- TA$(T.pH,ail
= A;H(T,pH,ail
The midpoint
of the transition
=
is determined
is used as a variable,
transition
between
the native
determined
esperimentally
peak
1).
(Fig.
transition
A!$TGI
where
=
of the heat
sianificant
absorption
7, the entropy
of
16)
proteins
heat
(Fig.
are equal
specific
capacity
It.
at which
concentrations
of the native
and A$&O.
features
increment
of protein that always
denaturation accompanies
IT)
ACp/T = (?AS)/(irrI,
(8)
that the enthalpy
and denatured
is a this
As
= (6AH)/(aTt,
it follows
of
can be
I$H(TGWTG
One of the most
p
states
is
and denatured
AC
from equation
TG is the temperature
process
then the enthalpp
and denatured
from the area
As follows
(6)
=o
- TA$(T,pH,ai)
If the temperature
by the condition
states
should
and entropy strongly
A;,,,,
=
+ J-; +CpdT. G‘
A;SfTf
= $S(Tf
+ _f';~(~~p)/T~dT G
difference
depend
of the native
on temperature:
19)
(10)
36 ?'!:edeaendence be observed protein
directly
temperature
states
the activity point
obtained
a universal
function
of
for the heat It follows
by heating,
are similar
In all cases
can
temperatures,
corrected
bindine: (ref. 3) (Fid. 2). of protein
on temperature
pH or denaturant
is always
is properly
of denaturants
of view.
denaturation
by varying
The enthalpy
if its value
of lisand
denatured
caused
enthalpy
at different
by studying,
denaturation
concentration.
effect
of the denaturation
that
the
pH variation,
or
from the thermodynamic
the heat capacity
increments
of
are indistinguishable.
I
L
I
8
60
SO
I
20
40 ?(Q
Fig.
2.
Enthalpies
of lysozyme
denaturation
obtained
by various
methods and under different conditions plotted against the temperature of denaturation. Circles indicate solutions without guanidinium chloride fGuHC11: open, denaturation by temperature at fixed pH; filled, denaturation by pii at fixed temperatures. Triangles indicate solutions with GuHCl: filled, denaturation by temperature at fixed concentration of GuHCl; open, denaturatian by GuHCl at fixed temperatures (ref. 3). It appeared increment Fig.
l),
in earlier
of denaturation
studies
4) showed
enthalpy
carried
protein
increases
(Fig. 3).
is likely
to be a linear
the range
from zero
temperature
to 80°C,
However,
the native
(ref.
and that
as temperature
of the native
of temperature
in which
(see dependence
range
protein
in parallel
the heat capacity
function
linear
(Fig. 2).
of the native
do not change
capacity
on temperature
on temperature
capacity
While
the heat
by the observed
out over a broad
that the heat
of the denatured
that
does not depend
and this was confirmed
of the denaturation recent
experiments
state
(in any case state
can be
in
37 practically studiedf, the heat capacity of the denatured state is a nonlinear function, k%hiGh asymptotically approaches some constant level at high temperature.
Linear extrapolation of the
beat capacity of the native state above 8Q'C indicates that A$R
is 1iheJ.y to decrease to zero at about 140'.
Fig. 3, Temperature dependence of the partial specific heat capacity of pancreatic ribonucfease (RNaseI, hen egg white lysozyms fbysl, sperm whale myosfobin (Mb), and catslase from Xhermua thermophil us (CTT) . The flattened curves represent RNase and Lys with disrupted disulfide cross-links and apomyoglobin whose polppcptide chains have a random coil conformation uithout noticeable residual structure (ref. 4), The main consequence of the decrease in the beat capacity increment of denaturation with temperature increase is that the e~~ba~~y
and
entropy
of
protein denaturation are increasina:
functions of temperature that as~~tot~~~l~~ Levels at about 140°C iPia. 4).
approach definite
These levels are likely to be
universal for the specific values of the enthafpy and entropy of all compact globular proteins irefs, 4, 5).
If one neglected the temperature dependence of AbC R RI then the specific enthalpy and
entropy of all globular proteins would CQ~B to the same values, However, in this case one could hardly suggest any physical meaning for these universal values, as the
but at about IlOOC.
enthalpy and entropy functions increase continuously above 11O'C. Since in reality an infinite increase of these functions is improb~b~e~ the assumption that AbC N F decreases with temperature increetse is more justified than the assumption that it is ~n~~p~n~~n~ of temperature.
38
T
(Kf 350
so0
Temperature
Fig. 4. RNase
and Mb when
or constant (ref. 5). Another
their
peculiarity
temperature assume
TR,
to be decreasinu
follows
temperature
$H=O.
from equations
9 and
line)
increases
10 is
become
zero and then change
A$%=GI must
Indeed,
always
for simplicity
exceed let us
which is not very ADC R P' in the vicinity of TS and TR, does
approximation
to the temperature
that
In this case
not depend% on temperature.
(solid
of
decrease upon temperature
TS, at which
at which
functions
line) as the temperature
functions
Temperature
in the first
sensitive
that
and A$
and at some
sign.
of the Aili and A$
is assumed
A$P
(dotted-and-dashed
that both the A$ decrease
dependence
400
from equations
9 and
10
we have: A+(T)
= A;H(TG)
A;SIT)
= fAzH(TGkl/TG
Thus
A@Tl
- (T-T
G
I Ilf
,ADC Np'
+ ep
(12)
In T/TG.
for AiG(T),
= A;H(Tl - TA;SiT)
= UTG-T)tTGlA$TGl
+ jT ADC (T)dT - T.fT A$Zp(Ttdtln RP
TG
TG
T)
39 U ({TG-TJ/TGIAiH(TGl + [T-TGl$Cp
-
TA$
In (TO/T,.
$13)
Then D D TH p1TG - iANH(TGWANCpr TS r TG/IiA~WtTG)i/t~&pTG,
2 + 1t = TG/(ZTG-TW),
and for the difference of these values D D Tg: - TH = USjH(TGW(ANCpTCt
2 TG,
(16)
which is always positive (ref. 6). The temperature shift of the enthalpp and entropy functions is very important for stabilization of the native Protein structure,
500
250 0
._.W.--.
-2Jo -500
Fib?.
5.
0
5a
Temperature dependence of the AgG function of RNase and
?lb when A?C is assumed to be decreasing (solid linef or constant ."P (dotted-and-dashed line) as the temperature increases (ref. 51. Stability of pratein is usually expressed in the Gibbs energy values, since A$Z is the work required for disruption of the native protein structure. estremum of A$
As can be seen from Fig. 5, the
iequation 13) is not very sensitive to the ADC HP dependence on temperature. The maximum A:0 is reached when
40
i.e. at most
temperature
stable
the native
Ts.
and denatured
only by the enthalpy .4t temperatures
should
state
temperature observed,
cold
that while
proceeds
should
with
decrease,
that
below
of
(ref. 3). decrease
and,
of the native
zero stability
of the
temperatures,
at high
of protein
is
T&,
the breakdown feature
proceeds
of the enthalpy
a release
since
the stability
also cause
heat denaturation increase
is
difference
should
the heat denaturation
The distinguishing
of protein
states A$
at two difference
TG, at which
structure.
TS,
therefore,
and a low temperature
at which
thus with
of these
and below decrease
is reached
state
the entropy
is zero, and it is stabilized
difference
One can expect,
native
the native where
states
above
correspondingly, state.
Thus,
at the temperature
of heat,
of the native
of these with
two processes
heat
absorption
and entropy,
i.e. with
T s and TN both
cold
enthalpy
these
functions
is
and
denaturation and entropy change
their
sign. Cold denaturation themodynamic expect
a priori
structure increase
that
could
crucial
test
of cold
proposed
phenomena
predicted
values
below
release
point
upon cooling solutions
various
experimental
under
phenomena
direct
However,
of protein
below
It has been
different
was
of
task,
proteins
denaturation
as
a
the since
were
far
with
heat
the
by supercoolins
O°C and studying
including shown
solvent
regarded
thermodynamic
observation
achieved
techniques,
in an
experimental
solutions.
of protein
was recently
native
experimental
of aqueous
the
one can hardly
i.e.
Therefore
of the protein
was a complicated
aqueous
proteins
denaturation
demonstration
microcalorimetry.
decrease,
of Th for all the studied
the freezing
The direct
from
since
of the ordered
in an entropy
above.
follows
paradoxal,
of the system.
for the correctness
presented
which
appears
the breakdown
result
in the order
demonstration
theory
of pro&ins,
formalism,
for a number
conditions
them by
scanning of globular
lmyoFflobin (ref. 61,
41
apomyonlobin (ref. 7) and staphylococcal
nuclease
(ref. 811 that
upon coolinn, the compact protein structure unfolds, releasing heat (Fig. 61, This process is highly reversible; if the cooled protein solution is heated, the protein folds back, absorbinu the beat.
Ali
the studied
cases
IFirc.
71
are satisfactorily
described by the thermodynamic equations presented above, which confirms that the denaturation of a single-domain protein can be regarded
in
good approximation as a transition between the two
macroscopic states, rihich differ significantly in their heat capacity.,
-IO
IO
Jo
50
70
T SC, Fig, 6, The heat effect observed upon eoolina and heating of apomyo~fobin solution (ref. 7).
subsequent
One of the most general characteristics of protein denaturation is the heat capacity increment, which is the same upon heat, cold, acid, or guanidinium chloride (GuHClf denaturation and is specific for a rtiven protein.
As for the
entbalpy and entropy of denaturation, they can be positive, negative, and even zero, depending on temperature.
However, it is most surprisina that at hiah temperatures the specific values
of these functions are similar for all globular proteins studied, What do the asymptotic values of the denaturation enthalpy and entropy mean and why are they apparently universal for very different proteins? Why should the denaturation enthalpy and entropy depend so much on temperature and consequently have negative values at low temperature?
42
20
0
40
60
80
T (t) Fig, 7. The temperature dependence of the partial heat capacities of metrn~o~~ob~n fMb1 (ref. 61, aponyoalobin fa%b! (ref. 7!, and staphylococcal nucfease (Nase) (ref. 8) in solutions with different pH values. Since neaa$ive
entropy
is a measure of disorder in R system9 the
entropy of denaturation might only mean that at low
temperature the native state is less ordered than the denatured one, notwithstanding the much higher order in the arrangement. of the polypeptide chain in the native protein.
This paradox can be
resolved if one takes into account that the partial entropy of 8 protein molecule in solution is determined not only by its conformation, but also by the state of the solvent in the vicinity of the protein. Dissolution of a nonpolar molecule in water leads to a decrease in the entropy of the system owing to ordering of the water molecule3, while a decrease in this order with :emperature aives the safution
excess
beat capacity.
It is assumed that
unfolding of the compact protein structure with the exposure of internal nonpolar droups ordering
of water
to water
molecules:
should also lead to the
the extent
of this ordering
should
43 decrease with increasing temperature.
An extra energy
expenditure upon the gradual "meltintf" of water ordered by the exposure of nonpolar groups of the unfolded protein is indeed the only reasonable explanation for most of the denaturation heat capacity increments observed upon a temperature increase (ref. 6). On the other hand, if the observed heat capacity increment of denaturat~o~ is caused mainly by melting water ordered by exposed nonpolar groups of protein, one would expect that the influence of these groups on the surrounding water should vanish at some temperature TO. I propose that this should be the temperature at which the specific enthalpies and entropies of denaturation of various proteins become equal, which is about 14GoC according to recent calorimetric studies.
This suggestion was confirmed by
the finding that at about the same temperature the entropy of transfer of various nonpolar substances from the liquid phase to water becomes zero, i.e. these substances no longer affect the order of the water {ref. 61. If one neglects the dependence of the denaturat~on heat capacity increment on temperature, then T G decreases to about llO°C (Pi!?. 4).
This modification does not notably affect the
AiG function fFig, 3).
Therefore, considering protein stability,
in the first approximation one can neglect the temperature dependence of hiCp
and present the A$
and AiS functions in the
folloKinft way: = A+TGI
- IT,-TtA% N p'
1191
A;S(T) = A;S(TGI - fin tTG/T)lA$2p
Here A~H~T~~ and $SITOk
are temperature -independent parts of the enthalpg and entropy of protein denaturatlon that do not include
the effects of water ordering by the protein's nonpolar Ltroups. Then, for the Gibbs energy of stabilization of the native protein structure we have:
44 A;G(TI
= A;HIT) - TA;StT)
In this expression, only the first term is positive, i.e. the enthslpy of protein unfolding in the absence of water salvation effects from the protein's non-polar groups.
This term
represents the temperature-independent contribution of van der Waals and hydrogen bonds in the stabilization of the protein's compact structure.
The second term, which is negative and
increases with temperature, represents the disordering action of dissipative forces.
The third term merely expresses the
contribution of water soivation by nonpolar groups in the stabilization of the protein's native structure.
The most
remarkable feature of this term is that it is negative and that its value decreases to zero a? the temperature increases to TO (Fig. 81.
Therefore at all temperatures below TO, water
salvation by protein nonpolar Ffroups leads to a decrease in the stability of the native compact state; stability is maintained only by the enthalpic interactions, i.e. by van der Waals and hydrogen bonding. This conclusion disagrees with the widespread opinion that water salvation by nonpolar groups is responsible for the hydrophobicity of these groups and for the stability of the compact state of a protein molecule.
Hydrophobic interactions
are usually taken to include, in fact, not onfy the effect of the hydration of nonpolar groups, but also the van der Waals interaction between these groups, which is far from neafisible, contrary to what was previously supposed (for details see ref. 6).
It is important that these two effects are of opposite sign
and of a different range; the van der Waals interaction is short-range, while the hydration effect is long-range. Therefore, the hydrophobic interaction should be attractive at short distances and repulsive at long distances (exceedins the size of a water molecule).
This might be one
of the reasons for
the extreme cooperativity of a tightly packed native domain. Et is remarkable that for a31 proteins studied the value of the A$2 function determining the stability of the native state does not exceed 50 kJ,mof-'(ref. 3).
Since the cooperative
domain usually includes about 100 amino acid residues, it appears
that the contribution of each of the residues in stab~~i~at~o~ of tke native structure does not exceed 500 S per mole sf residue,
FiR”
8.
Contributions of the dissipative forces TAS(TOI and the
wetel: salvation effect ~A~~~~~~T~-T~/TI' Lhe
native
This
state
is five
value
motion at ordered
of
~hbulas
times
protein
less
temperature.
room
native structwe
than Lt
to
fsef.
that
follows
of that
the
stabilization
of
6).
the
energy
protein
of kharmaX ha3
an
onfy because protein is a c~a~erat~~?e
system whose oom~o~e~ts can change their state o&y Zn other words, the stability
cooperatively. deterntined
the system,
by
an
inte~rai
of
m-w+
a
srstem
contribution of all the com~~~e~ts of
The stability af a c~~p~rat~ve domain exceeds bp
almost 20 times the energy of thermal motion; this st;abilitp is quite sufficient ta ensure the existence of the ordered structure of the domain,
Et is this requirement for stability
that seema to determine the lawer limit of the size of the eo~~~~~t~~~e unit:
The must include
ressfdues to be stable amu&
at
least
50 amincl acid
at pb~~s~c~c~~ca~ temperature,
is
46
dissipative
action
cooperation
of these
A% present
of thermal
we know
of the intrazaoiecular
motion,
fittfe
about
interactions
cooperativity
when
all the elements
into a single
unit
seems
tight
and unique
eooperativity looks
interlacing groups
packing
of cooperation Extreme
of the system
to be achieved
of all the short-
only
can provide
and long-range
chain
that
are integrated
in molecules
In other
words,
of the aperiodic
such a structure
in the polypeptide
the mechanism in proteins.
of groups.
is a peculiarity
as if only
but in the effective
interactions.
with
extreme
structure.
It
the complex interactians
is necessary
between
for their
cooperation.
REFEHENCES 1
2
3 4
5
6
7
3
P.L. Privalov and V.V. Plotnikov, Three generations of scanning microcaforimeters; Thermochim, Acta, 139 (1988) 267-277. P.L. Privalov and N.N. Khechinashvili, Thermodynamic approach to the problem of stabilization of globular protein structure, J, Mol. Biol., 86 tl9741 666-684. P.L. Privafov, Stability of proteins. Small globular Advances in Protein Chemistry, 33 (1979) 167-241. proteins, P.L. Privalov, E.I. Tiktopulo, S.Yu. Venyaminov, Yu.V. Griko, G.I, Makhatadze and N.N. Rhechinashvili, Heat capacitp'and conformation of proteins in the denatured state, J. Mol. Biol., 205 (1989) 737-750. P.L. Prisalov and S.J. Gill, _IStability of protein structure and hydrophobic interaction, Advances in Protein Chemistry, 39 (1988) 191-234, P.L. Privalov, Yu.V. Griko, S.Yu. Venyaminov and V,P, Kutyshenko, Cold denaturation of myoglobin, J. Mol. Biol., 190 (1986) 487-498. Yu.V. Griko, P,L, Prisalov, S.Yu. Venyaatinov and S.P. Kutyshenko, Thermodynamic study of the apozayoglobin structure, J. ?lol. Biol,, 202 119881 X27-138. Yu.V, Griko, P.L. Privalov, J.M. Sturtevant and S.Yu. Venyaminov, Cold denaturation of staphylococcal nuclease, Proc, Natl. Acad. Sci. USA, 85 (1988) 3343-3347.