close

Вход

Забыли?

вход по аккаунту

?

Calorimetric Studies of Biopolymers and Aggregates of Phospholipids.

код для вставкиСкачать
Calorimetric Studies of Biopolymers and Aggregates
of Phospholipids
By Theodor Ackermann *
Dedicated to Professor Christoph Riichardt on the occasion of his 60th birthday
The structural transformations that biopolymers undergo in aqueous solution are complex
processes, whose mechanisms can only be explained by the coupling of various partial processes. A spectroscopic analysis of the states of the system often requires considerable effort. In
many cases, calorimetric studies have proven to be sufficient for a general characterization of
the behavior of the system. These measurements provide information on the stability of the
initial state and on the cooperativity of the total process. From the thermodynamic parameters
of the system, knowledge on the structure-determining influence of the various types of interand intramolecular interactions is obtained. This not only applies to solutions of biopolymers
and analogous model substances, but also to solutions of biopolymer complexes with lowmolecular-weight ligands and to aqueous suspensions of self-aggregating phospholipids. Possibilities and limitations of the calorimetric methods of measurement are demonstrated for
typical examples from the numerous polypeptide and polynucleotide systems and the phospholipid bilayer systems that have been studied. In addition, the special problems involved in
carrying out measurements on dilute solutions are pointed out. Here, the latest advances made
in measuring techniques are readily apparent.
1. Introduction
The structure of proteins and nucleic acids is characterized
by a high degree of molecular order, which is not limited to
the covalently bonded atomic building blocks (primary
structure). The helix sequences of the secondary structure,
stabilized by hydrogen bonds, are a characteristic feature of
these biopolymers. The functional diversity of these substances reflects the diversity of the three-dimensional folding
pattern (tertiary structure). Hydrophobic interactions, together with other intermolecular interactions, stabilize the
aggregation of molecules having defined tertiary structures
(quaternary-structure formation). These molecular aggregates, often embedded in the matrix system of the lipid bilayer of biomembranes, are the building elements of biological
functional units.
Phospholipids are the most important structural components of biological membranes. Their fatty acid chains are
more or less closely packed in the hydrophobic region of the
lipid matrix. In the hydrocarbon chains of phospholipid
molecules, rotation about the bond between neighboring
CH, groups is limited. The ideal conformation of this chain
is all trans. However, this ordered state seldom occurs in
membranes. Normally, the CH, groups of lipid molecules
are in a state characterized by several gauche-trans-gauche
kinks. In regard to the molecular dynamics of the lipid system, this means that the mobility of the molecular chains in
biomembranes is much higher than in the ideal quasi-crystalline lipid bilayer. In an exact description of the state of
order of phospholipids, the relative orientation of the neighboring polar head groups must also be considered.
The biological functions of proteins and nucleic acids are
optimal only in a narrow temperature range and for an almost constant composition of the surrounding aqueous
['I
Prof. Dr. T. Ackermann
Institut fur Physikalische Chemie der Universitat
Albertstrasse 23a, D-7800 Freiburg (FRG)
Angew. Chem. In[. Ed. Engl. 28 (1989) 981 - 991
medium. Any temperature change, as well as any pH shift or
change in the solvent composition, leads to a conformational
change of biopolymers associated with a reduction in function (denaturation). Examples of this kind of process include
the helix-random coil transition of polypeptides and the
denaturation of the DNA double helix. The transformation
of a polypeptide chain from the ordered helix state to the
disordered randomly coiled state is schematically shown in
Figure 1. During this conformational change, a variety of
different intermediate states can occur in a group of dissolved macromolecules. These intermediate states must be
Fig. 1. Helix-random coil transition of a polypeptide chain (schematic).
t> V C H Verlagsgesellrchufi mhH, 0-6940
Weinherm, 1989
0570-0833189jo808-0981$02.5010
98 1
considered in any statistical thermodynamic theory of helixrandom coil transitions.“ - 31 Molecular states characterized
by a relatively small number of randomly coiled regions between c o n h u o u s helix sequences are preferred. In the helixrandom coil transformation, concerted transitions of connected helix regions to the randomly coiled state take place.
The overall process is cooperative and is similar to a phase
transition. Therefore, the thermally induced helix denaturation is often compared with a melting process, although it is
not a first-order phase transition.
The transition from the ordered helix state to the disordered state is reflected in numerous physical changes when
solutions are studied. For example, the optical rotatory power of polypeptide solutions is influenced by the helix-random
coil transition. A simple procedure for recording transition
curves, in which the degree of helix transformation is represented as a function of temperature, is based on this effect.I41
In measurements on polynucleotide solutions, the helix denaturation curves can be recorded in a similarly easy manner, because base unstacking, which progresses with denaturation, causes a reduction of the hypochromic effect (i.e.,
increased UV absorption).[’]
Information on the stability of secondary structures with
different units can be obtained from denaturation curves. An
exact analysis of the combined effects of the various types of
weak interactions is only possible if the thermodynamic
parameters of the system are known. The most important
thermodynamic parameters are the enthalpy of transition
and the entropy of transition as well as additional parameters describing the degree of cooperativity of the conformational change. If the observed transformation is coupled
with the release of low-molecular-weight ligands, the relevant values for the binding enthalpy and binding entropy
must be determined.
Phospholipid bilayer vesicles[61 in aqueous suspension
may be considered as lipid-matrix model systems for biological membranes. These lipid systems also change from a quasi-crystalline gel state to a less ordered liquid-crystalline state
above a characteristic transition temperature. This transformation, characterized by a reduction in the packing density
and increased chain mobility, also takes place within a narrow temperature interval. This cooperative process is called
a lipid phase transition!” The characteristic transition temperature not only depends on the state of the polar phospholipid head groups, but also on the number of double bonds
in the hydrocarbon chains of the lipid and the chain length.
The phase transition of lipid bilayers is also influenced by the
addition of other components (e.g., cholesterol). In lipid systems with low water content, in addition to the simply structured bilayer vesicles, other more complicated lipid aggregates can also occur. The areas where these “phases” exist
can be assigned by a lipid phase diagram. Knowledge of the
thermodynamic equilibrium parameters of lipid systems is
required for a molecular theoretical interpretation of the
phenomena observed in these systems.
There are several methods for determining the thermodynamic parameters of biopolymer solutions and phospholipid
systems. Based on studies of simple model systems, several
research groups have prepared extensive experimental data
which have recently been critically studied and summarized
in review articles[*. 1’ and data collections.[’01 It is not possi-
982
ble to discuss all these data within the framework of this
article. Instead, its aim is to present the calorimetric methods
described in Sections 4 and 5 as an important procedure for
studying biopolymers and phospholipid systems and to exemplify the results obtained with these methods.
2. Degree of Conversion and Enthalpy of Transition
The simplest model system for a biopolymer solution with
a temperature-dependent helix-random coil equilibrium is a
group of homopolypeptide molecules with the same chain
lengths. The peptide units of these molecules are called “segments”. If the cooperative interactions are neglected, the
state of this system can be described by a two-state equilibrium (segment in a-helix state (B) 2 segment in randomly
coiled state (A)). The fraction of the segments bound in the
helix state is called the degree of helix formation B.[’ - 31 The
helix formation constant s is the equilibrium constant for
process (1).
’.
. ABBAA . F ? , ABBBA . . .
(1)
The temperature dependence of the quantity 1 - B is
shown in the upper part of Figure 2. The temperature T, (or
!y
II
I
,.i
I
‘m
T-
Fig. 2. The temperature dependence of the degree of transformation I 8 (top)
and the conformational excess heat capacity C, (bottom) of a thermally induced
helix-random coil transition according to Equations (3) and (7).
~
the corresponding 9, of the Celsius scale) that is assigned to
0 = 0.5 is called the transition temperature. If no cooperative
interactions need to be considered, the typical shape of the
curve obtained by plotting 1 - 8 versus T is derived from the
temperature dependence of the equlibrium constant s, which,
according to the laws of thermodynamics, is expressed by
Equation (2).
In Equation (2),
represents the van? Hoffs enthalpy of helix formation o r the enthalpy of transition per
mole of segmental unit or per mole of lipid. At the transition
Angew. Chem. Inl. Ed. Engl. 28 (1989) 981-991
point, s = 1 and 8(1 - O ) = 1/4. From this and the simplifying assumptions already mentioned,f3] Equation (3) is derived.
(3)
Based on Equation (3),
can be obtained from the
slopes of the transition curves recorded by optical methods.
The helix-random coil transition facilitates the addition of
solvent molecules to the dissolved polypeptide molecules
(solvation). Solvation is an exothermic process. The state of
association of the solvent also changes with the state of solvation of dissolved substances. Thus, the overall helix-random coil transition is a process in which endothermic and
exothermic partial processes overlap and, except for a relatively insignificant amount, are compensated in the energy
balance. Therefore, the molar enthalpy of transition calculated for a peptide segment with an average hydrogen bond
dissociation energy"'] of 30 kJ mol- cannot be significantly greater than 5 kJ mol-'. For a comprehensively studied
peptide system (poly(y-benzyl-L-glutamate)
1 3 1 dissolved
in a mixture of dichloroacetic acid and 1,2-dichloroethane;
see Section 7.1), however, a much higher AH,,,. value of
approximately 400 kJ mol-' was obtained from the evaluation of transition curves. Apparently, the stabilizing effects
of neighboring segments of a peptide chain cannot be neglected, because the formation of a helix sequence is strongly
favored by the immediately neighboring segments already in
the helix state. This facilitation means that the neighboring
segments of a helix sequence in peptide chains change from
the helix state to the random state in a concerted process.
The influence of this cooperativity is exhibited in the pronounced sharpness of the transition curves.
'
f
'
3 '
3. Cooperativity Parameters
and Average Cooperative Lengths
Theoretical formulations" - 3 - 1 3 ] allow the derivation of a
quantitative relationship for the temperature dependence of
the frequencies and neighboring distribution of the segments
in the helix and randomly coiled states. This provides information on the average number of united segments still in a
helix sequence a t temperature T,. The number N o of segments that cooperatively transform to the randomly coiled
state a t the transition point is called the average cooperative
length.['' When the cooperativitiy is considered, the true molar enthalpy of transition AH, is not used for AHv,", in Equation (3), but rather the product No . A H l . Equation (4)
applies, since, instead of the individual peptide segment, the
number of segments ( N o ) is used as a molar stoichiometric
unit for the conformational change. Evaluation of transition
curves that are, for example, determined by measurements of
the temperature dependence of optical rotatory power always provides only the product N o . A H , . Normally, the average cooperative length No is not known. Often, it cannot
be theoretically calculated if there is no exact data on the
Angeu Chpm Int Ed Engl 28 (1989) 981-991
chain lengths of the polypeptide molecules studied and
their influence on the sharpness of the transition curves.
In the theory of the helix-random coil transformation,
the cooperativity parameter CJ should be given central importance. If s is the equilibrium constant for the growth
step defined by Equation (l), the product CJ . s is the equilibrium constant for the nucleation step ( 5 ) (a < 1). AppleAAAAA . & . ABAAA . . .
(5)
qz.~ist[~'
showed that there is a simple relationship [Eq. (6)]
between the cooperativity parameter and the average cooperative length N o . For example, for o =
No = 100. If
the value for No is not known, no information on the true
molar enthalpy of transition per segment ( A H , ) can be obtained with the described procedure. However, knowledge of
the AHt values is especially important for an understanding
of the enthalpy balance that contributes to the stability of
secondary structures. Therefore, direct calorimetric procedures should be used to determine AHl values.
The influence of the cooperative interactions discussed
here for the helix-random coil transition of a polypeptide
system is similarly reflected in the thermally induced denaturation of nucleic acid double-helix sequences. Thus, if chainlength effects are considered, the true enthalpies of denaturation per mole of base pairs (bp) for poly- or oligonucleotides
can only be obtained from the different slopes of the socalled melting curves determined by recording UV absorbance of substances with different chain lengths.
The cooperative effect has a particularly dramatic effect
on the transition of a lipid bilayer system from the gel state
to the liquid-crystalline state. The strong similarities of this
change of state with a first-order phase transition lead to the
conclusion that this process is highly cooperative. In the
physico-chemical characterization of these systems, the exact
determination of the true enthalpy of transition can often
only be achieved by means of direct calorimetric measurements. Suitable methods for these measurements are explained below.
4. True Enthalpies of Transition
and the Conformational Excess Heat Capacity
If the temperature is limited to the relative narrow range
of a cooperative conformational transformation, the molar
enthalpy of transition A H , can be considered a constant system parameter. Since the total mole number ntgfor all segments is given, with the differential mole number change
dn = - n,dO and the increase in enthalpy d H = AHdn,
Equation (7) is obtained, where Cl is the conformational ex(7)
cess heat capacity per mole of segment. The temperature
dependence of Cl is schematically represented in Figure 2
(bottom). The symmetric curve shows the proportionality
between the value of C, and the differential quotient dQjd7:
983
Since the total enthalpy of transition of the system does not
depend on the type of intermediate state, the true molar
enthalpy of transition AH can be determined by measuring
the area under the curve shown in Figure 2 (bottom). Since
the transition curve reaches its maximum sharpness at temperature T,, C, ( T )also has a maximum at this point. Consideration of Equations (4) and ( 4 ) gives Equation (8) for T,.
1 (AH)'
Ct(Tm)=-A
4a"'
RT;
If the temperature dependence of the heat capacity of a
solution of biopolymers is exactly recorded, the parameters
AH, and C,(T,,,)can be determined. If these two parameters
are inserted into Equation (8), o and thus No can be calculated based on Equation (6). The prerequisite for an exact determination of the cooperativity parameters is, however, a
distortion-free reproduction of the function C,( T ) , which
can only be achieved if the rate of heating is low and the
measuring arrangement is relatively inert to thermal lag effects.
In cases of model substances with uniform primary structures (homopolypeptides and homopolynucleotides) and
some transfer ribonucleic acids, the conformational transformation can be reversed by cooling via the same path
achieved by heating. In enzyme proteins, deviations from the
renaturation equilibrium
are also rarely observed.
In contrast to polypeptides partial renaturation of native
DNA samples following thermal denaturation can only be
achieved by very slow cooling." Between these two extremes, complete renaturation and nonreversible denaturation, there are numerous hysteresis-like phenomena. The importance of these phenomena in regard to possible information storage effects are discussed in detail by E. Neurnann.[161
Hysteresis effects are observed during variations of pH values and during thermally induced conformational transformations of several polynucleotide solutions and lipid suspensions. An example of a polynucleotide system with thermally
induced hysteresis is the mixed system of poly(deoxyriboadenylic acid) (poly(dA)) and poly(ribothymidy1ic acid)(poly (T)).[' '] Phosphatide acid systems are lipid systems
exhibiting thermal hysteresis.["] The hysteresis effect is also
reflected in the course of the C, ( T )curves. When the equilibrium of hysteresis-free systems is quickly reached, the C , ( T )
curves recorded upon heating and cooling must agree. Since
the primary objective of thermodynamic measurements is
the characterization of changes of state occurring between
equilibrium states, it should always be checked whether the
system studied goes through a series of equilibrium states
during the conformational transformation or whether significant deviations from equilibrium take place. Thus, the
development of heating-cooling calorimeters,[*91which
make it possible to record C , ( T ) curves during heating and
cooling, should receive special attention. If the enthalpy area
under the C , ( T ) curve is constant, the problems related to
distortion-free recording of these curves are insignificant if
only the AH, values are to be determined.
5. Experimental
The first measurements on solutions of polypeptides[20'
and polynucleotides[2'1were made with adiabatic recording
984
calorimeters. For special measurements of some samples dissolved in nonaqueous corrosive solvents, glass calorimeter
In most cases, measurements on
vessels had to be used.r22.231
aqueous solutions can be done with suitable metal calorimeter vessels (e.g., gold-plated silver vessels'241).The most suitable method for recording C, curves is the DSC method
(differential scanning calorimetry).
The principle of this
twin calorimeter method (recording the additional electrical
heating power needed for the compensation of the conformational excess heat capacity when heat exchanges with the
surroundings are largely excluded) is well-known and described in detail in the latest specialized literature on measuring procedures in
However, only some of the
specialized types of commercial DSC apparatuses are suitable for measurements on solutions of biopolymers and
analogous model substances. Special demands on sensitivity
inevitably result from the fact that the major contribution to
the enthalpy increase observed in thermally induced conformational transformations is not caused by the structural
change of the solute but by the high heat capacity of the
solvent. While the increase in enthalpy caused by structural
changes strongly contributes to the total heat capacity in the
study of homogeneous solid samples (e.g., solid polymer materials), the conformational excess heat capacity for biopolymer solutions is normally no more than one or two percent
of the calorimetrically measurable total effect. Thus, the sensitivity for the difference effect to be determined must also be
particularly high when the major contribution to the heat
capacity caused by the solvent is eliminated by compensation
based on the principle of the twin calorimeter. In addition,
only instruments with hermetically sealable sample chambers can be used for calorimetric measurements on solutions
of biopolymers owing to the volatility of the solvent.
The requirements mentioned are met by the DASM-1M
calorimeter (Mashpriborintorg, Moscow, USSR) equipped
with gold vessels, which was developed by P. L. Privalov et
al.12'] An improved version (DASM4M) is now available. With this type of apparatus, measurements on biopolymer solutions with relatively small concentrations (1 2 mg mL-') can be made without particular difficulty. The
Microcal MCI (Microcal Inc., Amherst, MA, USA) developed by X E Brandb et a1.[281is also suitable to make measurements on biopolymer solutions. The Perkin-Elmer DSC
apparatus (Perkin-Elmer Corporation,. Nonvalk, CT, USA)
can also be used to study concentrated lipid suspensions
(5-10 wt YO).
However, with this apparatus, the small sample volumes (20 to 25 pL) have a negative influence on the
calorimetrically measurable effect. In view of the relatively
small number of suitable calorimeters offered, the further
development of DSC instruments for the above mentioned
purpose would be useful.
Figure 3 shows the construction scheme and the twin vessel unit of a DPSC apparatus (differential power scanning
calorimeter) developed in our institute by M . G r ~ b e r t . [In
~~]
Figure 4, as example, a C,( T j curve of a polynucleotide solution recorded with this type of apparatus is shown.[301Since
the AH, value obtained from the area under the C,( T ) curve
corresponds to a small difference effect of two endothermic
effects eliminated by compensation, this AH?value cannot be
assigned a particularly high degree of relative accuracy. Normally, the AH, values determined with the DSC method are
A n ~ e wChem.
.
Inf. Ed. Engl. 28 (1989)981-991
'...'
~
,
Fig 3 A construction scheme (left) and twin-vessel unit
(right) of d DPSC apparatus (according to M Gruher1[2Y]) The twin-vessel unit fixed to the inlet tubes is
surrounded by an ddiabatic shield regulated by means of
multijunction thermocouples A constant temperature of
the surroundings is maintained by means of d thermostated outer shield with d Peltier battery, which surrounds the
adidbatic shield (outer protective end aluminum block
with d presure-resistant lid, lining of the space between
the outer shields polyurethanefoam fillers)
G
[rnJK-'l
I
~
ues increase (see Fig. 5): that is, they show a T, dependence
that corresponds to the dependence of T, values on GC
content in DNA samples determined by Marrnur and
DO~Y.[~]
15 -
--
-- - ----
I
y,r;-
(A
IkJ/mol bpi
reproducible with an accuracy of 2%. In most cases where
thermodynamic data are needed for numerical calculations,
this accuracy has proved to be sufficient.
For the thermodynamic characterization of biopolymer
systems with ligand binding,[3'1 direct calorimetric determination of the enthalpy of binding should also be carried out
if possible. The LKB mixture calorimeter[321(model 10 7002; LKB, Broma, Sweden) is suitable for these kinds of experiments. Such measurements have also been made by several
research groups.[331
6. Results from Measurements
on Polynucleotide Solutions
Calorimetric measurements on nucleic acids and synthetically prepared poly- and oligonucleotides have been made.
The results are summarized in the data collection mentioned
previously.[''] Here, several typical examples will be briefly
described and explained.
6.1. Deoxyribonucleic Acids
The magnitude of the AHl values per mole of base
pairs (mol bp) for the thermally induced denaturation of
dissolved deoxyribonucleic acids is approximately 35 kJ/
mol bp.134-371
For a number of different DNA samples, the
dependence of AH, values on guanine/cytosine (GC) content
has been studied.f38.391 From the AH, values determined for
different T , values, the entropy of denaturation (AS, values)
and the AG, values can be calculated and summarized in a
state diagram."'' With increasing GC content, the AH, valAngew Chem. Inl.
Ed. Engl. 28 (1989) 9 8 t - Y 9 /
1
0
0
42
I
50
XG[
-
100
l%I
Fig. 5. Dependence of the DNA enthalpy ofdenaturation A H , on the guaninecytosine content x of the samples [34, 391.
Exact measured values of the enthalpy of denaturation
can therefore be used as an additional analytical criterion for
the relative base composition of DNA samples. However,
the dependence of AHl values on the GC content is not as
pronounced as might be expected in a double helix stabilized
solely by hydrogen bonds. From all the AHl values measured
on DNA samples, it can be concluded that the hydrogenbond-enthalpy effect does not contribute significantly to the
stabilization of secondary structures and that the doublehelix structure is primarily stabilized by means of base stacking. This is understandable if it is considered that solvated
single strands are formed during denaturation. The AH values for the partial processes of base unstacking listed in
Table 2 (Section 9) are approximately 25 kJ/mol bp. The
comparison with the value for the enthalpy of DNA denaturation (35 kJ/mol bp) implies that approximately 70 % of the
total enthalpy effect of the double helix denaturation can be
attributed to base unstacking. Thus, the hydrogen-bond systems are of primary importance in regard to the specificity of
base pairing, but only of secondary significance for the stability of secondary structures.
In the discussion of AH, values, it is to be noted that the
thermodynamic data of dissoIved electrolytes are normally
related to the standard state of ideal dilution.t401Calorimetric measurements on polynucleotide solutions of different
polynucleotide concentrations have shown that the AH, values of these substances exhibit only a small degree of concen985
tration dependence.[411The deviations of the AH: values
extrapolated to the state of ideal dilution and the measured
AHt values from moderately concentrated solutions are within the stated limit of error of f 2 %. Therefore, the experimentally determined AH, values can be referred to as standard values of first approximation. U p to now, however, this
simplified assumption has proven valid only for aqueous
polynucleotide solutions.
By means of measurements on selected D N A samples, it
was possible to show that the DSC method could also be
used in the characterization of highly repetitive sequences in
eukaryotic DNA. The repetitive sequences occurring in certain satellite DNAs can be assigned to different discrete
peaks in the C,(7') diagram of the samples studied. From the
T,,, values determined with this method, the average base
composition of the repetitive sequences can be obtained.[421
6.2. Ribonucleic Acids
The specific transfer ribonucleic acids (tRNAs), which are
indispensable for protein biosynthesis, exhibit a characteristic base-pairing pattern (cloverleaf model) and a distinctive
tertiary structure (see Fig. 6). Their molecular weights differ
3 - O H end
1
5'-phosphale end
Viroids are circular, single-stranded ribonucleic acids
(molecular weight of approximately 120000), which as infectious particles lead to a number of plant diseases.i471Detailed information on the secondary structure of these ribonucleic acids was derived from precise measurements of
the temperature dependence of UV absorption of dissolved
viroid
The model of the secondary structure
adapted to the denaturation behavior of the viroid samples[491is the basis for an estimate of the enthalpy of denaturation, which can be calorimetrically determined using the
methods described in this paper. For the CPFV (cucumber
pale fruit viroid), enthalpies of denaturation of 3930 i
150 kJ mol-' were measured with the DPSC method.[501
This value agrees relatively well with the value of
4080 kJ mol - derived from the model secondary structure.
Thus, the model of secondary structure of CPFV suggested
by Riesner et al.i491is essentially correct.
6.3. Model Substances (Poly- and Oligonucleotides)
The double helix-random coil transition of equimolar
mixtures of complementary homopolyribonucleotides is
characterized by a high degree of cooperativity and is easily
accessible to studies using the calorimetric methods of
measurement described. This is exemplified by the results of
measurements on aqueous solutions of polyriboadenylate
(poly A) and polyribouridylate (poly U), a system that has
been comprehensively studied.[Z'.30.s1. T he formation
and thermally induced degradation of triple helices can be
followed relatively well by means of measurements on this
system. In Figure 7 Ct(7')curves, recorded at pH 6.8 and a
t
ct
Fig. 6. Simplified drawing of the secondary structure (tRNAAl".left) and the
tertiary structure (tRNAPhe,right) of transfer ribonucleic acids.
only slightly. However, characteristic differences in the secondary structures of different specific transfer ribonucleic
acids can be recognized in the number of base pairs determined by various physicochemical methods.[43JIn the largely reversible thermally induced denaturation of dissolved
tRNA samples, the endothermic degradation processes of
the tertiary and secondary structures overlap.[44.4s1 Despite
this overlapping of two effects, a significant, nearly linear
dependence of the enthalpies of denaturation on the number
of hydrogen bonds given by the secondary structure was
determined by means of calorimetric measurements on solutions of different specific ~ R N A s . [ The
~ ~ IAH, value for the
denaturation of tRNA samples is approximately
1500 kJ mol- ' per mole of tRNA. This value corresponds to
an average of approximately 20 base pairs[431and is thus
significantly higher than the calculated values, which were
determined using the base-pair denaturation increments listed in Fable 2 (Section 9).[461Based on this, a substantial proportion (approximately 30 %) of the total enthalpy of transition of the tRNA sample must be used for the denaturation
of the tertiary structures.
986
I J K-'I
T ['CI
-
Fig. 7. C,(T)curve of an eqnimolar mixture (0.0078 M) of polyriboadenylate
(poly A) and polyribouridylate (poly U). The experimental conditions are stated
in the text.
Na@concentration of 0.5 mol L-I , are presented. The transition peak at 54.3 "C corresponds to the "disproportionation" (9).
2 poly(A
+ U) z=? poly(A +'2U) + polyA
(9)
During transformation, a triple helix composed of poly
(A + 2 U ) is formed, accompanied by the release of a molar
equivalent of poly A. The pronounced maximum of the heat
capacity at 72.1 "C corresponds to a transformation in which
the components polyA and polyU are released from
the triple-helical complex [Eq. 101. The model system
poly(A
+ 2 U ) s polyA + 2 polyU
(10)
A n p r . Chrm. Int. Ed. Engl. 28 11989) 981-991
poly(1 + C), that is, an equimolar aqueous mixture of
polyriboinosylate (poly I) and polyribocytidylate (poly C)
has also been largely characterized with calorimetric metho d ~ . [ ~ ~ ~
From the AH, values of polynucleotide systems, which are
related to a double helix denaturation, information on the
relative stability of different complementary base combinations can be obtained, because the corresponding entropies
of transition differ only slightly.[’0] For example, the AH,
value[541of 30 kJ/mol bp determined for alternating polyd(A-T) fits well into the explanations in regard to the system described in Figure 5. In this context, the experiments on
the thermodynamic characterization of the “wobble base
pairs” GU postulated by Crickr551(e.g., see Fig. 6, left)
should be noted.[561Using spectroscopic and calorimetric
measurements on equimolar aqueous mixtures of poly (A,
G) with poly U (also poly (C,U) with poly G), it was possible
to show that the AH, vaIue of 28 kJ/mol bp determined for
the guanine-uracil base combination differs only slightly
from the value for the AU base pair (33 kJ/mol bp). In these
cases, the different stabilities of the base combinations are
reflected in the different values of the entropies of transit i ~ n . ‘At
~ ~the
] standard temperature of 25 “C, the ribonucleotide secondary structures are also stabilized by the guanine-uracil base combination, although this base combination is
characterized by a particularly high dynamic flexibility.[571
The poly A released during transformation based on Equation (9) is not in the randomly coiled form but rather in a
partially ordered secondary structure. The enthalpies of
transition assigned to the thermally induced degradation of
this secondary structure can also be determined with calorimetric measurements.[’01In the determination of enthalpies
of transition, the enthalpy changes caused by changes in the
secondary structure of single-stranded polynucleotides must
in any case be taken into account.
The thermodynamic characterization of oligonucleotide
systems[581is important because these systems provide an
especially good basis for the study of the effects of certain
base sequences on the relative stability of various secondarystructure elements (e.g., hairpin loops, see Fig. 6). Sequencespecific effects can be better determined by studying short
oligonucleotide sequences than by determining the characteristics of polynucleotides. With a suitable set of thermodynamic parameters of oligonucleotides, the stability of certain
models of secondary structures (e.g., of viroids) can be approximated. By means of measurements on oligonucleotide
systems, it was shown that there is no significant difference
between the enthalpies of base unstacking of ribonucleotides
and deoxyribonucleotides with identical base
The
stabilizing or destabilizing effects of dangling ends on the
stability of helix structures can be quantitatively determined
by denaturation experiments on solutions of double-helical
oligonucleotide complexes.[591It should be noted that a relatively small slope of the transformation curves is expected
even for extremely high degrees of cooperativity, because the
average cooperative lengths [see Eq. (4)] cannot be greater
than the sequence length of the molecule. For the same reason, only broad and flat C,(r) curves are obtained from
calorimetric measurements on oligonucleotide systems.
Thus, the amount of substance used for calorimetric determination in such systems is much greater than the quantity
Angeu Chem. Int. Ed. Engl. 28 (1989) 981-991
used for measurements on polynucleotide systems. Despite
this disadvantage, calorimetric measurements should always
be carried out when oligonucleotide systems are studied,
provided that there is a sufficient quantity of sample substance. The calorimetrically measured AH, values do not
depend on the type of model (two-state model or multistep
model) that must be assumed when transition curves are
evaluated. A process that corresponds to the simple twostate model exists if the calorimetrically measured total molar enthalpy of transition agrees with the van’t Hoff enthalpy
of transition AHV.”,derived from the transition curves. This
applies, for example, to the deoxyribonucleotide system
d(GC),
= 240 kJ mol-’) studied by Bresluuer et
a1.[601If, on the other hand,
is less than the calorimetrically measured total enthalpy of transition, the system goes
through a series of intermediate states (for example,
rA7U7)r6’1during the thermally induced helix-random coil
transition. Further measured data and experimental details
have already been summarized.[581
6.4. Ligand Binding (Intercalation)
A typical example of an interaction of dissolved biopolymers with low-molecular-weight ligands is the binding of
ethidium and acridine dyes to deoxyribonucleic acids. Here,
the noncompetitively bound ligands (dye cations) are largely
shielded from their surroundings by intercalation.[62.631 In
contrast, the competitively (externally) bound ligands can be
displaced by competitor salt ions. These types of binding
differ in the values of the binding constants; the standard
entropy of binding is greater for external binding. In regard
to the van’t Hoff enthalpies of binding (- 30 kJ mol- I )
derived from the temperature dependence of the binding
constants, both types of binding are similar. For the thermodynamic characterization of the binding of dyes to deoxyribonucleic acids, systematic studies with physicochemical
methods have been carried out by H. Zimmermnnn and his
The enthalpies of ligand binding can be
calorimetrically determined with a suitable mixing calorimeter and the resulting AH values agree relatively well with
values derived from binding studies under comparable conditions. This agreement is remarkable, because often difficulties can occur in the calculation of formal standard enthalpies of binding[641from the temperature dependence of
the binding constant owing to the complexity of the binding.
Calorimetrically determined formal standard enthalpies of
binding are temperat~re-dependent.[~
‘I This (“apparent”)
temperature dependence can be primarily attributed to the
overlapping of the two binding processes mentioned above
and to the effects on the electrolyte equilibria and buffer
dissociation equilibria associated with these processes. Detailed discussions of these temperature-dependent binding
equilibria have been presented el~ewhere.1~
’* 641
7. Results from Measurements on Solutions
of Polypeptides and Proteins
The results from measurements of enthalpies of protein
denaturation have been described and summarized in ta987
b l e ~ . [ Some
~ ~ I of the most important aspects of these studies
are explained here using several examples.
7.1. Polypeptide Systems
Several authors[2Z* have reported exact calorimetric
measurements on solutions of the polypeptide poly(y-benzyl-L-glutamate) 1 in a mixture of dichloroacetic acid and
0
I1
-C-CH-NHI
7%
7%
1
C-0-CH,C,H
II
0
n
1,2-dichloroethane. The nonaqueous solvent mixture containing the strongly solvating component dichloroacetic acid
was chosen for this model system, because relatively simple
assumptions could be made for the polypeptide-solvent interactions (quasi-stoichiometric solvation). This system has
already been comprehensively studied by Zimrn, Doty, and
Z S O [ ~ and
'
is characterized by the enhanced formation of a
polypeptide ct-helix (see Fig. 1) when samples are warmed.
At low temperatures, solvated polypeptide molecules are
largely in the randomly coiled state. Thus, a necessary precondition for helix formation is the release of solvent
molecules (dichloroacetic acid) from the polypeptide, thereby making available hydrogen-bond-acceptor and -donator
sites. Therefore, the overall process of a-helix formation in
this solvent mixture is endothermic. For the thermodynamic
characterization of this polypeptide system, all preconditions for the use of the equations developed by Zimrn and
BraggI2l for quantitative determination of the transition
curves are satisfied. The optical rotatory power and the heat
capacities (measured with a suitable adiabatic calorimeterIZ3]) of a solution of poly(y-benzyl-L-glutamate) 1
(0.257 mole per kilogram of solvent) in a mixture of
dichloroacetic acid and 1,2-dichloroethane (81 :19) result in
a temperature curve corresponding to the scheme presented
in Figure 2. The enthalpy of transition determined from this
data is dependent on the concentration of the dissolved
poylpeptide.[22]The A H , value of 3.99 kJ per mole of segment extrapolated to standard ideal dilution agrees well with
the AH, value determined from the temperature dependence
of the optical rotatory power, where the length of the polymer chain was taken into consideration. From the maximum
value of the conformational excess heat capacity, it was possible to calculate the cooperativity parameter g for this system from Equation (8). A value of G =
was determined
for the standard state of ideal dilution. This value agrees well
with the value approximated by Zimm, Doty and ZSO.~~]
The
effects of solvent composition on the magnitude of the A H ,
values has also been studied.[67]It should be noted that not
all polypeptide systems are suitable for a physico-chemical
characterization using calorimetric methods. For example,
the helix-helix transformation of poly(L-proline) comprehensively studied by Engel et
cannot be induced by a
change in temperature ( A H z 0).
988
7.2. Proteins
The free energy of denaturation of proteins (Gibbs free
energy change associated with protein unfolding) is an important parameter for the evaluation of the ability of organisms to live under extreme conditions (thermophilic, psychrophilic behavior).[691 In this context, the
determination of enthalpies of denaturation of dissolved
proteins is of fundamental importance.165*
701
The
values of the calorimetrically determined enthalpies
of denaturation range from 1134 kJ mol-' (pepsinogen,
M , = 38000 g mol-') to 97 kJ mol- (tryptophane synthase, a subunit, mutant (Ser49), M , = 29000 g mol-'). The
numerous interactions between various molecular regions
are reflected in the size of the AH values. This applies especially to the systems with protein-protein association[7'1
that have also been thermodynamically characterized (example: tobacco mosaic virus protein TMVP[721).
In the study of
association or polymerization of polymer subunits, both
positive and negative values for the standard entropies of
transition have been found.17z1
Another important aspect of protein thermodynamics is
the protein-Iigand intera~tion."~'Of the numerous ligand
binding systems that have been studied, but which cannot be
discussed in detail here, five kinds of interactions are particularly noteworthy, namely, the binding of inorganic ions
including H' ions, the binding of denaturing agents (e.g.,
urea),[741the hapten-antibody interaction," 5 *761 the interaction of proteins (e.g., muscle proteins) with ADP and
ATP,[77,781 and the binding of NAD' and analogous ligands
to enzymes (e.g., lactate dehydrogenase from porcine heart
mus~Ie).[~~]
The interaction between protein subunits[80] and the
protein nucleic acid interactions L811are considered to be especially important. A typical example of the interaction of
protein subunits is that between S-peptide and S-protein of
ribonuclease S'.
The results from these studies are the basis for the investigation of model concepts on the structure-function relationship of protein systems.
8. Aqueous Lipid Systems
The discussion of thermodynamic parameters of lipid suspensions is limited to the explanation of a few typical examples. For measurements on lipid systems with low water contents, the reader is referred to a recent summary19]and to the
original
8.1. Transition Temperatures, Enthalpies of Transition,
and Apparent Molar Heat Capacities
In Table 1 the thermodynamic parameters for 16 different
phospholipids, each containing two identical hydrocarbon
chains, are summarized. The abbreviations almost always
used in the relevant literature[6.81(e.g., DSPC for distearoyl
phosphatidylcholine) are derived from the trivial names of
the fatty acids and of the head-group components linked to
the phosphate groups (see footnote to Table 1). The T,,, valAngew. Chem. Int. Ed. Engl. 28 (1989) 981-991
Table I . Transition temperatures (T,,,,)and transition enthalpies ( A H , ) of pretransition as well as transition temperatures, enthalpies of transition, and entropies of transition (T,,,,,A H 2 , ASz) of the main transformation of phospholipids with two identical hydrocarbon chains in dilute aqueous suspension
(taken from A . Blurne [9,83, 841).
Lipid [a]
Tm,
["C]
Tm2
["C]
DMPC
DPPC
DSPC
DAPC
DBPC
DLPE
DMPE
DPPE
DSPE
DAPE
DLPA (pH 6)
DMPA (pH 6)
DPPA (pH 6)
DMPA (pH 12)
DPPA (pH 12)
DHPC
DHPE
DHPA (pH 6)
DHPA (pH 12)
15.3
35.5
51.0
62.1
24.0
41.5
54.3
64.1
12.5
30.5
49.9
63.9
72.5
AH1
[kJ mol-'1
5.4
6.1
1.5
7.1
81.1
33.0
33.5
52.2
65.0
22.4
43.1
43.5
68.5
73.3
53.8
5.9
AHz
[kJ mol-'1
As2
[J mol-' K-'1
27.2
41.9
45.6
51.5
62.4
18.0
21.6
36.0
43.9
51.1
14.2
23.8
33.1
17.2
23.9
35.6
31.8
29.1
24.3
91.5
115.9
139.4
157.4
180.4
59.4
85.4
106.7
121.2
144.4
46.5
73.2
97.9
58.2
14.5
111.8
93.3
85.8
74.5
[a] DMPC, dimyristoylphosphatidylcholine; DHPC, 1,2-dihexadecyl-snglycero-3-phosphocholine;DHPE, 1,2-dihexadecyl-sn-glycero-3-phosphoethanolamine; DHPA, 1,2-dihexadecyl-sn-glycero-3-phosphoric
acid.
ues and the enthalpies of transition ( A H , ) increase with increasing hydrocarbon chain lengths.
Figure 8 shows, as example, the DSC curves of aqueous
suspensions of different phosphatidylethanolamines record-
DMPE
I
DPPE DSPE
DAPE
TIOCl-
Fig. 8. C , ( n curve of aqueous suspensions of five phosphatidylethanolamines with different hydrocarbon chain lengths (c = 1 mg mL- ').
ed under identical conditions and at a concentration of
1 mgmL-'. The entropies of transition (AS,) shown in
Table 1 vary systematically in an analogous way. The magnitudes of these A S values are much lower than the entropies
of melting of corresponding long-chain fatty acids or corresponding pure hydrocarbons. This finding clearly shows that
the degree of order of the lipid bilayers of liposomes[71in the
liquid-crystalline phase is significantly greater than it is in an
isotropic liquid phase. In some lipid systems (e.g., DPPC) a
so-called pretransition 16. I" is observed. Where it was possible to determine enthalpies of transition for this transformation, these values ( A H , ) are also listed in Table 1 . The magnitude of the A H 2 values depends not only on the length and
Angen. Chem. (nt. Ed. Engl. 28 (1989) 981-991
degree of saturation of the hydrocarbon chains, but also on
the type of phospholipid head groups. As a comparison of
the values for distearoyl phosphatidylethanolamine (DSPE)
and distearoyl phosphatidylcholine (DSPC) listed in Table 1
shows, the T,,, values of phosphatidylethanolamines (PEs)
are greater and the A H , values are smaller than the corresponding values for phosphatidylcholines (PCs) having identical hydrocarbon chain lengths. This finding, which is reflected in the different values of methyl-group increments[841
of A H , values, indicates that the PE molecules are more
closely packed than corresponding PC molecules in the liquid-crystalline phase because of a possible hydrogen-bond
linkage of NH: head-group components. For quantitative
determination of the influence of the size of head groups,
Hinz et a1.[86]made calorimetric measurements on suspensions of different synthetically prepared glycolipids. It is remarkable that for the phospholipids with two different fatty
acid chains largely present in biomembranes, a significant
dependence of the AHl values on the binding position of
these fatty acid chains at the glycerol backbone is observed.IS7]Since, for steric reasons, a line joining the three
glycerol carbon atoms cannot run parallel to the surface of
the bilayer, the effective number of methylene groups in the
lipophilic zone is reduced by approximately two units (effective chain shortening effect). Based on this, a lipophilic zone
with a largely regular arrangement of hydrocarbon chains
with identical effective chain lengths results for
MPPC. In contrast, the PMPC bilayer is characterized
by a packing of hydrocarbon chains with different
effective lengths, resulting in a lower AH2 value. Finally, it
should be noted that with the DSC method, the apparent
molar heat capacities[", 89,9 5 *961 defined by Equation (1 1)
can be determined and that information on the state of order
of lipid molecules can also be derived from these apparent
molar heat capacities.[90*9 1 *9 5 * 961 Contributions to the apparent molar heat capacities come not only from the excitation of internal degrees of freedom of the molecules, but also
from interactions with the surrounding aqueous medium.
The negative contributions come from the hydration of the
polar groups and the change in the association structure of
waterE9'] caused by hydration. Normally, additional positive
"C, contributions can be attributed to the hydrophobic interaction of nonpolar groups with the bordering regions of
the water structure.193*941
8.2. Multicomponent Systems, Lipid-Protein Interaction,
and Mixing and Titration Calorimetry
In the formal description of the mixing behavior of multicomponent lipid systems, the methods of classical phase
thermodynamics can be used in largely unchanged form.[971
The phase diagrams can be obtained from DSC curves
recorded for different mixing ratios.['' In connection with
lipid-protein interactions, the question concerning the state
of order of lipid molecules in the immediate surrounding
(annulus zone) of integrated membrane proteins is of particular interest.[981Figure 9 shows, as an example, a set of DSC
989
takes place (phase transition L, --t L p . )between 22 "C and
51 "C, the L, --* L, phase transition overlaps the deprotonation;['' above 51 "C, the experimentally determined AHdirs
values are nearly equal to the enthalpies of denaturation
because, in this temperature range, the phase state of lipids
does not change. These results are discussed in detail elsewhere.[']
4L
9. Summary
e
5
0
10
15
20
J
l°C1
25
An opportunity to compare the magnitudes of the AH
values of different important processes is provided by
Table 2. The range of application of calorimetric methods
30 35
-
Fig. 9. DSC curve of DMPC-bacteriorhodopsin mixed-vesicle suspensions
with the mixture proportions a) 00, b) 552, c) 316, d) 188, e) 91 (taken from
[991).
curves of aqueous suspensions of vesicles["*
formed
from dimyristoyl phosphatidylcholine (DMPC) and bacteriorhodopsin for different values of the mixing ratio. From
a comparison of simulated transition curves, based on a simple model, with experimentally determined transition curves,
it can be concluded that the state of order of approximately
60 DMPC molecules is influenced by an intercalated bacteriorhodopsin molecule accompanied by a reduction of the
individual AH, values of these molecules (approximately
30 YO).Accordingly, an effective "order effect" beyond the
immediate bordering zone can be attributed to the intercalated protein. Similar conclusions are drawn from measurements made in reaction and mixing calorimeters for the characterization of the phospholipid-melittin system.[']
The investigation of reaction (12 ) exemplifies measureDMPAe
+ OHe F? DMPA'" + H,O
(12)
ments made with a mixing c a l ~ r i m e t e r . ~The
~ ~ *reaction
~'
enthalpy AHdiesfor the DMPAe deprotonation reaction is
obtained by subtracting the enthalpy of neutralization from
the measured total reaction enthalpy when the dilution enthalpy of NaOH is eliminated by compensation in a twin
calorimeter arrangement. From the graphical presentation
of the temperature dependence of the AHdiss
values (Fig. lo),
Table 2. Magnitude of characteristic A H values for different types of structural
transformations in biopolymer solutions and aqueous lipid suspensions.
Process
Molecular
subunit
Specific transfer
ribonucleic acids
Denaturation of the
secondary
structure
Double helix-random
coil transformation
Base unstacking
Reversal
of the "intercalation"
Denaturation of the
secondary
structure
a-Helix
formation
tRNA molecule 1180
[44,46]
Base pair
35
[38, 391
Nucleotide
residue
EB molecule
15
[lo]
20
[31]
CPFV molecule 3900
[50]
Amino acid
residue
4
[221
0.5
[lo]
Deoxyribonucleic
acid
Polyriboadenylic
acid
DNA and ethidium
bromide (EB)
Viroid
Poly(y-benzylglutamate) in
dichloroacidic acid/
1,2-dichloroethane
(Pro,Pro,Gly),,
Poly(N5-hydroxypropyl-L-glutamine)
Protein
401.
Phospholipid vesicle
in aqueous suspension
Phosphatidic acid
1'
0
I
c
10
20
I
,
30
JI'Cl
-
40
50
,
60
Fig. 10. Temperature dependence of the dissociation enthalpy for the second
dissociation step of dimyristoyl phosphatidic acid [91.
three characteristic temperature areas can be recognized. For
temperatures below 22 "C, in addition to deprotonation, a
change in the angle of inclination of the fatty acid chains
990
AH
Ref.
[kJ mol-']
Model
substance
Polyproline I1
helix-random coil
transition
a-Helixrandom
coil transition
Denaturation
Amino acid
residue
Chain segment
a-Subunit of
tryptophan synthase mutant
(Ser 49) ( M , =
29000 g mol-')
DPPC molecule
Lipid
phase transition: gelliquid-crystalline
Deproton- DMPA molecule
ation (sec- or ion
ond step)
100
42
110, 65, 701
[83]
for thermodynamic characterization of biopolymers and
suspended lipids is also shown in this table.
I wish to thank the Deutsche Forschungsgemeinschaft and the
Fonds der Chemischen Industrie for their support of our scienAngew. Chem. I n f . Ed. Engi.
28 (1989) 981-991
tijk work. In particular, I would like to thank m y former and
present co-workers mentioned in the references for their tireless ejjorts and numerous helpfill discussions.
Received: June 16, 1987:
revised: November 15, 1988 [A 727 IE]
German version: Angew. Chem. /Of (1989) 1005
Translated by DipLChem. 5. Jackson, Brussels, Belgium
[l] J. Enge1,G. Schwarz, Angew. Chem. 82(1970)468; Angew. Chem. l n t . Ed.
Engl. 9 (1970) 389.
[2] B. H. Zimm, J. K. Bragg, J. Chem. Phys. 31 (1959) 526.
[3] J. Applequist, J. Chem. Phys. 38 (1963) 934.
[4] B. H. Zimm, P. Doty, K. Iso, Proc. Nafl. Acad. Sci. USA 45 (1959) 1601.
[5] J. Marmur, P. Doty, J. Mol. Biol. 5 (1962) 109.
[6] D. Chapman, R. M. Williams, B. D. Ladbrooke, Chem. Phys. Lipids 1
(1967) 445.
[7] J. A. Lucy in D. Chapman (Ed.): Biological Membranes, Academic Press,
London 1968, p. 261 ff.
[S] D. Bach in D. Chapman (Ed.): Biomembrane Sfrucfureand Function,
Verlag Chemie, Weinheim 1984, p. 1 ff.
[9] A. Blume in C. Hidalgo (Ed.): Physical Properties o/ Membranes, Functional Implications, Plenum Press, New York 1988, p. 71.
[lo] H.-J. Hinz (Ed.): Thermodynamic Data for Biochemisfry and Biofechnology, Springer, Berlin 1986.
[ I l l G. C. Pimentel, A. L. McClellan: The Hydrogen Bond, Freeman, New
York 1960.
[12] S . Lifson, A. Roig, J. Chem. Phys. 34 (1961) 1963.
[13j N. Davidson: Stafistical Mechanics. McGraw-Hill, New York 1962,
p. 385ff.
[14] R. Jaenicke, Angew. Chem. 96 (1984) 385; Angew. Chem. lnt. Ed. Engl. 23
(1984) 395.
[15] J. Marmur, D. Lane, Proc. Natl. Acad. Sci. USA 46 (1960) 453.
[16] E. Neumann, Angew. Chem. 85 (1973) 430; Angew. Chem. l n t . Ed. Engl.
12 (1973) 356.
[17] D. Barszcz, D. Shugar, Eur. J. Biochem. 5 (1968) 91.
[18] H.Trauble, H. Eibl in F. 0. Schmitt, D. M. Schneider, D. M. Crothers
(Eds.): Funcfional Linkage in Biomolecular Systems, Raven Press, New
York 1975, p. 59ff.
[19] G. Messner, M. Grubert, J. Phys. E 16 (1983) 131.
[20] T. Ackermann, H. Riiterjans, Z. Phys. Chem. (Frankfurt/Main)41 (1964)
116.
[21] E. Neumann, T. Ackermann, J. Phys. Chem. 73 (1969) 2170.
[22] T. Ackermann, H. Riiterjans, Ber. Bunsen-Ges. Phys. Chem. 68 (1964)
850.
1231 T. Ackermann, Ber. Bunsen-Ges. Phys. Chem. 62 (1958) 411.
1241 M. Grubert, T. Ackermann, Z . Phys. Chem. (FrankfurflMain) 93 (1974)
255.
1251 T. Ackermann in D. B. Davies, W. Saenger, S. S. Danyluk (Eds.): Sfructurd Molecular Biology, Plenum, New York 1982, p. 305ff.
[26] W. Hemminger, G. Hohne: Calorimetry, Verlag Chemie, Weinheim 1984.
[27] P. L. Privalov, V. V. Plotnikov, V. V. Filimonov, J. Chem. Thermodyn. 7
11975) 41.
[28] K. S. Krishnan, J. F. Brandts, Methods Enzymol. 49 (1978) 1.
[29] M. Grubert, Z. Phys. Chem. (FrankfurflMain) 93 (1974) 255.
[30] C. L. Stevens, G. Felsenfeld, Biopolymers 2 (1964) 293.
[31] W. Burkart, H. Klump, Ber. Bunsen-Ges. Phys. Chem. 85 (1981) 371.
[32] I . Wadso, Acfa Chem. Scand. 22 (1968) 927.
1331 W. Burkart, H. Klump, Ber. Bunsen-Ges. Phys. Chem. 85 (1981) 375.
[34] H.Klump, T. Ackermann, Biopolymers 10 (1971) 513.
[35] P. L. Privalov, D. R. Monaselidze, Zh. Eksp. Teor. Fiz. 47 (1964) 2073.
[36] L. G. Bunville, P. E. Geiduschek, M. A. Rawitscher, J. M. Sturtevant,
Biopolymers 3 (1965) 213.
[37] J. M. Sturtevant, S. Rice, P. E. Geiduschek, Discuss. Faraday Soc. 25
(1958) 138.
[38] H. Klump, Thermochim. Acfa 85 (1985) 457.
[39] H. Klump, K. Herzog, Ber. Bunsen-Ges. Phys. Chem. 88 (1984) 20.
[40] G. N. Lewis, M. Randall: Thermodynamics, 2nd ed. (revised by K. S.
Pitzer, L. Brewer), McGraw-Hill, New York 1961, p. 246ff.
[41] T. Ackermann in H. D. Brown (Ed.): Biochemical Microcalorimetry. Academic Press, New York 1969, p. 128 ff.
[42] H. Klump, Ber. Bunsen-Ces. Phys. Chem. 91 (1987) 206.
1431 U. Schernau, T. Ackermann, Biopolymers 16 (1977) 1735.
1441 D. Bode, U. Schernau, T. Ackermann, Biophys. Chem. 1 (1974) 214.
[45] H.-J. Hinz, V. V. Filimonov, P. L. Privalov, Eur. J. Biochem. 72(1977) 79.
[46] F. J. Schott, M. Grubert, W. Wangler, T. Ackermann, Biophysical Chemrstry 14 (1981) 25.
[47] D. Riesner, H. J. Gross, Annu. Rev. Biochem. 54 (1985) 531.
Angew Chent. l n f . Ed. Engl. 28 (1989) 981-991
D. Riesner, G. Steger, J. Schumacher, H. J. Gross, J. W. Randles, H. L.
Sanger, Biophys. Struct. Mech. 9 (1983) 143.
J. Langowski, H. Henco, D. Riesner, H. L. Sanger, Nucleic Acids Res. 5
(1978) 1589.
H. Klump, D. Riesner, H. L. Sanger, Nucleic Acids Res. 5 (1978) 1581.
E. Neumann, T. Ackermann, J. Phys. Chem. 71 (1967) 2377.
H. Krakauer, J. M. Sturtevant, Biopolymers 6 (1968) 491.
H.-J. Him, W. Haar, T. Ackermann, Biopolymers 9 (1970) 923.
D. W. Gruenwedel, Biochim. Biophys. Acta 395 (1975) 246.
F. H. C. Crick, J. Mol. Biol. 19 (1966) 548.
J. Stulz, T. Ackermann, Ber. Bunsen-Ges. Phys. Chem. 87 (1983) 447.
D. Sol], J. Mol. Biol. 29 (1967) 113.
K. J. Breslauer in H.-J. Hinz(Ed.): Thermodynamic Datafor Biochemisfry
Springer, Berlin 1986, p. 402.
and Biofechnolo~y,
S. M. Freier, M . Petersheim, D. R. Hickey, D. H. Turner. J. Biomol.
Sfrucf.Dynam. 1 (1983) 1229.
D. Albergo, L. A. Marky, K. J. Breslauer, D. H. Turner, Biochemisrry 20
(1981) 1409.
K. J. Breslauer, J. M. Sturtevant, I. Tinoco Jr., J. Mol. Biol.99 (1975) 549.
A. R. Peacocke in R. M. Acheson (Ed.): Acridines, 2nd ed., Wiley, New
York 1973, p. 723ff.
M. J. Waring in E. F. Gale, E. Cundliffe, P. E. Reynolds, M. H. Richmond, M. J. Waring (Eds.): The Molecular Basis ofAntibiofic Action. 2nd
ed., Wiley, New York 1981, p. 273ff.
H. W. Zimmermann, Angew. Chem. 98(1986) 115; Angew. Chem. l n f . Ed.
EngI. 25 (1986) 115.
W. Pfeil in H.-J. Hinz (Ed.): Thermodynamic Data for Biochernisrrj~and
Biotechnology, Springer, Berlin 1986, p. 349.
F. E. Karasz, J. M. OReilly, H. E. Bair, Nature (London) 202 (1964) 693.
T. Ackermann, E. Neumann, Biopolymers 5 (1967) 649.
V. Ganser, J. Engel, D. Winklmair, G. Krause, Biopotymers 9(1970) 329.
R. Jaenicke, Annu. Rev. Biophys. Bioeng. 10 (1981) 1.
W. Pfeil, P. L. Privalov in M. N. Jones (Ed.): Biochemical Thermodynamics (Studies in Modern Thermodynamics, Vol. l), Elsevier, Amsterdam 1979, p. 75ff.
P. D. Ross in H.-J. Hinz (Ed.): ThermodynamicData for Biochemistry and
Biotechnology. Springer, Berlin 1986, p. 227.
J. M. Sturtevant, G. Veliceleby, R. Jaenicke, M. A. Lauffer, Biochemistry
20 (1981) 3792.
H. Wiesinger, H.-J. Hinz in H:J. Hinz (Ed.): Thermodynamic Data for
Biochemistry and Biotechnology, Springer, Berlin 1986, p. 21 1.
M. I. Paz Andrade, M. N. Jones, H. A. Skinner, Eur. J. Biochem. 66
(1976) 127.
J. F. Halsey, R. L. Biltonen, Biochemistry 14 (1975) 6221.
M. F. M. Johnston, B. G. Barisas, J. M. Sturtevant, Biochemrsfry 13
(1974) 390.
T. Kodama, I. D. Watson, R. C. Woledge, J. Biol.Chem. 252 (1977) 8085.
T. Kodama, R. C. Woledge, J. Bioi. Chem. 254 (1979) 6382.
C. W. Niekamp, H.-J. Hinz, R. Jaenicke, C. Woenckhaus, R. Jeck, Biochemistry 19 (1980) 3144.
R. P. Hearn, F. M. Richards, J. M. Sturtevant, G. D. Watt, Biochemrsfry
10 (1971) 806.
H.-J. Hinz, S. Srinivasan, E. M. J. Jaspars, Eur. J. Biochem. 95 (1979) 107.
E. Jiirgens, G. Hohne, E. Sackmann, Ber. Bunsen-Ges. Phys. Chem. 87
(1983) 95.
A. Blume, T. Ackermann, FEBS Leu. 43 (1974) 71.
A. Blume, Biochemistry 22 (1983) 5436.
R. J. Wittebort, C. F. Schmidt, R. G. Griffin, Biochemistry 20 (1981)
4223.
H.-J. Hinz, L. Six, K.-P. Ruess, M. Lieflander, Biochemisrry 24 (1985)
806.
J. Stumpel, H. Eibl, A. Niksch, Biochem. Biophys. Acta 727 (1983) 246.
G. Wedler: Lehrbuch der Physikalischen Chemie, Verlag Chemie, Weinheim 1982, p. 237.
G. Kortum, H. Lachmann: Einyuhrung in die chemische Thermodynamik,
Verlag Chemie, Weinheim 1981, p. 124.
See [40], p. 375ff.
P. Privalov, N. N. Khechinashvili, 1 MoI. Bioi. 86 (1974) 556.
E. Wicke, Angew. Chem. 78(1966) 1 ;Angew. Chrm. lnt. Ed. Engl. 5 (1966)
106.
G. C. Kresheck in F. Franks (Ed.): Water, A Comprehensive Treatise,
Vol. 4, Plenum Press, New York 1975, p. 95ff.
F. M. Menger, Acc. Chem. Res. 12 (1979) 111.
N. Nichols, R. Skold, C . Spink, J. Suurkuusk, I. Wadso, J. Chem. Thermodyn. 8 (1976) 1081.
G. Roux, G. Perron, J. E. Desnoyers, Can. J. Chem. 56 (1978) 2808.
A. G. Lee, Biochem. Biophys. Acta 472 (1977) 285.
N. J. Singer, G. L. Nicholson, Science (Washington,DC) /75 (1972) 720.
M. P. Heyn, A. Blume, M. Rehorek, N. A. Dencher, Biochemistry 20
(1981) 7109.
M. P. Heyn, R. J. Cherry, N. A. Dencher, Biochemistry 20 (1981) 840.
99 1
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 407 Кб
Теги
calorimetric, aggregates, biopolymers, studies, phospholipid
1/--страниц
Пожаловаться на содержимое документа