heat induced denaturation of fibrous hard
TRANSCRIPT
Heat induced denaturation of fibrous hard
α-keratins and their reaction with various
chemical reagents
Von der Fakultät für Mathematik, Informatik und
Naturwissenschaften der RWTH Aachen University zur
Erlangung des akademischen Grades eines Doktors der
Naturwissenschaften genehmigte Dissertation
vorgelegt von
Diplom-Ingenieur
Daniel-Vasilica Istrate
aus Iasi, Rumänien
Berichter: Universitätsprofessor Dr. rer. nat. Martin Möller
Universitätsprofessor Dr. rer. nat. Walter Richtering
Tag der mündlichen Prüfung: 20. Juni 2011
Diese Dissertation ist auf den Internetseiten der Hochschulbibliothek online verfügbar.
Acknowledgements
First of all I would like to thank Prof. Dr. Martin Möller, Director of the DWI who
honoured me by accepting my PhD study under his coordination. I am greatly indebted to him
for giving me the opportunity to do the PhD thesis on a topic of such a great actual interest and
also for his support and innovative ideas that contributed essentially to the success of this
project.
I would like to express my deep gratitude to Prof. Dr. Crisan Popescu for the enlightening
and stimulating dialogue throughout the all years of my PhD study, for the permanent
encouragement and warm support. This thesis would hardly have become possible without his
critical and creative discussions.
I am furthermore indebted to Professor Franz-Josef Wortmann from School of Materials
Manchester for his contribution to the initiation and development of this project, as DWI
member.
I gratefully acknowledge the financial support and the prolific exchange of ideas during the
development of this project from Beiersdorf AG, CIBA Spezialitaetenchemie Grenzach GmbH,
COGNIS GmbH, HENKEL AG & Co. KGaA, KPSS-KAO Professional Salon Services GmbH
and WELLA Services GmbH through the German cosmetic industry committee ―DSC of Human
Hair‖.
In great recognition of his service in performing amino acid analyses I am deeply grateful
to Dr. Josef Föhles.
My thanks go also to Dr. Er Rafik Meriem for the scientific discussions and help, for
sharing her knowledge with me regarding the X-ray analyses involved in this work. Moreover, I
would like to thank Dr. Hô Phan and Franz Steffens for preparing the electron microscopic
pictures.
I am filled with a deep sense of gratitude to my beloved wife, Monica. I would like to
thank her, beside for the outstanding support and valuable advices regarding the tensile
properties experiments, for her tremendous confidence and for her special way of making come
to light the best of myself.
To come so far was also the merit of the wonderful professors I had the chance to work
Aknowledgements
with during my study at the Faculty of Textiles and Leather Engineering from the University
―Gh. Asachi‖, Iasi, Romania. My greatest appreciation and thanks go to Conf. dr. ing. Radu
Cezar Doru. Very grateful I am to Prof. dr. ing. Mureşan Augustin. I owe also enormously to all
members of the Textile Chemical Technology Department.
Last but not least, I am thankful to all former and present colleagues from DWI and
especially from the joint groups of Prof. Dr. Crisan Popescu, for the nice atmosphere and for the
time we spent together.
Though, many have not been mentioned, none is forgotten.
To Monica
List of Publications
Parts of this thesis are published, in preparation to be published or presented at
conferences:
Articles D. Istrate, C. Popescu, M. Möller, Nonisothermal kinetics of hard α-keratin
thermal denaturation, Macromolecular Bioscience, Volume 9, Issue 8, p 805-812,
2009
D. Istrate, C. Popescu, M. Er Rafik, M. Möller, Thermal denaturation of fibrous
hard α-keratins and the effect of pH, Polymer Degradation and Stability,
submitted 2010
D. Istrate, C. Popescu, F.-J. Wortmann, M. Möller, Micro-tubes of keratin. The
thermal stability of cortex and cuticle, Biomacromolecules, submitted 2010
D. Istrate, C. Popescu, M. Er Rafik, M. Möller, Differential scanning calorimetry
(DSC) analysis of structural changes in bleached, perm-waved and dyed hard
alpha-keratin fibres, Journal of the Society of Cosmetic Chemists, submitted 2009
M. Baias, D.E. Demco, D. Istrate, C. Popescu, B. Blümich, M. Möller,
Morphology and molecular mobility of fibrous hard α-keratins by 1H, 13C, and
129Xe NMR, The Journal of Physical Chemistry B, 113 (35), p 12136–12147,
2009
Posters D. Istrate, C. Popescu, Investigations by Differential Scanning Calorimetry of the
effects of hair bleaching on the main morphological components of human hair,
DWI Reports 130, P14 (2006)
M. Er Rafik, D. Istrate, C. Popescu, Behaviour of human hair at various
temperatures, TRI/Princeton Hair Conference 2006
M. Baias, D. Istrate, C. Popescu, D. E. Demco, M. Möller, B. Blümich, Thermal
denaturation of keratin by 1H solid-state NMR, Aachen Dresden International
Textile Conference, Aachen, Germany, November 2007
List of Abbreviations
H enthalpy
DSC Differential Scanning Calorimetry
DTA differential thermal analysis
HPDTA high-pressure differential thermal analysis
IFAPs intermediate filaments associated proteins
IFs intermediate filaments
LES Lauryl ether sulphate
NaOH natrium hydroxide
SAXS small- angle X-Ray scattering
S-S disulphide
Td denaturation temperature
Tmax maximum temperature
Tp transition temperature peak
WAXS wide-angle X-Ray scattering
Table of Contents
Acknowledgements .......................................................................................................................................5
List of Publications ........................................................................................................................................9
Summary .....................................................................................................................................................17
Zusammenfassung .......................................................................................................................................19
Chapter I : Introduction .........................................................................................................................21
1.1. Thermal stability of proteins .......................................................................................................21
1.2. Thermal stability of fibrous hard α-keratins ................................................................................25
1.2.1. Physical and mechanical models of keratin fibres ...................................................................26
1.2.2. Thermal analysis of keratin fibres ...........................................................................................28
1.3. Content of this thesis ...................................................................................................................31
1.4. References and notes ...................................................................................................................33
Chapter II : Micro-tubes of keratin. The thermal stability of cortex and cuticle* ...................................37
2.1. Introduction .................................................................................................................................37
2.2. Materials and methods .................................................................................................................37
2.3. Results and discussions ...............................................................................................................39
2.4. Conclusions .................................................................................................................................47
2.5. References and notes ...................................................................................................................47
Chapter III : Thermal denaturation of fibrous hard α-keratins and the effect of pH* ...............................49
3.1. Introduction .................................................................................................................................49
3.2. Materials and methods .................................................................................................................51
3.3. Results and discussions ...............................................................................................................54
3.4. Conclusions .................................................................................................................................67
3.5. References and notes ...................................................................................................................67
Chapter IV : Nonisothermal kinetics of hard α-keratin thermal denaturation* .....................................71
4.1. Introduction .................................................................................................................................71
Table of contents
4.2. Materials and methods ................................................................................................................ 72
4.3. Kinetic modeling: General description of the kinetic method .................................................... 73
4.3.1. The activation energy, Ea ....................................................................................................... 75
4.3.2. The kinetic function, f(α) and the pre-exponential factor, A .................................................. 75
4.4. Results and discussions .............................................................................................................. 76
4.5. Conclusions ................................................................................................................................ 84
4.6. References and notes .................................................................................................................. 84
Chapter V : Differential scanning calorimetry (DSC) analysis of structural changes in bleached, perm-
waved and dyed hard alpha-keratin fibres* ................................................................................................. 87
5.1. Introduction ................................................................................................................................ 87
5.2. Materials and methods ................................................................................................................ 89
5.3. Results and discussions .............................................................................................................. 93
5.4. Conclusions .............................................................................................................................. 105
5.5. References and notes ................................................................................................................ 105
Appendix A: Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
*
.................................................................................................................................................................. 109
A.1. Introduction .............................................................................................................................. 109
A.2. Materials and methods .............................................................................................................. 111
A.3. Theory of NMR spin diffusion ................................................................................................. 114
A.4. Results and discussions ............................................................................................................ 118
A.5. Conclusions .............................................................................................................................. 134
A.6. References and notes ................................................................................................................ 134
Appendix B: Nonisothermal kinetics of chemically damaged hard α-keratin thermal denaturation ........ 137
B.1. Introduction .............................................................................................................................. 137
B.2. Material and methods ............................................................................................................... 138
B.3. Results and discussions ............................................................................................................ 139
B.4. Conclusions .............................................................................................................................. 144
B.5. References and notes ................................................................................................................ 145
Table of contents
Appendix C: Factors influencing the DSC thermogram of hard alpha-keratin proteins and the
reproducibility of the experimental results ................................................................................................147
C.1. Introduction ...............................................................................................................................147
C.2. Factors relating to the DSC methodology .................................................................................147
C.2.1. Instrument baseline ............................................................................................................147
C.2.2. Analysis of the experimental thermograms .......................................................................147
C.2.3. Sample pans and crucibles.................................................................................................148
C.2.4. Pressure influence ..............................................................................................................148
C.2.5. pH influence ......................................................................................................................148
C.3. Factors relating to fibrous protein structure ..............................................................................148
C.3.1. Cortex-cuticle assembly ....................................................................................................148
C.3.2. Melanin pigment ................................................................................................................150
C.3.3. Ethnic differences ..............................................................................................................151
C.4. Reproducibility of the experimental results ..............................................................................152
C.5. References and notes .................................................................................................................153
General Conclusions..................................................................................................................................155
Summary
This dissertation is concerned with the thermal behaviour of fibrous proteins encapsulated
in rigid structures, among the most well-known representatives of this class being the α-keratins
in human hair. In spite of a lot of work in the field, there is still no mechanism proposed for
accounting on how thermal denaturation process occurs in hard α-keratins. This work aims at
proposing a model for the α-keratin fibres and a mechanism for their thermal denaturation
process. These are further used for understanding the effect of various cosmetic reagents on the
thermal stability of the fibres.
The use of differential scanning calorimetry, scanning electron microscopy and light
microscopy revealed strong structural modifications induced by high temperature in case of
heating keratin material in an opened atmosphere. The DSC in open environment was showed to
supply misleading information, due to the interference of pyrolysis with the process of interest.
Consequently the present work focuses mainly on using DSC of keratins in water excess.
The study of the influence of pH, particularly acid values, on thermal behaviour of hard
alpha-keratins, indicates limits of the two-phase model used so far to describe the fibrous
proteins. We propose a three-phase model for explaining fibrous hard alpha-keratins high
thermal stability and their reaction with various reagents. The approach is based on results from
DSC study of keratins under various conditions, and is supported by amino-acid analysis, X-ray
diffraction, Raman spectroscopy and tensile strength observations. According to the proposed
model, the third phase, the interface between crystalline and matrix phases, made of nonhelical
tail domains of keratin, scaffolds the intermediate filaments and controls their interaction with
chemical reagents as well as their thermal properties.
The differential scanning calorimetry measurements carried out in water excess and with
different heating rates were used for the kinetic analysis of the endothermic process assigned to
the denaturation of the helical material from human hair. We found that the kinetic mechanism is
autocatalytic and that the value of the activation energy is rather close to disulphide bond
scission than to protein denaturation. This allowed us proposing a multistep mechanism for the
thermal denaturation of hard α-keratins in water excess that relies on the 3-phase model which
describes their structure. The limiting step of the thermal denaturation process is then the
Summary
scission of S-S bonds between the main morphological components, namely intermediate
filaments (IF) and matrix (IFAP). The theoretical proposed model shows a good agreement with
the experimental recorded data.
The chemical damage induced by bleaching, permanent waving and oxidative dyeing on
the structure of hard alpha-keratin fibres (human hair) as revealed by modifications in their
thermal behaviour was investigated by using differential scanning calorimetry. Regression
analysis of the data from hair samples treated differently shows a linear correlation between the
enthalpy of the denaturation peak recorded by DSC and the cystine content of the fibre. The
experimental results are evaluated within the framework of a three-phase model in which the
nonhelical (globular) terminal domains of keratin promote filament interactions and control the
thermal properties of keratin intermediate filaments. Amino-acid analysis, X-ray diffraction and
tensile strength measurements provide evidence that the attack of chemical reagents occur
preponderantly in the matrix and at the interface between filament and matrix. A possible
intermediate state between native and denaturated crystalline helical material is suggested to
account for the increased disorder in the IFs-IFAP package induced by harsh treatments. The
DSC data suggests that hair keratin IFs can modulate their organisation and thermal properties
through chemical induced interactions.
Zusammenfassung
Diese Dissertation beschäftigt sich mit dem thermischen Verhalten von fiberartigen
Proteinen, die in starren Strukturen eingelagert sind. Zu den bekanntesten Vertretern dieser
Klasse gehören die α-Keratine aus menschlichem Haar. Trotz einer großen Anzahl an
wissenschaftlichen Arbeiten in diesem Feld wurde bis jetzt noch kein Mechanismus für die
thermische Denaturierung von harten α-Keratinen vorgeschlagen. Ziel dieser Doktorarbeit ist das
Aufstellen eines Modells für α-Keratin Fasern und die Klärung des Mechanismus für die
thermische Denaturierung. Die Ergebnisse tragen zum Verständnis der Wirkungsweisen von
verschiedenen kosmetischen Reagenzien auf die thermische Stabilität der Fasern bei.
Die Verwendung von Dynamische Differenzkalorimetrie (DSC), Rasterelektronen- und
Lichtmikroskopie offenbarte starke strukturelle Modifikationen, die durch die hohe Temperatur
im Falle von Erwärmen des Keratins in offener Atmosphäre induziert wurden. DSC in offener
Umgebung erwies sich als irreführende Methode infolge der Überlagerung von Pyrolyse und
dem hier zu untersuchenden Prozess. Konsequenterweise legt diese Arbeit ihren Fokus in erster
Linie auf DSC-Untersuchungen an Keratinen in einem Überschuss an Wasser.
Untersuchungen des Einflusses von pH, insbesondere im Sauren, auf das thermische
Verhalten von harten α-Keratinen zeigen Grenzen des Zweiphasenmodels auf, das bis jetzt für
die Beschreibung der fiberartigen Proteine diente. Wir schlagen daher ein Dreiphasenmodel vor,
das sowohl die gute thermische Stabilität der harten α-Keratinen als auch ihre Reaktivität
gegenüber verschiedener Reagenzien erklärt. Dieser Ansatz basiert auf die Ergebnisse der DSC-
Untersuchungen an Keratinen unter verschiedenen Bedingungen und wird durch
Aminosäurenanalyse, Röntgenstreung, Ramanspektroskopie und Zugfestigkeitsuntersuchungen
unterstützt. Gemäß des vorgeschlagenen Models hält die dritte Phase, die Grenzfläche zwischen
dem kristallinen Bereich und der Matrixphase, die aus nichthelikaler Keratindomänen besteht,
die dazwischenliegenden Filamente zusammen und kontrolliert sowohl die Wechselwirkung mit
chemischen Reagenzien als auch ihre thermischen Eigenschaften.
Die DSC-Messungen, die in Überschuss von Wasser und mit verschiedenen Heizraten
durchgeführt wurden, dienten für die kinetische Analyse des endothermen Prozesses, der der
Denaturierung der helikalen Bestandteile des menschlichen Haars zugeschrieben wird. Dabei
Zusammenfassung
fanden wir heraus, dass der kinetische Mechanismus autokatalytisch ist und die
Aktivierungsenergie ist sehr nahe dem Wert für die Disulfidespaltung und nicht dem Wert für
Proteindenaturierung. Diese Feststellung erlaubte einen Multischrittmechanismus für die
thermische Denaturierung von harten α-Keratine bei Wasserüberschuss vorzuschlagen, der auf
das Dreiphasenmodel, das ihre Struktur beschreibt, beruht. Dabei ist der
geschwindigkeitsbestimmende Schritt der thermischen Denaturierung die Spaltung der S-S
Bindungen zwischen den morphologischen Hauptkomponenten, nämlich den
dazwischenliegenden Filamenten (IF) und der Matrix (IFAP). Das vorgeschlagene theoretische
Model zeigt gute Übereinstimmung mit den experimentellen Daten.
Die Änderung im thermischen Verhalten aufgrund chemischer Schäden an den harten α-
Keratinfasern (menschliches Haar), hervorgerufen durch Bleichen, Einbringen von Dauerwellen
und durch oxidative Färbung, wurde mittels DSC untersucht. Die Regressionsanalyse der Daten
von Haarproben, die unterschiedlich behandelt wurden, zeigt eine lineare Korrelation zwischen
der Enthalpie des DSC-Denaturierungspeaks und dem Cystingehalt der Faser. Die
experimentellen Ergebnisse wurden im Rahmen des Dreiphasenmodels ausgewertet, in dem die
nichthelikalen (globulären) Enddömänen von Keratin Filamentwechselwirkungen begünstigen
und die thermischen Eigenschaften der IF Filamenten kontrollieren. Aminosäureanalyse,
Röntgenstreuung und Zugfestigkeitsuntersuchungen belegen den Angriff der chemischen
Reagenzien überwiegend in der Matrix und an der Grenzfläche zwischen Filamenten und Matrix.
Ein möglicher Zustand zwischen dem nativen und denaturierten kristallinen, helikalen Material
wurde vorgeschlagen, um die zunehmende Unordnung in den IF-IFAP Packungen, die durch
eine aggressive Behandlung hervorgerufen wurde, zu berücksichtigen. Die DSC Daten weisen
darauf hin, dass die IF des Keratins aus dem Haar ihre Organisation und thermische
Eigenschaften durch chemisch-induzierte Wechselwirkungen modulieren können.
Chapter I : Introduction
1.1. Thermal stability of proteins
Proteins are macromolecules (polypeptides) arranged in complex structures that are
important to their function. These structures have numerous levels namely primary, secondary,
tertiary, and quaternary1 ones. The primary structure is associated with the covalent bonds
between the atoms making up the protein molecule; the secondary structures involve primarily
hydrogen bonding between the atoms (although some disulfide bonding can also occur), thereby
creating the (well-known) alpha helix and beta sheet structures, whereas the ultimate 3D folded
structure of the whole (globular) protein is called the tertiary structure and is important to protein
function. Quaternary structure usually involves the conformational fitting of two proteins
together associated with specific function2. When this structure is changed or altered, the protein
is unable to carry out its specific function. This may involve either partial or total unravelling of
the protein through changes of the hydrogen bonding which define the higher-order native
structure of the protein. This process is called denaturation. It can be either partial or total,
meaning that the process may not necessarily complete for a specific condition, and it can also be
reversible or irreversible. Denaturation does not involve breaking of the individual covalent
bonds between the atoms of the polypeptide backbone of the protein molecule2. The denaturation
process can be accompanied by aggregation, coagulation, and gelation3. Aggregation is a general
term referring to protein–protein interactions with formation of complexes of higher molecular
weights. Coagulation is the random aggregation of already denatured protein molecules and is
usually a thermally irreversible process. Gelation is an orderly aggregation of proteins, which
may or may not be denatured, forming a three-dimensional network that may be thermally
reversible. The thermally irreversible loss of protein stability and function is rate limited initially
by the denaturation step, which may then be followed by coagulation, aggregation, and/or
gelation2.
Differential Scanning Calorimetry (DSC) has been widely used as a tool for biomolecular
studies4-7
. For globular/soluble proteins, the thermally induced process detectable by DSC is the
structural melting or unfolding of the molecule. The transition of protein from a native to a
denatured conformation is accompanied by the rupture of inter- and intra-molecular bonds, and
Chapter 1
22
the process has to occur in a cooperative manner to be discerned by DSC8. Analysis of a DSC
thermogram enables the determination of two important parameters: transition temperature peak
(Tp) or maximum (Tmax) or denaturation (Td) temperature, and enthalpy of denaturation (ΔH).
The denaturation temperature measures the thermal stability of proteins. The value is influenced
by the heating rate9 and protein concentration
10, as well by the biochemical environment
(especially pH)11,12
. The enthalpy value, calculated from the area under the transition peak, is the
heat uptake for the unfolding transition, independent of any denaturation model assumption13
.
Assuming a 2-state transition model for proteins denaturation, the heat uptake is correlated with
the content of ordered secondary structure of a protein10,13
. The ΔH value is actually the balance
of a combination of endothermic reactions, such as the disruption of hydrogen bonds determined
as 7.11 kJ per mole of hydrogen bond14
, and exothermic processes, including protein aggregation
and the break-up of hydrophobic interactions4,15
. The total amount of heat released during
denaturation (followed by coagulation, aggregation, and/or gelation) of purified proteins as well
as whole cell preparation is usually between 20 and 40 J/g protein14,16
and 10 and 60 J/g protein
for rat tail collagen, depending on hydration17
.
Activated state
Natured state
Denatured state
Aggregated /coagulated state
Activation
Denaturation
Aggregation/coagulation
STATE
EN
ER
GY
Figure 1.1 Energy states of protein denaturation2 (not to scale)
Thermodynamically speaking, the denaturation is the condition when sufficient energy is
transferred to a native protein such that an alteration of its molecular conformation can take
Introduction
23
place. At low temperatures the heat capacity of the protein increases monotonically with
temperature like for any organic solid. As the protein begins to unfold at higher temperature the
DSC shows the increase of heat capacity arising from heat energy uptake in the endothermic
unfolding transition. Once this transition is complete the thermogram reverts to a ―post-
transition‖ baseline, reflecting the heat capacity of the now-unfolded protein in solution13
. The
transferred energy usually has two parts: a kinetic component (the activation energy barrier), and
the enthalpic part (total heat absorption or release) –Figure 1.1.
The kinetic (activation energy) barrier determines the temperature and time dependence of
the denaturation process. The total enthalpic heat change is calorimetrically measured when the
phase transition takes place. As the temperature raise, it becomes thermodynamically favourable
for the protein to denature. The final denatured state can be at a higher or lower total energy than
the original state. The final state is at a lower energy than the initial state when coagulation,
aggregation, and/or gelation of denatured proteins occur, which is a strongly exothermic
process2.
Many experimental methods for estimating thermodynamic parameters for protein
transitions are based on the assumption/approximation of ―2-state‖ behaviour for the system. The
accuracy of the data thus obtained, and the validity of their interpretation are critically dependent
on the validity of this assumption. The simplest, but most widely used kinetic model to express
the ―2-state‖ behaviour is the first order irreversible rate reaction model, for protein denaturation
that assumes that the process of interest may be represented by a transition between two
experimentally distinguishable states, native (N) and denatured (D):
N Dk
(1.1)
where k is the rate constants, with no significant population of intermediate states , and the
transition may be brought about by changes in temperature, pH or denaturant concentration. The
D state does not, necessarily, have to become random coil, nor even fully unfolded during the 2-
state transition, and might continue to change – to become ―more unfolded‖ - as more denaturant
is added, or higher temperature reached, for example. Even if the experimental data are
satisfactorily described by this one-step model, the real mechanism of denaturation can be more
complex18
.
Higher-order kinetic models (i.e., models assuming that one or more intermediate states
exist between the native and denatured states) were also adopted in several studies18-20
. These
showed that higher-order models could fit experimental results better than the first-order model.
Chapter 1
24
When analysing the one-step model (eqn. 1.1) it was noted20
that the Lumry and Eyring model21
:
N U D
k1k2
k-1 (1.2)
(where N, U, and D are native, partially unfolded and denaturated protein form ; k1,k-1,k2 are the
rate constant for the corresponding reactions) is more realistic and that the one-step model is a
particular case of the Lumry and Eyring model. Irreversible protein denaturation is thought to
involve at least two steps: (a) reversible unfolding of the native protein (N); (b) irreversible
alteration of the unfolded protein (U) to yield a final state (D) that is unable to fold back to the
native one.
There are two main situations when the Lumry and Eyring mechanism (eqn. 1.2) reduce to
one-step irreversible model (eqn. 1.1). The first situation is when the value of k2 is much higher
than the values of k1 and k -1, so that the direct reaction of the first step is rate-limiting and the
reverse reaction is practically neglected. The second case is realized when the rates of the direct
and reverse reactions of the first step are much higher than the rate of the second step, but
equilibrium for the first step is shifted toward the form N20
.
Transformation of a protein between various conformational states might be brought about
by changes in temperature, pressure, pH, ligand concentration, chemical denaturants or solvent
nature. A transformation may only come about if the folded and unfolded states have different
affinities for these parameters. Temperature-induced protein unfolding (at equilibrium) arises
from differences in enthalpy (ΔH) between folded and unfolded states; pressure denaturation can
only occur if the folded and unfolded states have different partial molar volumes (the unfolded
state is normally of lower volume); unfolding at high or low pH implies differences in pKA of
protein acidic and/ or basic groups; ligand-induced unfolding or stabilization of the native fold
results from differences in binding affinity of the ligand regarding folded or unfolded states;
chemical denaturants may act as ligands, binding differently to folded or unfolded states, or may
act indirectly via changes in overall solvent properties.
There are, as well, modifiers (sensitizers and protectants) for denaturation mechanism2. In
the case of heat denaturation, pH is known to accentuate heat denaturation in individual
proteins11,12
. Other hyperthermic sensitizers in cells include methanol, ethanol, propanol, and
butanol. Other agents (thiol-specific oxidative agents) can also sensitize protein to denaturation.
There are also agents such as glycerol, and D2O, as well as other proteins (called Heat Shock
Proteins (HSPs)) that can retard or delay protein denaturation in cells22
.
Introduction
25
Modification of the denaturation process by mechanical and chemical loading can change
the rate constant, but not the activation energy of the process23
. This situation is suggested to be
due to a change in frequency factor, or entropy of activation, which is associated with the
configurational entropy of the protein molecules23,24
. A wide range of activation energy values
were reported in the literature to identify protein denaturation. Values less than 41.8 kJ/ mol are
typically associated with simple diffusion processes, while values in the 41.8–125.6 kJ/ mol
range can involve enzyme controlled metabolic processes including membrane transport. The
activation energies for protein denaturation can range from as low as 104.7 kJ/ mol to as high as
837.4 kJ/ mol, depending on the temperature and pH conditions25
. However, there are also
arguments that protein denaturation only occurs for activation energies above 418.7 kJ/ mol22
.
This argument has been important in identifying protein denaturation as critical to thermal injury
processes.
1.2. Thermal stability of fibrous hard α-keratins
Little systematic work has been done on the thermodynamics of fibrous proteins, with the
exception of the myosin / tropomyosin family of α-helical coiled-coil proteins26
. A possible
reason is their poor solubility and the difficulty to purify them in sufficient quantities for
biophysical studies. The fibrous proteins are generally high molecular weight, made up of
several long polypeptide chains that make them prone to aggregation and entanglement when
unfolded. Fibrous proteins are distinguished from globular proteins by their filamentous,
elongated form. Most of them play structural roles in animal cells and tissues. Among the most
well-known representatives of this class are the α-keratins in human hair, wool and finger nails,
fibroin in silk, actin and myosin in muscles, and collagen, the most abundant protein in
vertebrate bodies. The unfolding transitions are often irreversible on the experimental timescale,
and non-cooperative or non-2-state processes that makes thermodynamic analysis difficult13
.
Amino acid side chains in such proteins may frequently remain exposed to solvent, on the
outside of the elongated chain structure, even in the folded state, influencing the denaturation
pathway. Observations of the thermal stabilities of fibrous proteins reflect behaviours
characteristic of their structures at both the physical and chemical level. Much of our knowledge
about human hair behaviour at heating derives from wide-ranging research on sheep‘s wool
within the framework of the textile industry.
Over the time thermal behaviour of keratins fibres was studied by two DSC methods,
called for simplicity sake ―dry DSC‖ and ―wet DSC‖, respectively. The first method comprises
Chapter 1
26
investigations performed while allowing the moisture content of the keratin sample to evaporate
with increasing temperature. The ―dry‖ investigation puts into evidence endothermal effects
above 200°C, sometimes as a doublet27-30
. By the second method keratins are investigated in
water excess, in sealed pressure resistant capsules that keep the water during heating28,31-33
. This
method puts into evidence endothermal effects at about 150°C.
The analysis of the recorded endotherms observed with any of the methods relies on the
physical / mechanical models describing the keratin fibres behaviour and on the similarities with
the behaviour to heating of globular proteins.
1.2.1. Physical and mechanical models of keratin fibres
Explanations of the mechanical and most other properties of wool and hair fibres have
been dominated by the consideration of the behaviour of the two-phase microfibril/matrix fine
structure. Nearly all the available models rely on the stress–strain curve, on the basis of
knowledge of the microscopic and molecular morphology of α-keratin fibres. The most
important attempts to consistently interpret the shape of the stress/strain curve in relation to fibre
structure have been made by Hearle and Feughelman.
In 1959, Feughelman laid the foundations of structural interpretation of the stress–strain
curve of keratin fibres with his two-phase model of microfibrils imbedded in a matrix34
. The
model (Figure 1.2.a) consist of long, water-impenetrable relatively rigid cylindrical rods set
parallel to the fibre axis and embedded in a water absorbing matrix. In this model, adapted to the
current knowledge of keratin morphology, the α-helices, aggregated in the IFs, form a
crystalline, continuous, axially oriented, elastic filament phase, which is embedded in a matrix
phase that comprises the non helical proteins (IFAPs) and all other noncrystalline viscoelastic
components (intermacrofibrillar cement, nuclear remnants, cell membrane complex, cuticle,
etc.). The absorption of water by the matrix mechanically weakens this phase, whereas the rods
are water impenetrable, thus mechanically unaffected by the presence of water. During the time,
several corrections were implemented to the original two-phase model. In 1960- Haly and
Feughelman35
respectively in 1968-Bendit and Feughelman36
have developed the so-called series
zone model in which the microfibril contains two kinds of alternating zones, named X and Y,
endowed with different elastic properties. The mechanical properties of the matrix are supposed
to be driven by an entanglement of the matrix chains due to disulphide bonds. In 197937
,
Feughelman description of the model identified the X- and Y-zones with different zones of an IF
structure proposed by Fraser38
.
Introduction
27
Figure 1.2 Mechanical models redrawn & adapted from originals: a) Schematic of
microfibril/matrix assembly, after Feughelmann34
. b) Two-phase extended model after
Feughelman40
with staggered terminal domain links between microfibrils, the matrix existing as
separate globules. c) Schematic arrangement of units in IF after Wortmann/Zahn39
d) Crewter´s47
―beaded chain‖ model for the matrix, with protofibrils showing acid-labile and disulphide
linkages and side-chain interactions between microfibrils and matrix molecules e)
Chapman/Hearle model49
where the α-helical rods have head and tail domains extended into the
matrix
Chapter 1
28
Wortmann and Zahn (1994) have reinterpreted available biochemical data on the
microfibril‘s structure to derive a model that neglects the matrix proteins39
. The X and Y zones
of the series zone model are assigned to specific amino acid sequences along the length of the
keratin molecule (Figure 1.2.c). In 1994, Feughelman proposed a new model (Figure 1.2.b),
which treated the matrix as globular proteins with a hydrophilic surface and hydrophobic
interior, surrounded by water40
. In this model, water molecules are supposed to be ejected from
the matrix at high stress levels, which leads to a matrix proteins compression between the
microfibrils. This model is not specific on the IF structure, except to postulate that the terminal
domains link neighbouring IFs and define the spacing between them.
In 1969, an alternative model was proposed by Chapman41
. The matrix proteins are
supposed to be covalently linked to fundamental repeat units aligned along the microfibril. The
stress–strain curve of the fibre is modelled as a combination of the stress–strain curves of the
microfibril and of the matrix in permanent interaction. Some features of the Chapman model
were independently proposed by Hearle42
, further developments43-45
of this model leading to the
Chapman-Hearle model (Figure 1.2.e) which treats the microfibrils as ideal α-helical crystals46
.
The matrix is regarded as a rather highly cross-linked swollen rubber. The third structural feature
of Chapman-Hearle is a periodic linkage between microfibril and matrix, interpreted as occurring
through the cysteine-rich tails (terminal domains) on the keratin IF proteins. The matrix fills the
space between the microfibrils, with the globular IFAPs having inter- and intra-molecular
crosslinks and links to the terminal domains. Another model (Figure 1.2.d), which is close to
Chapman-Hearle model, was proposed by Crewther on the basis of the effect of chemical
treatments on the stress–strain properties of wool47,48
.
Crewther regards the matrix as a ‗chain of beads‘, in which the crosslinked protein globules are
weakly linked to one another by a few disulphide bonds. He also suggests a fairly uniform
distribution of disulphide bonds between the globular protein and the microfibrils. However,
now that it is known that most of the cystine in the high-sulphur protein is in the terminal
domains, an intermittent linkage is more likely46
.
1.2.2. Thermal analysis of keratin fibres
When allowing the moisture to evaporate during heating, one or two endothermic peaks
were usually distinguished on a DSC thermogram, above 200°C. Reducing the complex
morphological structure of fibrous proteins, as stated in the Feughelman two-phase model to the
low sulphur microfibrils (the α-helices forming the IFs -crystalline phase) embedded in a non
Introduction
29
helical high sulphur matrix (comprising also the cuticle- amorphous phase) leaves only two
primary contribution to the origin of the endotherms: the filament and the matrix. Using the first
order assumption of the irreversible rate reaction model for protein denaturation, these
endothermic peaks have been interpreted frequently contradictorily in terms of helix melting
points or irreversible helix unfolding28,50,51
(i.e. microfibrillar origin) and cystine decomposition
points27,52-54
(i.e. matrix origin). Additional DSC investigations of isolated microfibrilar and
matrix proteins in the disulphide form have shown that in case of endothermic doublets, the first
peak (lower temperature) has microfibril origins while the second one is a matrix peak55
.
Investigations on annealed keratins reported that the first peak, of microfibrilar origins, is not
only an irreversible helix unfolding, overlapping with various decomposition reactions56
. More
recently, Cao57,58
conducted DSC studies on Merino wool in an intermediate state between dry
and wet approaches, using silicon oil, with the intention to preserve a certain amount of water in
the fibres during the measurement. This result in a shift of the endotherms to lower temperatures
compared to the dry state. Similar to Spei56
, they interpreted the lower temperature peak of the
doublet at 170°C (at a heating rate of 5°C /min) as originating from melting or rather
denaturation of the α-helical, crystalline material of wool keratin while the broad, higher
temperature endotherms beyond around 185°C is considered to be due to the thermal degradation
of histological components. They rejected thus the hypothesis that the endothermal doublet
originate from the differential melting of the α-form crystallites in the domain of ortho- and para-
cortical cells.
The results of the second DSC method, which investigates keratins in water excess, are
also subject of controversy about the nature of the recorded endotherms. The effect, shifted in the
range of about 130-150°C (depending on keratin type) is associated with melting of the
crystallites or denaturation of proteins. Individual curves show pronounced variation in their
shapes and sizes, this variation being attributed to natural inhomogeneities between different
wool samples or to differences of their (physical and chemical) histories33
. Using differential
thermal analysis (DTA ) and pressure resistant sample containers, Ebert investigated the
transitions of wool fibres in various agents to study the phenomenon of supercontraction59
. For
some of the conditions he applied, they observed multiple transitions at temperatures above
approx. 100°C. Haly and Snaith28
also used DTA to examine the performance of wool samples
sealed into glass containers with various amounts of water. They observed a phase transition,
often a doublet that shifted with water content from approximately 230°C for dry wool to 140°C
for wool in excess water. They proposed that the two peaks of the bimodal endotherm might
Chapter 1
30
correspond to the melting points of α-form crystallites and β-form crystallites (alternative
crystalline form of keratin). The origin of this bimodal endotherm screened by heated wool in
water has attracted a lot of attention from researchers in wool science. Crighton and Hole
developed a measurement cell for high-pressure differential thermal analysis ( HPDTA) to study
the pressure dependence of the denaturation transition temperature of the helical material of
various keratins in water; they recorded also in some cases a doublet, at temperatures around
140°C60,61
. Additionally, they investigated the effects of chemical modifications of merino wool
on the melting endotherm and found that the bimodal endotherm shifted differently, depending
on the type of modifications. Crighton concluded that the bimodal endoderm was linked to the
differential stabilities of the ortho- and para-cortical cells. Wortmann and Deutz investigated the
correlation between the cystine content and melting point of a series of keratin materials (nail,
mohair, wool, etc.)33
. For seven keratins used in the study, the authors found a significant
positive correlation between the cystine level and the melting temperature. From this, they
concluded that the ortho- and para- explanation of the bimodal endotherm was satisfactory,
because the cystine content in the para- cortex of merino wool is, statistically, slightly higher
than that in the ortho-cortex. In a later study62
, the strong bimodality of the curves was re-
evaluated using isolated ortho- and para-cortical cells; the ortho/para hypothesis was
revalidated, being related to fractions of cortical cells differing in cystine content. However,
other studies show that the bimodal endotherm of wool in water excess depend on the
environment and the heating rate of DSC measurement57
.
Relying on these primary investigations and pushed up by the cosmetic industry the
research was extended also to human hair. Cosmetic treatments such as bleaching, perm-waving
and the use of the permanent colorants have been shown to cause changes to the fibre structure
of the hair63,64
, changes noticed by consumers as increased hair breakage, reduced shine, etc..
Leroy et al65
investigated virgin, bleached and perm-waved hair by DSC in the dry state. For the
virgin hair they observed a strong bimodality of the curves, which was considered as being
related to fractions of cortical cells differing in cystine content rather than having filament /
matrix origins. They observed that with bleaching the DSC peak for dry fibres shifts to higher
temperature and the denaturation peak area decreases. Spei and Holzem56
have reported that the
denaturation peak can usually be detected adequately and evaluated also for dry fibres but that
the effect is always secondary in size compared to a large background peak due to general
keratin pyrolysis66
.
To assess the effects of cosmetic processes on the main morphological components, the
Introduction
31
denaturation of human hair was investigated by means of DSC, under aqueous conditions33
, after
having been treated by relevant oxidative, bleaching and reductive, perm-waving processes66,67
.
By measuring in water, the peak shifts to around 150°C (heating rate 10°C/min) and exhibits no
background effects33
. Wortmann et al66
verified that the effect of bleaching and perm-waving
changes the denaturation temperatures and denaturation enthalpies. The resuming of perm-
waving and bleaching steps decreases further the denaturation temperature of hair as well as the
enthalpy. The results indicate that the enthalpy depends probably on the structural integrity of the
α-helical material in intermediate filaments, while denaturation temperature is kinetically
controlled by the density of cross linkages of the matrix, in which the intermediate filaments are
embedded. The data analysis showed that changes of the denaturation enthalpy and therefore the
damage formation in hair generally follows 1st
order kinetics. The work concluded that the DSC
yields the denaturation enthalpy ΔHD which depend on the amount and structural integrity of the
α-helical material in the intermediate filaments (IF), and the temperature TD which is kinetically
controlled by the cross-link density of the matrix (IFAPs) in which the IFs are embedded66
.
However, there is literature data indicating that a decrease of denaturation enthalpy or of other
parameters indicating extensive damage to the IFs is not necessarily accompanied by a loss of X-
Ray measured crystallinity30,68-70
. It was noted that a decreased enthalpy can in fact be not a
genuine denaturation, that is destruction of the α-helix, but a decrease of the ―native‖ fraction
that become undenaturable within the experimental range , due to induced crosslinking of the α-
helical material.
Specific information about the denaturation mechanisms of fibrous proteins and their
activation energies were searched by means of different, experimental and theoretical methods of
non-isothermal solid state reactions kinetics. First approaches in this area have been presented by
Popescu et al71
. It was found that the course of the denaturation process remain largely
unchanged through oxidation, despite the fact that pronounced decreases of denaturation
temperature, as well as of enthalpy occur72
, the denaturation taking place along a pathway that is
largely independent of the temperature and of the previous treatment73
. Therefore, properties and
interactions of the main morphological components of human hair are considered that are
specifically related to the various aspects of their thermal stability.
1.3. Content of this thesis
This thesis is concerned with the behaviour to heat of human hair, as one of the most
important representatives of fibrous proteins family. Analysis of the structural changes induced
Chapter 1
32
by various chemical reagents to the hard alpha-keratin fibres (human hair) are also discussed
together with their way of action.
Chapter 1 (the present chapter) gives a short introduction on the thermal stability of
globular/soluble proteins and summarises the actual understanding of the behaviour of fibrous
proteins when controlled heated
Chapter 2 deals with ―dry DSC‖ experiments and adds evidences to the origin of the
endothermal doublets contradictorily interpreted over the time. By sampling hair fibres while
heating at moments corresponding to thermal events shown on DSC plot, we put into evidence
the melting of cortex followed by pyrolysis of the material through the solid cuticle layer. The
result was the obtaining of tubes made from fibres emptied of cortical material and keeping the
structure and sorption properties of the initial keratin fibre. In spite of similar amino-acid
composition of the cuticle and cortex the two components of the keratin fibre differs
significantly from the point of view of thermal behaviour, which appears like a cortex-cuticle
paradox.
Chapter 3 studies the influence of pH, particularly acid values, on thermal behaviour of
hard alpha-keratins, indicating the limits of the two-phase model used so far to describe the
fibrous proteins. An alternative three-phase model for explaining fibrous hard alpha-keratins
high thermal stability and their reaction with various reagents is proposed. The model is based on
results from DSC study of keratins under various conditions, and is supported by amino-acid
analysis, X-ray diffraction and tensile strength observations. According to the proposed model,
the third phase, the interface between crystalline and matrix phases, made of nonhelical tail
domains of keratin, scaffolds the intermediate filaments and controls their interaction with
chemical reagents as well as their thermal properties.
In Chapter 4 the differential scanning calorimetry (DSC) measurements carried out at
different heating rates were used for the kinetic analysis of the endothermic process assigned to
the denaturation of the helical material from human hair in water excess. We found that the
kinetic mechanism is autocatalytic and that the value of the activation energy is rather close to
disulphide bond scission than to protein denaturation. This allowed us proposing a multistep
mechanism for the thermal denaturation of hard α-keratins in water excess that relies on the 3-
phase model which describes their structure. The limiting step of the thermal denaturation
process is then the scission of S-S bonds between the main morphological components, namely
intermediate filaments (IF) and matrix (IFAP). The theoretical proposed model shows a good
agreement with the experimental recorded data.
Introduction
33
In Chapter 5 the modifications of thermal behaviour of hard alpha-keratin fibres induced
by bleaching, permanent waving and oxidative dyeing are investigated by differential scanning
calorimetry (DSC). Regression analysis of the data from hair samples treated differently shows a
linear correlation between the enthalpy of the denaturation peak recorded by DSC and the
cystine content of the fibre. The experimental results are evaluated within the framework of the
proposed model in which the nonhelical (globular) terminal domains of keratin promote filament
interactions and control the thermal properties of keratin intermediate filaments. Amino-acid
analysis, X-ray diffraction and tensile strength measurements provide evidence that the attack of
chemical reagents occur preponderantly in the matrix and at the interface between filament and
matrix. A possible intermediate state between native and denaturated crystalline helical material
is suggested to account for the increased disorder in the IFs-IFAP package induced by harsh
treatments. The DSC data suggests that hair keratin IFs can modulate their organisation and
thermal properties through chemical induced interactions.
The Appendixes complete this work by adding additional evidences regarding the validity
of our observations. Appendix A deal with the morphology and molecular mobility of fibrous
hard α-keratins as shown by 1H,
13C, and
129Xe NMR, a qualitative model describing the changes
induced in hard α-keratin protein by chemical transformation being developed, that could be
correlated with the changes in domain thickness, phase composition, and molecular dynamics. In
Appendix B the nonisothermal kinetics of previously damaged hard α-keratin thermal
denaturation it is debated while Appendix C critically review the factors influencing the DSC
thermogram of hard α-keratin proteins and the reproducibility of the experimental results.
1.4. References and notes
1. Alberts, B. Essential Cell Biology: An Introduction to the Molecular Biology of the Cell; Garland
Publishing: New York, 1998.
2. Bischof, J. C.; He, X. Ann. NY Acad. Sci. 2005, 1066, 1-22.
3. Gossett, P. W.; Rizvi, S. S. H.; Baker, R. C. Food. Technol. 1984, 38, 67-96.
4. Jackson, W. M.; Brandts, J. F. Biochemistry 1970, 9, 2294-2301.
5. Privalov, P. L.; Potekhin, S. A. Methods. Enzymol. 1986, 131, 4-51.
6. Sturtevant, J. M. Annu. Rev. Biophys. Bioeng. 1974, 3, 35-51.
7. Sturtevant, J. M. Annu. Rev. Phys. Chem. 1987, 38, 463-488.
8. Ma, C. Y.; Harwalkar, V. R. Advances in food and nutrition research 1991, 35, 317-366.
9. Ruegg, M.; Moor, U.; Blanc, B. J. Dairy Res. 1977, 44, 509-520.
10. Koshiyama, I.; Hamano, M.; Fukushima, D. Food Chem. 1981, 6, 309–322.
11. Eyring, H. In The Theory of Rate Processes in Biology and Medicine; Johnson, H. F.; Eyring, H.; Stover,
Chapter 1
34
B. J., Eds.; John Wiley & Sons Inc: New York USA, 1974.
12. Joly, M. A Physico-chemical Approach to the Denaturation of Proteins; Academic Press: London 1965.
13. Cooper, A. In Protein: A Comprehensive Treatise; Allen, G., Ed.; JAI Press Inc: Stamford CT, 1999, p
217–270.
14. Privalov, P. L.; Khechinashvili, N. N. J. Mol. Biol. 1974, 86, 665-684.
15. Arntfield, S. D.; Murray, E. D. I Differential scanning calorimetry as an indicator of protein denaturation
Can Inst. Food Sci. Technol. J. 1981, 14, 289-294.
16. He, X.; Wolkers, W. F.; Crowe, J. H.; Swanlund, D. J.; Bischof, J. C. Annals of Biomedical Engineering
2004, 32, 1384-1398.
17. Miles, C. A.; Avery, N. C.; Rodin, V. V.; Bailey, A. J. Mol. Biol. 2005, 346, 551-556.
18. Lyubarev, A. E.; Kurganov, B. I. J. Therm. Anal. Cal. 2000, 62, 51-62.
19. Cravalho, E. G.; Toner, M.; Gaylor, D. C.; Lee, R. C. In Electrical Trauma: The Pathophysiology,
Manifestations and Clinical Management; Lee, R. C.; Cravalho, E. G.; Burke, J. F., Eds.; Cambridge University
Press: Cambridge, 1992, p 281-300.
20. Sanchez-Ruiz, J. M. Biophys. J. 1992, 61, 921-935.
21. Lumry, R.; Eyring, H. J. Phys. Chem. 1954, 58, 110-120.
22. Lepock, J. R. Int. J. Hyperthermia 2003, 19, 252-266.
23. Chen, S. S.; Wright, N. T.; Humphrey, J. D. J. Biomech. Eng. 1998, 120, 382.
24. Wright, N. T.; Humphrey, J. D. Annu. Rev. Biomed. Eng. 2002, 4, 109-128.
25. He, X.; Bischof, J. C. Crit. Rev. Biomed. Eng. 2003, 31, 355-422.
26. Privalov, P. L. Adv. Protein Chem. 1982, 35, 1-104.
27. Felix, W. D.; McDowall, M. A.; Eyring, H. Text. Res. J. 1963, 33, 465-471.
28. Haly, A. R.; Snaith, J. W. Text. Res. J. 1967, 37, 898-907.
29. Schwenker, R. F.; Dusenbury, J. H. Text. Res. J. 1960, 30, 800-801.
30. Spei, M.; Holzen, R. Meliand Textiber 1989, 70, 371-376.
31. Crington, J. S. Proc 8th Int. Wool Text. Res. Conf., Christchurch, New Zealand, 1990, pp 419.
32. Deutz, H.; Wortmann, F. J.; Höcker, H. Proc. Int. Wool Text. Org. Conf., Istanbul, 1993.
33. Wortmann, F. J.; Deutz, H. J. Appl. Polym. Sci. 1993, 48, 137-150.
34. Feughelman, M. Text. Res. J. 1959, 29, 223-228.
35. Feughelman, M.; Haly, A. R. Kolloid Z. 1960, 168, 107-115.
36. Bendit, E. G.; Feughelman, M. Encyclopedia of Polymer Science 1968, 8, 1-44.
37. Feughelman, M. J. Macromol. Sci. B 1979, 16, 155-162.
38. Fraser, R. D.; MacRae, T. P.; Suzuki, E. J. Molec. Biol. 1976, 108, 435-452.
39. Wortmann, F. J.; Zahn, H. Text. Res. J. 1994, 64, 737-743.
40. Feughelman, M. Text. Res. J. 1994, 64, 236-239.
41. Chapman, B. M. Text. Res. J. 1969, 39, 1102-1109.
42. Hearle, J. W. S. J. Polym. Sci. Part C 1967, 20, 215-251.
43. Chapman, B. M.; Hearle, J. W. S. J. Macromol. Sci., Part B 1968, 2, 697-741.
44. Chapman, B. M.; Hearle, J. W. S. J. Macromol. Sci., Part B 1970, 4, 127-151.
45. Hearle, J. W. S.; Susutoglu, M. 7th Int. Wool Text. Res. Conf., Tokyo, 1985, pp 214-223.
Introduction
35
46. Hearle, J. W. S. Int. J. Biol. Macromol. 2000, 27, 123-138.
47. Crewther, W. G. Text. Res. J. 1965, 35, 867-877.
48. Crewther, W. G. Text. Res. J. 1972, 42, 77-85.
49. Hearle, J. W. S. J. Mater. Sci. 2007, 42, 8010-8019.
50. Bendit, E. G. Text. Res. J. 1966, 36, 580.
51. Menefee, E.; Yee, G. Text. Res. J. 1965, 35, 801.
52. Spei, M. Proc 6th Qinquennial Int. Wool Text. Res. Conf., Pretoria, 1981, pp 263.
53. Spei, M.; Jörrisen, K.; Hack, R.; Föhles, J. Kautsch Gummi Kunstst 1980, 33, 345.
54. Zahn, H.; Spei, M. In Makromolekulares Kolloquium Chemiker-Zeitung: Freiburg, 1978.
55. Spei, M.; Thomas, H. Colloid & Polymer Sci. 1983, 261, 968-969.
56. Spei, M.; Holzem, R. Colloid & Polymer Sci. 1987, 265, 965-970.
57. Cao, J. J. Appl. Polym. Sci. 1997, 63, 411-415.
58. Cao, J.; Joko, K.; Cook, J. R. Text. Res. J. 1997, 67, 117-123.
59. Ebert, G.; Muller, F. H. Int. Wool Text. Res. Conf. Paris, 1965, pp 487.
60. Crington, J. S. 8th Int. Wool Text. Res. Conf., Christchurch, 1990, pp I419.
61. Crington, J. S.; Hole, E. R. 7th Int. Wool Text. Res. Conf., Tokyo, 1985, pp 283.
62. Wortmann, F. J.; Deutz, H. J. Appl. Polym. Sci. 1998, 68, 1991-1995.
63. Strassburger, J.; Breuer, M. M. J. Soc. Cosmet. Chem. 1985, 36, 61-74.
64. Zalfen, A. M.; Wortmann, G.; Wortmann, F. SOFW Journal 2005, 131, 40.
65. Leroy, F.; Franbourg, A. In 8th International Hair Science Symposium; German Wool Research Institute:
Kiel, 1992.
66. Wortmann, F. J.; Springob, C.; Sendelbach, G. J. Cosmet. Sci. 2002, 53, 219-228.
67. Popescu, C.; Wortmann, F.-J. Revue roumaine de chimie 2003, 48, 981-986.
68. Bertrand, L.; Doucet, J.; Dumas, P.; Simionovici, A.; Tsoucaris, G.; Walter, P. J. Synchrotron Rad. 2003,
10, 387-392.
69. Gillespie, J. M. Science 1970, 170, 1100-1102.
70. Schmidt, H.; Wortmann, F. J. Text. Res. J. 1994, 64, 690-695.
71. Popescu, C.; Sendelbach, G.; Wortmann, F. J. The 10th Int. Wool Text. Res. Conf. Aachen, Germany 2000.
72. Wortmann, F. J.; Popescu, C.; Sendelbach, G. Biopolymers 2006, 83, 630-635.
73. Wortmann, F. J.; Sendelbach, G.; Popescu, C. J. Cosmet. Sci. 2007, 58, 311-317.
* Biomacromolecules, submitted 2010
Chapter II : Micro-tubes of keratin. The thermal
stability of cortex and cuticle*
2.1. Introduction
The hair, the filamentous appendage of the skin of vertebrates serving to protect the body
against coldness and wetness, is made of filamentous proteins, the hard alpha-keratins1. The
keratin fibres have a composite structure, with a core-shell arrangement at various levels of
organisation, from the cortex wrapped by cuticle down to the intermediate filament (IF)
surrounded by intermediate filament associated protein (IFAP), or keratin associated protein
(KAP)2.
The keratin fibres exhibit a relatively high thermal stability, the fibre properties remaining
almost intact until 200°C3. The ability to preserve the properties at high temperature, apart from
the obvious benefit for the body protection, is of interest for the potential to design high-
thermally stable proteins.
The DSC investigations of keratin fibres put into evidence an endothermal effect,
sometime a doublet, whose peak at 10 K/min heating rate ranges from 230 to 260°C, depending
on the fibre source4-7
. The endothermal effect is generally attributed to the thermal denaturation
of the -helix which makes the crystalline part of the keratin fibre5-7
.
Cuticle is formed of four layers with different content of disulfide and isodipeptide bonds, from
outside towards cortex laying epi-cuticle, a-layer, exo-, and endo-cuticle, respectively1. Because
the volume fraction of the cuticle is of the order of 10% 8 of the total keratin fibre and because it
has similar chemical structure with the cortex, cuticle contribution is often neglected when
keratin properties are measured. Our results shown here suggest that cuticle has to be understood
as a different component of the fibre and its morphology requires more investigation.
2.2. Materials and methods
We used hair fibres of European brown hair as supplied by Kerling Int. Haar Fabrik
GmbH. The L-amino-acids of analytical grade were purchased from Merck.
DSC measurements
Chapter 2
38
The heating experiments were carried out in a DSC-7 Perkin Elmer, using closed
aluminium pans in which two holes were pierced. An empty crucible was used as reference. The
DSC device was calibrated using indium and palmitic acid, both of high purity.
Prior to the measurements, the hair samples were cut into snippets of 1-2 mm and stored under
constant, ambient room conditions (approx 22°C, 55% relative humidity) to ensure invariant
water contents. Samples of 7...10 mg were heated with a heating rate of 10 K/min under a
nitrogen flow of 20 mL/min, for temperature ranging from 50 to 300°C. After recording the
endothermal peak on DSC plot, during subsequent experiments the heating was stopped at
temperature before, on the peak, and after the peak for allowing taking the samples for further
investigations.
Cortex & Cuticle isolation
Cuticle isolation: 4-5 g of hair fibres were cut in 3-5 mm snippets and swelled over night in
a tumbler with 200 ml water. Next day, the fibre-water solution was transferred in a mixer and
hackled 10 times, 1 minute each, between two mixing steps the system being cooled down in an
ice bath. The dispersion of water-cuticle was further separated by the remaining fibres on a
suction filter and centrifuged twice, 30 min 12000rpm. Before analysis, the cuticle residue was
dried over night in an exsiccator.
Cortex isolation: 100 mg of hair fibres (snippets) and 2.5 g corundum were weighed in four
50 ml plastic bottles, filled up with water and shacked it 8 times, one hour each in the Baired-
Tatlock Shaker, after each step the system being cooled down in an ice bath for 1 hour. Then, the
fibres were separated from the solution, washed several times with distilled water and dried over
night in an exsiccator, in the next day being separated from the corundum.
Thermo-gravimetrical analysis
The thermo-gravimetrical analysis (TGA) was performed on a Netzsch Iris TG209C.
Sample of around 10 mg of pure amino-acid was placed in the alumina crucible and heated with
10 K/min under a nitrogen flow of 20 mL/min from room temperature to 400°C. The curve of
weight loss and its derivative indicates the thermal stability of the reagent.
Microscopy & Thermo-microscopy
Scanning Electron Microscopy photos were taken on gold sputter-coated snippets sampled
at temperatures chosen from the DSC curve to lie before, on the peak and after the peak using a
SEM S360 (Zeiss NTS GmbH, Oberkochen) at an acceleration voltage of 15 kV.
Mettler Toledo Thermo system FP 90 with FP82 hot stage and Optical Microscopy was used to
follow the events noticed on DSC under the normal atmosphere.
Micro-tubes of keratin. The thermal stability of cortex and cuticle
39
Amino-acid analysis
Amino-acid analysis was carried out on the snippets before and after the peak. Each
sample was hydrolysed in 3 mL 6N HCl at 110 °C for 24 h. The hydrolysed samples were dried
in a rotary evaporator under heating. Amino acid analysis was carried out using an Analysator
type Alpha-Plus II, Pharmacia.
Moisture sorption-desorption
Moisture sorption-desorption isotherms were measured on a IGA Sorp Moisture
Sorptiometer Analyser, at 25°C, using a programme of increasing in steps of 10% the relative
humidity from 1% RH to 95% RH.
2.3. Results and discussions
The DSC plot in Figure 2.1 shows the endothermic effect occurring at around 240°C.
20
22
24
26
210 215 220 225 230 235 240 245 250
Heat
Flo
w E
nd
o U
p (
mW
)
Temperature (°C)
Peak1 = 233.7 C
ΔH = 7.9 J/g
Peak2 = 243.4 C
ΔH = 2.1 J/g
Onset= 230.2 C End= 248.6 C
Figure 2.1 DSC trace of hair heated with a rate of 10 K/min. The relevant parameters of the
curve are shown
The effect is attributed to the melting of the crystalline phase, that is the unfolding of the
alpha helices, and is termed as the thermal denaturation of keratin fibres9. The process is
irreversible and kinetically controlled by the surrounding environment, which is the amorphous
matrix phase in which the alpha-helices are embedded.
We sampled snippets at various temperatures up to 300°C and examined them at scanning
electron microscope recording photos like those in Figure 2.2. One notices that at temperature
beyond 240°C the cortex seems to vanish and the original fibres turned into tubes made out of
Chapter 2
40
cuticle. Because the diameter of the tubes matches those of hair fibres of around 50 micron we
term them as ―micro-tubes‖. It is also interesting to notice that the scales on the surface of the
fibres vanished too and the surface appears relatively smooth. At around 300°C the micro-tubes
crack down into small pieces.
The results showed in Figure 2.2 suggest that the thermal stability of cortex and of cuticle
differ largely. Similar results were obtained on keratin fibres from various sources, indicating
that this is a more general property10
.
The amino-acid composition of fibre, cortex, cuticle and micro-tubes, given in Table 2.1, is
quite similar. By comparing the composition of tubes with those of the original fibres one notices
the loss of cystine, cysteic acid and serine from the tube composition, while the amount of
glutamic acid and ornithine increases.
Figure 2.2 SEM of keratin fibre snippets sampled at temperature indicated on photo. (Heating
rate of 10 K/min)
Micro-tubes of keratin. The thermal stability of cortex and cuticle
41
Because of the complete loss of cystine in micro-tubes we recalculated (see Table 2.2) the
amino-acid composition of cuticle by considering the vanishing of cystine. The last column lists
the ratio of the amino-acids of the two materials (micro-tubes to recalculated cuticle) and
highlights the values of more than 40 % change.
Amino acid Abrev. Total
Fibre
Cuticle Cortex Micro-tubes
@ 255°C
Cysteic acid CysO3 0.71 1.84 0.68 0.39
Aspartic acid Asp 5.57 3.47 5.73 5.96
Threonine Thr 7.7 4.96 8.08 3.9
Serine Ser 10.78 16.32 11.7 4.24
Glutamic acid Glu 13.44 11.06 13.9 18.09
Proline Pro 8.95 10.92 8.86 12.49
Glycine Gly 6.34 9.66 6.37 10.38
Alanine Ala 4.56 6.06 4.96 8.1
Lanthionine Lan - 0.46 - 0
Valine Val 5.67 8.05 6.13 7.97
Cystine (Cys)2 8.83 9.39 8.94 0
Methionine Met 0.51 0.47 0.29 0.42
Isoleucine Ile 3.38 2.22 2.86 3.39
Leucine Leu 7.46 4.79 6.8 9.24
Tyrosine Tyr 2.03 1.29 1.68 2.49
Phenylalanine Phe 2.32 1.67 1.92 2.47
Ornithine Orn 0.38 0.25 1.49
Lysine Lys 2.95 3.56 2.9 2.42
Histidine His 1.06 0.57 0.99 1.06
Arginine Arg 7.74 2.87 6.96 5.49
Table 2.1 Amino-acid composition (in mol / 100 mol) of whole fibre, cuticle, cortex and micro-
tubes sampled at 255°C
Chapter 2
42
Amino acid Abrev. Recalculated
Cuticle
Micro-tube
@ 255°C
Ratio
Cysteic acid CysO3 2.04 0.39 0.2
Aspartic acid Asp 3.85 5.96 1.55
Threonine Thr 5.50 3.9 0.71
Serine Ser 18.10 4.24 0.23
Glutamic acid Glu 12.27 18.09 1.47
Proline Pro 12.11 12.49 1.03
Glycine Gly 10.71 10.38 0.97
Alanine Ala 6.72 8.1 1.21
Lanthionine Lan 0 0 0
Valine Val 8.93 7.97 0.89
Cystine (Cys)2 0 0 0
Methionine Met 0.52 0.42 0.81
Isoleucine Ile 2.46 3.39 1.38
Leucine Leu 5.31 9.24 1.74
Tyrosine Tyr 1.43 2.49 1.74
Phenylalanine Phe 1.85 2.47 1.33
Ornithine Orn 0.42 1.49 3.54
Lysine Lys 3.95 2.42 0.61
Histidine His 0.63 1.06 1.68
Arginine Arg 3.18 5.49 1.72
Table 2.2 Recalculated amino-acid composition of cuticle compared with those of tubes. The
last column indicates the change of amino-acid concentration of micro-tubes reported to those in
recalculated cuticle.
The last column of Table 2.2 shows the decreasing of serine and lysine and the unexpected
increase of glutamic acid, leucine and ornithine in tubes compared to cuticle initial composition.
It has also to be noticed that, with the exception of cystine, there is no other amino-acid
vanishing totally from the composition of tubes.
Micro-tubes of keratin. The thermal stability of cortex and cuticle
43
Amino acid
Abrev. Weight loss (%) Temperature
interval (°C)
Amino-acid
variation
(%)
Cysteic acid CysO3 n.m. n.m. -80
Aspartic acid
Asp
30
30
192-275
360-451
55
Threonine Thr 90 206-287 -29
Serine Ser 65 158-262 -77
Glutamic acid Glu 60 194-376 47
Proline Pro 100 192-297 3
Glycine Gly 50 212-313 -3
Alanine Ala 100 267-306 21
Lanthionine Lan n.m. n.m. 0
Valine Val 100 220-312 -11
Cystine (Cys)2 85 206-285 -100
Methionine Met 100 260-331 -19
Isoleucine Ile 100 240-307 38
Leucine Leu 100 240-313 74
Tyrosine Tyr 80 270-433 74
Phenylalanine
Phe
60
40
213-294
294-369
33
Ornithine Orn n.m. n.m. 354
Lysine
Lys
35
30
30
222-285
285-379
379-489
-39
Histidine
His
26
35
148-197
247-361
68
Arginine
Arg
15
55
214-283
283-401
72
Table 2.3 Thermal stability in terms of temperature interval of weight loss and percentage of
weight loss over that interval as measured by TGA at 10 K/min under nitrogen draft. The last
column gives the percentage of amino-acid loss or gain in micro-tubes compared to the amount
in recalculated cuticle (see Table 2.2)
Note: ―n.m.‖ stands for ―not measured‖.
Chapter 2
44
Investigating the thermal stability of the individual amino-acids by thermogravimetry
(TGA) with 10 K/min we found, in line with other results from the literature11,12
that all of them
decompose mainly within 200-300°C (see Table 2.3). Particularly glutamic, or aspartic acids as
well as leucine should be decomposed at least 50 % each at the temperature at which the micro-
tubes are obtained.
The amino-acid composition measured by chemical analysis for micro-tubes suggests that
the thermal stability of the amino-acids arranged in the keratin chains of cuticle is higher than
those of the individuals. This still does not explain why the amino-acids arranged in the keratin
chains of the cortex pyrolyse at around 240°C.
Figure 2.3 Snapshots showing the behaviour of keratin snippets heated in silicon oil with 10
K/min. The temperature of each snapshot is indicated on the photo
In order to understand how the micro-tubes emerge we have examined under microscope
snippets of keratin fibre immersed in silicon oil while heating from room temperature to 300°C
with 10 K/min on the Mettler hot stage. The silicon oil isolates the snippets from the
Micro-tubes of keratin. The thermal stability of cortex and cuticle
45
environment, provides a good thermal media at temperatures above 200°C and allows optic
examination under the microscope. Some selected snapshots from the film we recorded are
shown in Figure 2.3.
It appears that the cortex starts melting at a temperature within the interval of DSC endotherm.
Following melting the viscous liquid ―boils‖, and pyrolysis gases escape through pores in cuticle,
or through the ends of the snippet. At the end of the DSC endothermic peak the evaporation
process is completed and the micro-tubes are obtained.
The snapshots in Figure 2.3 show firstly that the DSC endothermic peak recorded on keratin
fibres should be interpreted with care; it cannot be ascribed only to the denaturation of
intermediate filaments.
Secondly, the snapshots indicate how the tubes form. When escaping pores are not available the
pressure of gases achieved inside may blow the cuticle and this explains the broken micro-tubes
found among the others.
Also the snapshots suggest why the micro-tubes form, by showing that the melting is a
prerequisite stage for the process. The cortex contains 21-22 % ordered (crystalline) material13
which melts, while the cuticle is made of amorphous crosslinked material which does not melt.
The different morphology is, very likely, the reason for the micro-tubes formation.
Keratin fibres are known to absorb moisture up to 33 % of their weight and to exhibit a large
hysteresis at desorption (a large difference between sorbed and desorbed moisture amount at the
same relative humidity, RH)14,15
. As indicated by the amino-acid analysis the micro-tubes retain
most of the original structure. Consequently we have investigated how much the sorption-
desorption properties changed for the micro-tubes following the thermal treatment.
Figure 2.4 compares the behaviour of native fibre, separated cortex, separated cuticle and
the tubes. The results show that native fibre and isolated cortex behave almost identically, while
isolated cuticle absorbs less moisture than both of them at humidity values higher than 50 %. The
micro-tubes absorb less moisture than cuticle, but, in absolute values, still a significant amount,
reaching 20 % of the weight at 95 % RH.
The plot (see Figure 2.5) of the differences between the absorbed and desorbed moisture
amount at the same relative humidity for the fibre and tubes, respectively, points out the fact that
the micro-tubes retain significantly larger amount of moisture than native fibres over the most of
the range of relative humidity.
Chapter 2
46
0
5
10
15
20
25
30
0 20 40 60 80 100
Mois
ture
conte
nt
(%)
Relative humidity, RH %
Native
Cortex
Cuticle
Micro Tubs @ 255 C
Figure 2.4 Moisture sorption curves recorded on IGASorp machine for native keratin fibre,
isolated cortex, isolated cuticle and tubes obtained at 255°C, for relative humidity, RH, ranging
from 1 % to 95 % at 25°C
0
1
2
3
4
5
6
0 20 40 60 80 100
Mois
ture
dif
fere
nce
(%
)
Relative humidity, RH %
Native
Micro Tubes @ 255 C
Figure 2.5 The difference between absorbed and desorbed moisture by the samples, for relative
humidity ranging from 1 to 95 %, at 25°C
Micro-tubes of keratin. The thermal stability of cortex and cuticle
47
2.4. Conclusions
The microscopy investigation of the keratin fibres heated with a controlled programme
reveals that the endotherm effects above 220°C are related to more processes than the
denaturation of proteins. The melting and volatilisation of the cortical substance, following its
pyrolysis, occurs also at that temperature.
Although made of similar keratin chains the different morphology of cortex and cuticle,
respectively, leads to different thermal stability of the two components. While the cortex melts
and evaporates above 230°C, the cuticle resists over 250°C. This leads to obtaining keratin-based
micro-tubes which maintain most of the original chemical structure and moisture sorption-
desorption properties.
2.5. References and notes
1. Popescu, C.; Höcker, H. Chemical Society Reviews 2007, 36, 1282-1291.
2. Feughelman, M. Text. Res. J. 1959, 29, 223-228.
3. Istrate, D. Unpublished results
4. Felix, W. D.; McDowall, M. A.; Eyring, H. Text. Res. J. 1963, 33, 465-471.
5. Haly, A. R.; Snaith, J. W. Text. Res. J. 1967, 37, 898-907.
6. Spei, M.; Holzem, R. Melliand Textilber 1989, 70, 786-787.
7. Wortmann, F. J.; Deutz, H. J. Appl. Polym. Sci. 1993, 48, 137-150.
8. Hocker, H. In Wool: science and technology; Simpson, W.; Crawshaw, G., Eds.; Woodhead Publishing:
Cambridge, UK, 2002, p 60-79.
9. Popescu, C.; Wortmann, F.-J. Revue roumaine de chimie 2003, 48, 981-986.
10. Jörissen, K.; DWI an der RWTH Aachen e.V. , 1982.
11. Rodante, F.; Fantauzzi, F.; Catalani, G. Thermochim. Acta 1998, 194, 197-213.
12. Rodriguez-Mendez, L.; Rey, F. J.; Martin-Gil, J.; Martin-Gil, F. J. Thermochim. Acta 1988, 134, 73-78.
13. Cao, J.; Leroy, F. Biopolymers 2005, 77, 38-43.
14. A.F.El-Shimi. Colloid & Polymer Science 1978, 256, 105-114.
15. Robbins, C. R. Chemical and Physical Behavior of Human Hair; Springer-Verlag: New York, USA, 1994.
* Polymer Degradation and Stability, submitted 2010
Chapter III : Thermal denaturation of fibrous hard
α-keratins and the effect of pH*
3.1. Introduction
The hard alpha-keratin is a filaments protein found in mammalian epidermal appendages
(hairs, quills, horn, nails, etc.) distinct from beta/feather keratin (beta sheet-based) found in avian
and reptilian tissues. Keratin-containing tissues were first studied for the economic importance
of animal fibres in the textile industry (wool), along with cosmetic related aspects such as hair
growth and epidermis substitutes. Like other filamentous family members, hard alpha-keratin
fibres act mainly as a mechanical support and are the topic of many investigations1-5
.
The structure of hard alpha-keratin is characterized by three structural hierarchy levels6. At
high resolution, the intermediate filament (IF) protein is made of a central rod domain of
sequences (lA, lB, 2A, 2B) containing a heptad repeat that favours the formation of α-helical
structure, and separated by loop links (L1, L12 and L2)2,4
. At the extremity of the rod domain are
located the globular C- and N-terminal domains arranged mostly in ß-sheet and formed of rich of
sulphur compounds6,7
. Two -helices form a parallel, super helical dimer. At medium resolution,
i.e. the intermediate level arrangement of the heterodimers inside IFs, the molecules are
assembled both longitudinally and laterally in an ensemble called microfibril8. Radially, the
number of molecules, across a keratin IF section, is assumed to be 26–349. The dimers are
associated as straight tetramers with a random orientation6 and this organisation forms a long
cylinder-shaped intermediate filament8 with uniform density. The terminal domains play a
crucial role in directing molecular and filament aggregation2. At low organisation, the bundles of
parallel intermediate filaments are organised in distorted crystalline lateral network and
embedded in a sulfur-rich protein matrix of intermediate filament associated proteins (IFAPs)
and form a macrofibril, the main morphological components of hard alpha-keratin fibres6.
The properties and interactions of the main morphological components of keratin fibres
(IFs and IFAPs) are still under academic debates for understanding how these are specifically
related to the various aspects of fibre stability and properties. The head and tail domains in
keratin molecules generally contain a multitude of sites that allow keratin IFs to form covalent
Chapter 3
50
bonds with other proteins. Characteristically, the end domains of the keratin fibre IF lack the
extended runs of glycine residues found in epidermal IF and contain many cysteine residues.
This enables them to participate in extensive disulphide bond crosslinking with the abundant
cysteine-rich proteins of the fibre3,10,11
the non helical terminal domains of IF chains projecting
into the interfilamentous space and linking with the matrix proteins5. Lysine to glutamine
crosslinks have been also found between the head and the tail domains in all keratin chains7.
Besides, the matrix that fills the spaces between filaments is made of the small keratin proteins
of the cysteine-rich and glycine / tyrosine-rich protein families. The potential interactions of so
many proteins could attain a bewildering complexity3.
Nearly all the available physical models of keratin fibres ground their mechanical
description by a two-phase microfibril / matrix fine structure, the interactions of the keratin
proteins being at the origins of the divergences of the structural mechanic models. The two most
important attempts to consistently interpret the shape of the stress/strain curve in relation with
fibre structure have been made by Hearle10,12-15
and Feughelman16-20
, respectively.
Differential scanning calorimetry (DSC) has been widely used as a tool for protein
studies21-24
. For soluble proteins, the thermally induced process detectable by DSC is the
structural melting or denaturation of the protein. The DSC records an endothermal process
whose peak temperature (Tp) and enthalpy (ΔH) are used for characterising the denaturation of
protein. The peak temperature measures the thermal stability of proteins. Its value is influenced
by the heating rate25
and protein concentration26
, as well as by the biochemical environment
(especially pH)27-29
. The ΔH value is the heat uptake for the unfolding transition, independent of
any denaturation model assumption30
. Assuming a two-state transition model for proteins
denaturation, the fractional heat uptake is correlated with the content of ordered secondary
structure of a protein26,30
.
Little systematic work has been done on the thermodynamics of fibrous proteins,
particularly keratins, because of being poorly soluble and difficult to purify in sufficient
quantities for biophysical studies.
Over the time thermal behaviour of keratins fibres was studied by two major DSC methods,
namely by so-called ―dry DSC‖ and ―wet DSC‖, respectively. The first method comprises
investigations performed while allowing the moisture content of the keratin sample to evaporate
with increasing temperature. The ―dry‖ investigation puts into evidence endothermal effects
above 200°C, sometimes as a doublet31-34
. By the second method keratins are investigated in
water excess, in sealed pressure resistant capsules that keep the water during heating32,35-37
. This
Thermal denaturation of fibrous hard α-keratins and the effect of pH
51
method puts into evidence endothermal effects at about 150°C.
The analysis of the recorded endotherms observed with any of the methods relies on the physical
/ mechanical models describing the keratin fibres behaviour and on the similarities with the
behaviour to heating of globular proteins. Assuming the validity of the two-phase Feughelman
model and a two-state transition model for the denaturation transition as in case of globular-
proteins, Wortmann and Deutz37
suggested that the endothermal process recorded on the DSC
curve of keratins reflects also the progress of the thermodenaturation of the (alpha helix)
crystalline sections of the intermediate filaments. Consequently, it was concluded that the DSC
yields the denaturation enthalpy ΔH, which reflects the amount and structural integrity of the α-
helical material in the IFs, and the peak temperature, Tp, which reflects the cross-link density of
the matrix in which the IFs are embedded38
. Later evidence showed that the dry experiments are
very much obstructed by the pyrolysis reaction occurring at similar temperature and, therefore,
the description of the endothermal peak as showing the denaturation process is not coherent39
.
The wet DSC experiments were, thus, the only ones valid for investigating the thermal
denaturation process of hard alpha-keratins.
Working with wet DSC experiments under controlled pH we present results which
question the accepted view of alpha-keratin thermal denaturation and suggest a three-phase
model for giving account of the facts.
3.2. Materials and methods
The alpha-keratin fibres used for analysis were of Caucasian dark-brown hair, supplied by
KERLING International Haarfabrik GmbH. The fibres were cleaned with 1% Lauryl ether
sulphate (LES) and dried at room temperature prior to work with them. The pH of their aqueous
extract was found to be 6.5 to 7.
DSC measurements
Prior to the measurements the samples were cut into fine snippets (~2mm) and stored
under controlled conditions (~ 24 hours, 22°C, 55% relative humidity) to ensure invariant water
contents. 7…10mg of each sample snippets were weighted and placed in crucibles.
Prior to sealing a crucible, 50 μL of distilled water (pH 6.7) was added, and the sealed crucible
was stored over night (~14 hours preceding the measurement), to allow the fibres to wet.
The DSC experiments were run in a DSC-7 Perkin Elmer, using pressure resistant stainless steel
large volume capsules. DSC calibration was done with indium and palmitic acid, both of high
purity. The temperature ranged from 50 to 180°C at a heating rate of 10 K/min. For each sample
Chapter 3
52
we performed 3…5 measurements and the peak temperature, Tp, and enthalpy, H, of the
endothermal effect were reported as mean values and standard deviations.
Several heating rates of 5, 7.5, 10, 15 and 20 K/min were used for the samples submitted to
kinetic analysis.
Tensile measurements
The measurements were performed in wet conditions, considered to reflect best the
changes at the level of intermediate filaments40-42
.
The tensile measurements were performed using the Miniature Tensile Tester Model 675
(MTT675) and the Fibre Dimensional Analysis Unit Model 765 (FDAS765) of Dia-Stron, UK. A
minimum of 35 single fibres were tested for each sample at a stretching rate of 20 mm/min and a
gauge force of 1 gf, as initial condition.
Prior to loading in the circular cassette, the samples were immersed in distilled water for 120
minutes to allow them wetting. During the measurements the cassettes were also filled with
distilled water to ensure the 100% humidity content during the measurement.
The stress-strain curve recorded for each fibre allows calculating the Young‘s modulus, the yield
strength and the breaking extension and total work, which characterise numerically the fibre
mechanic.
Amino acids analyses were conducted conventionally on ―Alpha Plus‖ Amino acid
Analyser, manufactured by Pharmacia LKB, Freiburg, Germany. The results are expressed in
molar percentage.
X-Ray microdiffraction experiments were performed at the European Synchrotron
Radiation Facility (Grenoble, France) on microfocus Beamline ID1343
. A high intensity
monochromatic beam (wavelength k = 0.961 Å), coming from an undulator and a Si-111 double
crystal monochromator, was focussed with an ellipsoidal mirror (focal spot 20(h) * 40(v) m2)
and then size-limited down to a 5 m diameter circular section by a collimator placed in the focal
plane. A guard aperture (Pt–Ir, 10 m diameter) reduced diffuse scattering from the collimator
exit. Samples were made of 10 hairs mounted on a frame with the hair axis perpendicular to the
X-ray beam on a computer-controlled gantry coupled with a microscope which permitted sample
positioning with a 0.1 µm resolution.
Data collection:
The experiments were carried out with a 320 mm sample–detector distance, which was
calibrated using silver behenate, the first order spacing of which is 58.38 Å. Using a small beam
stop of 300 µm diameter, two-dimensional X-ray scattering patterns were collected from 0.006
Thermal denaturation of fibrous hard α-keratins and the effect of pH
53
to 0.4 Å. Patterns were recorded with 1 s exposure times on a MAR-CCD camera (16 bit
readout; 130 mm entrance window; 2048 · 2048 pixels; pixel size of 78.94 · 78.94 µm2).
Radiation damage on the structure has been verified to only occur after exposure times longer
than 10 s and it is indicated by the strong weakening and broadening of the scattering features.
About five patterns were collected along each hair.
Data analysis has been focussed on the scattering regions that provide information at the various
structural organization levels, viz.:
(i) To the meridian arc located in the 5Å region, which is produced by the regular -
helical coiled-coil packing. The strong intensity of this arc was shown to be related to
the fine configuration of residues44-46
.
(ii) To the fine meridian scattering arc at 67Å; this is indicative for the periodic
architecture of the molecules along the IF.
(iii) To the equatorial small-angle X-ray scattering zone; this is related to the radial
geometry of the filaments and to their lateral packing organization in the matrix. In
particular, the distorted crystalline lateral organization gives rise to a strong
equatorial reflection observed around 90Å47-49
.
The analysis was carried out following two complementary procedures. The position and
intensity of the main scattering features were first estimated and compared from a visual
inspection of all patterns. For the most representative patterns, one-dimensional equatorial
profiles passing through the origin, both along the equator and the meridian, were extracted from
the two-dimensional patterns integrating the intensity around the equator on a 10 pixels thick
rectangular strip. These profiles yielded the precise positions and intensities of the main
scattering features and made the comparison between patterns and modelled profiles more
straightforward. In the SAXS zone the huge scattering intensity (proportional to S-2.3
) was shown
to proceed from nonkeratinous zones in hair47
. This component has been subtracted from the
profile.
Raman Spectra
All Raman spectra were recorded on a RFS 100/s Raman device (BRUKER OPTIK,
Ettlingen). The laser type was Nd: YAG operating at 200 mW of 1064 nm wavelength. A bundle
of keratin fibres was fixed on an aluminium support, the Raman spectra being recorded for the
entire bundle.
A spectra resolution of 4 cm-1
was used. By collecting three spectra from the samples, and taking
an average of these, it was possible to ensure that no sample degradation occurred and that the
Chapter 3
54
spectrum obtained was quite reproducible. The OPUS 4.0 software was used to analyse the
recorded Raman spectra.
The bands of particular interest lie in the wave number range of 500–1800 cm-1
. These are
vibrations assigned to the S—S and C—S bonds of cystine; the amino acids tryptophan, tyrosine,
and phenylalanine; the amide I and amide III vibrations; and the C—C skeletal stretching
vibration of the -helix50
.
The disulfide (-S-S-) content of the keratin samples was compared by estimating the ratio of the
peak area of the S-S band (calculated from the peak to a baseline which was drawn between 470
and 560 cm-1
) divided by the peak area of the C-H band (calculated from the peak to a baseline
which was drawn between 1375 and 1500 cm-1
)51
. Also, the component content at 1671 cm-1
assigned to the β-sheet and/or random coil forms and the component content (α) at 1652 cm-1
assigned to the -helix form of the hair samples was compared by estimating the ratio of the
peak area of each component divided by the peak area of the C-H band (calculated from the peak
to a baseline that was drawn between 1375 and 1500 cm-1
) as described by Kuzuhara50-52
.
Treatments
Damaging: The damage of the keratin fibres was induced by an oxidative bleaching with
IGORA VARIO BLOND PLUS bleaching powder and IGORA ROYAL 20 vol. 6% H2O2
bleaching lotion, commercial products kindly supplied by Schwarzkopf. The bleaching
procedure followed the instructions of use. The fibres were rinsed after bleaching and the pH of
the aqueous extract was checked to be 7.
pH treatment: At a liquor ratio of 1 gram fibres to 200 mL solution, the hair tresses were
immersed in aqueous solutions with different values of pH, for 30 minutes at room temperature.
Afterwards, the fibres were rinsed under tap water for 3 minutes, 2 times subsequently washed
with a solution of Texapon N70, 0.1 mL/L, (70% Natrium Laurethsulfat), warm water 1 minute,
tap water 3 minutes, and immersed over night in distilled water to completely neutralise. The
fibres were eventually dried under hot air blow.
We used acetic acid for adjusting pH from 1 to 7; the values of 7 to 12 were adjusted with
ammonia and pH 13 was achieved with NaOH.
3.3. Results and discussions
The current analysis of DSC data of hard alpha-keratin fibrous proteins is based on the
two-phase model for describing the structure of the fibre. This assumes that a reducing of the
value of enthalpy reflects a reduction of the helix content or a decrease of the ―native‖ fraction
Thermal denaturation of fibrous hard α-keratins and the effect of pH
55
that can be thermally denaturated40
.
Feughelman model16
describes keratin fibres as composed of long, water-impenetrable
relatively rigid cylindrical rods set parallel to the fibre axis and embedded in a water permeable
matrix. According to the model, the α-helices, aggregated in the rods (IFs), form a crystalline,
continuous, axially oriented, elastic filament phase, which is embedded in an amorphous matrix
phase that comprises the non helical proteins (IFAPs) and all other noncrystalline, viscoelastic
components (intermacrofibrillar cement, nuclear remnants, cell membrane complex, cuticle, etc).
The absorption of water by the matrix mechanically softens this phase, whereas the rods are
water impenetrable, thus mechanically unaffected by the presence of water. The viscosity of
matrix phase is believed to control kinetically the thermal stability of the -helical part.
Many experimental methods for estimating thermodynamic parameters for protein
denaturation are based on the assumption of ―two-state‖ mechanism for the system. The
accuracy of the data thus obtained, and the validity of their interpretation are critically dependent
on the validity of this assumption.
The simplest, but most widely used kinetic model to express the ―two-state‖ behaviour is the first
order irreversible rate reaction model. For the case of proteins this is to assume that the process
of denaturation may be represented by a transition between two experimentally distinguishable
states, native (N) and denatured (D):
N Dk
(3.1)
where k is the rate constant. This approximation implies that there are no significant population
of intermediate states, and the transition may be brought about by changes in temperature, pH or
denaturant concentration.
Even if the experimental data are satisfactorily described by this one-step model, the real
mechanism of denaturation can be more complex53
.
Earlier studies showed that in spite of recording a pronounced reduced value of the
enthalpy for weathered wool, no significant changes in X-ray pattern were observed40,54
. It was
then proposed that a decreased enthalpy may in fact do not represent a genuine denaturation - a
destruction of the α-helix - but a decrease of the ―native‖ fraction due to radiation-induced
crosslinking of the material, and this fraction becomes undenaturable within the experimental
range.
The pH is known to have a major influence on the denaturation process of soluble
proteins27-29
and thus is expected to play a major role during thermal denaturation of hard alpha-
Chapter 3
56
keratins too.
Table 3.1 details the influence of a relatively short pH treatment on the wet DSC parameters of
Caucasian hair samples. One notices that the peak position shifts with approximately 5°C
towards higher temperatures and the value of enthalpy increases as a result of low pH treatments,
while the high value of pH (alkaline treatment) has an opposite effect.
Assuming that peak temperature and enthalpy obtained from DSC experiments reflects the two-
phase Feughelman model, their changes showed in Table 3.1 are understood as the change of
matrix cross-link density, which is responsible for the shift of peak temperature, and,
respectively, as the change of the amount of α-helical material, for giving account of the
enthalpy variation. The higher is the cross-link density in the IFAPs, the higher their viscosity is
and the more hindered is the thermo-denaturation in the IFs, and vice versa38
.
Alkaline pH is favourable to cystine disruption and to side reactions such as hydrolysis of
the peptide and amide bonds and formation of new cross-links as lanthionine or lysinoalanine11
.
The variation of both ΔH and Tp as a result of alkaline pH seems, thus, to fit well the classical
interpretation. The decrease of enthalpy indicates that the denaturation of the helical segments of
the IFs occurred as a result of high pH value and there is less amount of crystalline material to
thermal-denaturate; the break of cystine disulphide bond, located preponderantly in the matrix,
leads to the decrease of viscosity and the drop of Tp value is the consequence.
Treatment Tp ± st.dev
(°C)
ΔH ± st.dev
(J/g)
Treatment Tp ± st.dev
(°C)
ΔH ± st.dev
(J/g)
N 150.4 ± 0.3 14.7 ± 0.2 pH 7 150.8 ± 0 14.2 ± 0.5
pH 1 156.0 ± 0.2 16.8 ± 1.6 pH 8 148.7 ± 0.9 13.2 ± 0.1
pH 2 156.1 ± 0.2 17.5 ± 1.7 pH 9 146.8 ± 0.8 12.1 ± 2.6
pH 3 156.4 ± 0.2 16.9 ± 0.2 pH 10 143.6 ± 0.1 12.1 ± 0
pH 4 155.6 ± 0.3 15.1 ± 0.7 pH 11 144.1 ± 0.6 12.0 ± 0.5
pH 5 154.3 ± 0.4 14.9 ± 0.1 pH 12 143.4 ± 0.4 10.9 ± 0.2
pH 6 152.8 ± 1.1 14.7 ± 0.2 pH 13 144.6 ± 0.6 8.0 ± 0.1
Table 3.1 Peak temperature, Tp, and value of enthalpy, ΔH, with standard deviations for wet
DSC experiments recorded at 10 K/min after a pH treatment of Caucasian hair samples. (N)
refers to untreated material which has a pH of the aqueous extract of 7
The effect of pH may be understood by investigating the swelling behaviour of keratin fibre11
.
Thermal denaturation of fibrous hard α-keratins and the effect of pH
57
The swelling of fibres in aqueous solutions after 24 hours or longer exposure at different pH
values exhibits four distinct parts: a minimum of swelling for pH ranging from 4 to 9; above pH
10 a large increase in swelling; for pH from 3 to 1 a slight increase in swelling; below pH 1 a
slight decrease in swelling. The large increase in swelling above pH 10 is largely due to
ionization of di-acid residues of the amino acid in the hair and partly to material. The increase of
swelling for pH ranging from 3 to 1 was reported to be due to the combination of acid with the
dibasic amino-acids.
An increase of swelling leads to a drop in matrix viscosity and thus of peak temperature, Tp.
While this is matched by the results of Table 3.1 at high pH value, the records show an opposite
behaviour at low pH. A sudden reorganization and folding of parts of originally amorphous
matrix is also less probable to occur as a result of an acid treatment in order to justify the ΔH
increase.
Even more striking results were obtained after a relatively short time exposure to low pH of a
previously damaged (through bleaching) hair material (see Table 3.2).
A strong recovery (~30°C) and a plateau for pH of the treatment ranging from 1 to 3 is
readily noticeable (Table 3.2) for the peak temperature of bleached samples subjected to an acid
treatment. This comes at odds with the suggestion of the two-phase model according to which a
decrease of both peak temperature and value of enthalpy indicates a permanent damage of hair.
As data of Table 3.2 suggest, these seems to be reversible effects tuneable by pH variation.
As the results of Table 3.3 show the duration of treatment with pH solution does not play a key
role. It appears to be enough to adjust the pH of the water in the pan of DSC experiment prior to
heating for reaching similar results with a 15-30 minutes separate treatment.
Treatment Tp ± st.dev (°C) ΔH ± st.dev (J/g)
Untreated 150.4 ± 0.3 14.7 ± 0.2
Bleached 128.6 ± 0.7 9.8 ± 0.8
pH1 159.4 ± 0.6 12.5 ± 0.4
pH2 159.4 ± 0.9 12.1 ± 0.6
pH3 159.7 ± 0.6 11.8 ± 0.7
pH5 138.7 ± 0.7 9.8 ± 1.3
Table 3.2 Peak temperature, Tp, and value of enthalpy, ΔH, with standard deviations, recorded
on virgin hair at pH 7, 3 times bleached hair at pH 7, and bleached Caucasian hair samples after
30 minutes exposure to low pH. Wet DSC experiment at 10 K/min
Chapter 3
58
Sample Tp ± st.dev (°C) ΔH ± st.dev (J/g)
Untreated 150.4 ± 0.3 14.7 ± 0.2
Bleached 128.6 ± 0.7 9.8 ± 0.8
t1=15 159.8 ± 1.0 13.1 ± 1.4
t2=5 158.7 ± 0.8 12.6 ± 0.6
t3=0 159.4 ± 0.7 12.7 ± 0.5
Table 3.3 Variations of DSC parameters recorded for hair bleached 3 times and immersed in
water of pH 3 in DSC pan. The capsules were kept for t1, t2, and t3 (minutes) respectively before
starting the heating. The values of untreated hair were recorded at pH 7. Data recorded with
heating rate of 10 K/min
Aminoacid 3x 3x & pH 3 Aminoacid 3x 3x & pH 3
Cysteic acid 4.88 4.88 Methionine 0.19 0.21
Aspartic acid 5.88 5.98 Isoleucine 2.66 2.74
Threonine 8.24 8.51 Leucine 6.36 6.40
Serine 12.07 12.15 Tyrosine 1.38 1.32
Glutamic acid 12.88 12.68 Phenylalanine 1.84 1.83
Proline 7.91 7.71 Ornithine 0.09 0.12
Glycine 6.26 6.24 Lysine 2.88 2.83
Alanine 4.67 4.83 Lanthionine 0.00 -
Valine 6.12 6.20 Histidine 1.02 1.05
Cystine 7.68 7.33 Arginine 6.97 7.05
Table 3.4 Amino acids composition (mol %) of bleached 3 times Caucasian sample at pH 7 and
bleached 3 times Caucasian samples subsequently submitted to a pH3 treatment
In order to clarify whether the acid environment caused damages to the main peptide
chains in hair, amino-acid analysis was carried out before and after exposure to low pH.
The results indicate no significant degree of main chain hydrolysis and also no difference (within
experimental error) of the total amount of S-S content in hair as a consequence of exposure to
acid (Table 3.4).
It is accepted that the extensional fibre properties in the wet state are largely controlled by
the properties of the intermediate filaments, because IFs are water-impenetrable, and the matrix
is weakened by water19,40,42
. As a consequence the variation of the wet tensile strength should
Thermal denaturation of fibrous hard α-keratins and the effect of pH
59
correlate with the change of DSC parameters, mainly with the change of enthalpy. This is
suggested by the usual DSC interpretation, which, as mentioned, is derived from physical models
of stress– strain measurements.
Figure 3.1 Wet tensile properties recorded after a pH treatment for native Caucasian hair,
respectively bleached, as ―box & whisker‖ plots, characterized by arithmetic means, the standard
errors (box) and the expectation ranges for the 95 % confidence limits (whisker). (N) refers to
untreated material and (B) to previously bleached damaged keratin, both at pH 7
On the other hand, the investigation of the tensile strength of dry fibre could lead to
unreliable evidence because of ionic and hydrogen bonds that become effective and hide the
damages of a chemical attack15,38,55
.
Figure 3.1 summarises the influence of the pH treatment on the wet tensile strength. In
spite of the large changes recorded on DSC plot after a short pH exposure of hair, the variation
of the fibre tensile strength appears to be insignificant. This has been already expected from the
amino-acid analysis, which did not identify any break of covalent bonds or formation of new
Chapter 3
60
cross-links of the keratin. The value of the yield strength (the stress at which a material begins to
deform plastically), a property that was found56
to be generally sensitive to structural
modifications of keratin fibres, varies also insignificantly.
The X-ray diffraction patterns of hard alpha-keratins is of central importance in structural
studies of this material because the agreement between calculated and observed intensities
provides a searching test of the correctness of any proposed model57
.
The structure of hard alpha-keratin in hair shaft was inferred from X-ray scattering and electron
microscopy analyses. Along the meridian axis (fibre axis), the dimmers characterized by a
regular alpha-helical coiled-coil folding in the rod central domain give rise to the wide-angle X-
ray scattering (WAXS) meridian arc located in the 5.15 Å region. The strong intensity of this arc
was shown to be related to the fine configuration of residues45,46
. At the small-angle X-ray
scattering (SAXS) region, the strong and fine 67 Å meridian scattering arc is related to an axial
stagger between molecules or group of molecules along the microfibril58-60
, its position being
almost insensitive to humidity variations59,61
. Along the equatorial axis, the X-ray pattern gives
poor information about the intermediate scale arrangement of the chains inside IFs. At WAXS
equatorial region, the broad scattering maximum located at 9.5 Å peak is supposed to be due to
interferences between coiled coil chains48,49
or chains distance from others structures62
. In the
SAXS equatorial region three broad peaks corresponding to the distances 90 Å, 45 Å, and 28 Å
(respectively located at S = 0.012, 0.022, and 0.036 Å-1
) are provided from the dense lateral IF
packing. Hair contains crystallized lipids,63
more precisely soaps,64
that give rise to a series of
rings, of which the first order is superimposed on the peak at 45 Å. The signals due to soaps are
the only variable scattering signals displayed by hair; the signals due to keratin are fairly sample-
independent. The pioneering X-ray scattering analyses of Fraser have established that the IFs are
located at the nodes of a distorted two-dimensional quasi-crystalline array48,49
.
This model was later refined using an analytical description of the corresponding small-angle X-
ray scattering (SAXS) equatorial X-ray scattering pattern47
. So, the dense lateral packing of the
microfibrils embedded in the matrix namely the microfibril-matrix network can be investigated
in this region. The position and the intensities of these peaks are characteristic of microfibril
diameter and of the mean of the centre-to-centre distance between microfibrils59
; when the hair
fibre is immersed in water, the 90 Å peak position increase indicating a matrix swelling59,65
.
Thermal denaturation of fibrous hard α-keratins and the effect of pH
61
Figure 3.2 Changes in meridian (top) and equatorial (bottom) intensities of X-ray pattern of
native and three times bleached Caucasian hair samples after 30 minutes exposure to pH1
Figure 3.2 (top) shows that beside an increased disorder degree between the IFs, in the
WAXS region (0.22 to 0.066 Å-1
), the coiled coil configurations remain virtually unaffected by
acid uptake through keratin structure. As it appears, there is no change of the crystalline degree
to account for the increase of enthalpy. The high intensity peak at 66 Å of the sample of three
Chapter 3
62
times bleached hair and kept at pH 1 is probably from the parasitic scattering from the air. This
point of view is also supported by the absence of large intensity differences at the peak from 5.17
Å.
In the SAXS zone (0.066 to 0.01 Å-1
) in the Figure 3.2 (bottom), the huge scattering intensity
(proportional to S-2.3
) was shown to proceed from non keratinous zones in hair47
. This component
has been subtracted from profile. The reflection at 45 Å is provided by crystallised lipids. The
calculation of inter-microfibril distances gives 93.121 Å for native hair and 90.211 Å for hair
kept at pH 1 (both native and bleached 3 times) respectively, while the diameter of the
microfibril was found of 37.5 Å in all cases. Furthermore, from the lateral organisation (i.e.
equatorial intensities), no significant changes were recorded for the 90 Å peak position.
Consequently we may assume that there is no significant change of the fibre structure (matrix
swelling or shrinking, IFs amount increase or decrease) to justify the shift of the Tp, or enthalpy,
H, with such a magnitude.
200400600800100012001400160018002000
Wavenumber (1/cm)
Inte
nsi
ty (
a.u.)
.
Bleached 3x
Bleached 3x & pH
S-S
C-H
Amide III
200400600800100012001400160018002000
Wavenumber (1/cm)
Inte
nsi
ty (
a.u.)
.
Bleached 3x
Bleached 3x & pH
S-S
C-H
Amide III
Figure 3.3 Raman spectra of bleached 3 times respectively pH 1 treated wool bundles
Raman spectroscopy as analytical tool for studying keratin fibres provides information
about –SS– groups status and as well about the state of ordered, respectively unordered proteins.
White wool bundles were used as sentinel sample (followed an identical damaging (i.e.
bleaching) / pH treatment as the hair samples) in order to prevent fluorescence, as white wool
does not contain melanin granules. Even in the case of using samples that included only small
Thermal denaturation of fibrous hard α-keratins and the effect of pH
63
amounts of melanin granules, no usable spectrum from the samples could be obtained because of
an increasing baseline due to fluorescence.
The Raman spectra of bleached 3 times respectively bleached 3 times and pH 1 treated
wool samples are shown in Figure 3.3 As it is revealed the S–S band intensity do not change
significantly indicating that the –SS–groups do not reform as a consequence of an acid treatment.
The S-O band intensity at 1040 cm-1
, assigned to cysteic acid51
remained as well unchanged .
The band component observed at 1671 cm-1
has been assigned to the β-sheet and / or random coil
forms, and the band component at 1652 cm-1
has been assigned to the -helix form51
. No shifts
of these peaks or modifications of their area were noticed as a result of low pH exposure.
Also, the amide III (unordered) band intensity at 1243 cm-1
, assigned to the random coil form do
not suffer significant modifications that may explain the shift of the Tp or ΔH.
As it has been previously mentioned, the current analysis of DSC data of thermal behaviour
of keratin materials is based on the two-phase model, according to which the status of crystalline
IFs is linked to the value of endothermal enthalpy, and those of the amorphous matrix, to the
peak temperature, Tp. The experimental data from amino-acids analysis, tensile measurements
and X-ray diffraction experiments reported in this study indicate some flaws of this
interpretation. It is generally accepted that the decrease of both peak temperature and enthalpy
indicates the damage of keratin. The results of pH influence suggest a reversible effect of both
parameters, which comes at odds with the usual model. The reversibility observed after a pH
treatment of a severely bleached keratin fibre is likely to be due to a change of the environment
of the intermediate filaments, rather than a change of the amount of crystalline material in the
IFs.
Since two separated phases, as Feughelman‘s model assumes, cannot explain the data, we
consider a variant in which the nonhelical terminal domains that project into the interfilamentous
space and link with the matrix proteins control and enhance the thermal properties of keratin
filaments. In other words we propose that the interface phase, lying between crystalline and
matrix phase plays a more important role than it was assumed so far.
Figure 3.4 gives a simplified schema of a microfibril with protofibrils showing the α-
helical rods and the non helical terminal domains projecting into the interfilamentous space and
linking with the matrix proteins through disulphide bonds, as suggested by Chapman and
Hearle‘s model14
. The terminal domains contain, besides cystine, glycine, threonine, valine,
alanine and serine, acidic sites as glutamic and aspartic acid66
. This way the electrostatic forces
may well play a role for the stability of IFs in native fibre. In our view, this scaffolding structure
Chapter 3
64
at the IFs surface made by the side-chain interactions that anchor microfibrils to matrix (interface
phase) assists the thermal stability and the primary control over the denaturation of helical
material of keratins materials when heated. We consider, as Crewther suggested,67
that the
matrix structure is in globular form, with a high density of disulphide cross-links and other
bonds. It has a protective role and the capacity to participate to the formation of a solid interface.
The model allows explaining the effect of pH on thermal denaturation of hard alpha-
keratins in terms of mechanism of ions (ligand) binding. The ligands are the aqueous protons
(H+) that bind to the specific sites in both folded and unfolded states of the IFs and IFAPs, in a
similar manner as in the case of soluble proteins30
. The plateau recorded for peak temperature,
Tp, when pH ranges from 1 to 3, implies that both IFs and IFAPs phases are fully liganded, the
interface recording the maximum cross-link density attainable under the given conditions.
The heat of the enthalpic process recorded by DSC for fibrous keratins is a cumulative
effect of all participating groups. The observed increase of H arises from the additional energy
required to remove the ligand prior to unfolding of the helical components and collapse of the
IFs. Analogous, a decrease of enthalpy as a result of damaging treatments, which does not reflect
in a change of the X-ray diffraction pattern, is due to interface weakening, rather than to a
change of fibre crystalline degree.
As it appears, the interface is playing the role of a shield for the intermediate filaments. The
shield rigidity is tuned with aqueous protons concentration and it controls the thermal
denaturation behaviour of the intermediate filaments after the matrix fails.
The S-S cross-links from the interface bring also their influence over the denaturation process.
As shown schematically in Figure 3.4 the disulphide bonds anchor the interface on the matrix
and any damage of the scaffold required for inducing the denaturation process needs the
breaking of these bonds.
The mechanism of thermal denaturation of keratins should therefore follow several steps.
Beyond a certain temperature, the temperature rise lead to the break of scaffold structure of IFs.
Once set free, the IFs denaturate. This inevitably involves a transition from a relatively compact
ordered structure to a more flexible, disorganized, open polypeptide chain. As the process of
denaturation proceeds the protein molecules unfold and the intern hydrophobic regions expose to
the outside of the molecules. The hydrophobic groups in water tend to cluster, leading to
associates of molecules. This is only possible after the polypeptide chains are set free from the
proposed scaffold structure. Since the covalent S-S bonds control the strength of this interface,
their break is the rate determining step of the denaturation process, and the total heat recorded
Thermal denaturation of fibrous hard α-keratins and the effect of pH
65
(ΔH) is the sum of all those heats required for the stability of the IF structure.
Thermodynamically, denaturation is viewed as the transfer of enough energy to a native protein
such that an alteration in its molecular structure can take place.
Figure 3.4 A 3-D structure of hard-alpha keratin intermediate filament embedded in the matrix,
build for giving account to thermal behaviour of the keratinous fibres. The central rod domain is
dominated by α-helical subsegments (1A, 1B, 2A, 2B, graphed as small rectangles inside the
rod) separated by short linker regions (L1, L12, L2, graphed as lines between the rectangles).
The rod is flanked by nonhelical head and tail domains at the NH2- and COOH- termini that
extend into the matrix through cystine bonds and link with the matrix proteins. Together with
other linkages, the interface, as the scaffolding structure at the surface of IFs, controls and
Chapter 3
66
enhances the thermal properties of keratin filaments. For the sake of clarity, some terminal
domains projections are omitted.
This energy includes a kinetic component, as the activation energy required to overcome the
barrier, and an enthalpic part which is the heat absorbed or released27
.
The mechanism proposed for thermal denaturation of fibrous keratins is more complex than the
two-state one used for soluble proteins. We assume a multi-step process, with a rate limiting
reaction which is the breaking of the scaffold.
The reaction rate of the endothermal process recorded on DSC is given by:
fk(T)dt
dα (3.2)
where dα/dt is the reaction rate, k(T) is rate constant as a function of the absolute temperature T,
and f(α) is the function of conversion. The rate constant is an Arrhenius type function:
k(T) = A exp(-E/RT) (3.3)
where A is the pre-exponential factor and E is the activation energy.
The activation energy (kinetic barrier), E, that determines the temperature and time
dependence of the endothermal process, was calculated by applying the iso-conversional
differential method of Friedman to the DSC data collected at several heating rates68
. We obtained
E = 118.8 ± 13.9 kJ/mol for the untreated sample. The value is close to the lower limits of the
range generally associated with protein denaturation (i.e.104.6-836.8 kJ/mol)27
, some authors69
arguing that the protein denaturation usually only occurs for activation energies above 400
kJ/mol. On the other side the activation energy we calculated lays down within the range of those
recorded for the thermal decomposition of polysulphurs (105-210 kJ/mol) and polysulphones
(155 kJ/mol), through a homolytic fission of S-S bond in the aliphatic portion of these
polymers70
. The values calculated for the bleached fibres, as well as previous work, reports that,
despite recording a pronounced decreases of Tp as well of ΔH for a damaging treatment, the
activation energy does not change significantly71
.
These considerations on activation energy support our view of the damaging of the
scaffold, probable by the scission of the disulphide bonds, as a rate limiting step of the
denaturation.
After calculating the activation energy, the function describing the mechanism of the process,
f(α), is found by a methodology described elsewhere72
. We have calculated here the kinetic
function of the form f(α) = αn (1-α)
m, where the indexes n and m were found to be 2/3 and 1,
Thermal denaturation of fibrous hard α-keratins and the effect of pH
67
respectively. It has to be underlined that the analytical form of this function is regarded merely as
fitting parameter than related to a certain mechanism. It is, however, a clear indication that the
mechanism of fibrous hard alpha-keratin thermal denaturation cannot be expressed by the first-
order kinetics demanded by the two-state transition mechanism.
Summing up, the proposed model describes the fibrous keratins as comprised of an
amorphous matrix, a crystalline phase and an interface between the two phases. The role of the
interface appears more important for thermal behaviour than for the mechanics of the fibre. The
thermal denaturation of alpha-keratins requires, as a rate determined step, the breaking of the
scaffold structure, most probable by scission of S-S bonds, for setting free the -helical domains
which unfold.
3.4. Conclusions
We report a strong influence of pH treatment of keratin fibre on its thermal stability as
recorded by wet DSC experiments. The amino-acid analysis, tensile measurements, X-ray
analysis and Raman spectroscopy do not indicate significant changes of chemistry or
crystallinity of the keratin to account for the shifts of peak temperature or enthalpy of the
endothermal process. The results indicate some limits of Feughelman‘s two phase model for
interpreting the structure of hard alpha-keratins. Based on these data we propose a three-phase
model in which the nonhelical (or globular) terminal domains of keratin promote filament
interactions and control the thermal properties of keratin intermediate filaments. The interface
phase scaffolds the intermediate filaments and controls their thermal stability. The thermal
denaturation process of the intermediate filaments can occur only after the scaffold is damaged
irreversibly. Consequently the two-state mechanism for describing thermal denaturation of
proteins is also not applicable to keratin fibres, for which a multistep reaction is more probable.
Kinetic investigation of the data acquired at several heating rates supports this view and suggests
the scission of S-S bonds as the limiting step of the thermal denaturation process.
3.5. References and notes
1. Fraser, R. D. B.; MacRae, T. P.; Sparrow, L. G.; Parry, D. A. D. Int. J. Biol. Macromol. 1988, 10, 106-112.
2. Parry, D. A. D.; Fraser, R. D. B. Int. J. Biol. Macromol. 1985, 7, 203–213.
3. Powell, B. C.; Rogers, G. E. In Formation and Structure of Human Hair; Jolles, P.; Zahn, H.; Höcker, H.,
Eds.; Birkhäuser Verlag: Basel Switzerland, 1997, p 59-148.
4. Steinert, P. M.; Torkia, D. R.; Mack, J. W. In The Biology of Wool and Hair; Rogers, G. E.; Reis, P. J.;
Ward, K. A.; Marshall, R. C., Eds.; Chapman & Hall: London, 1989, p 157-167.
Chapter 3
68
5. Zahn, H. Melliand Textilber 1991, 72, 926-931.
6. Er Rafik, M.; Doucet, J.; Briki, F. Biophys. J. 2004, 86, 3893-3904.
7. Parry, D. A. Adv. Protein Chem. 2005, 70, 113-142.
8. Birbeck, M. S. C.; Mercer, E. H. J Biophys. Biochem. Cytol. 1957, 3, 203-214.
9. Engel, A.; Eichner, R.; Aebi, U. J. Ultrastruct. Res. 1985, 90, 323-335.
10. Hearle, J. W. S. Wool Tech Sheep Breed 2003, 51, 95-117.
11. Robbins, C. R. Chemical and Physical Behavior of Human Hair; Springer-Verlag: New York, USA, 2002.
12. Hearle, J. W. S. J. Polym. Sci. Part C 1967, 20, 215-251.
13. Hearle, J. W. S. Int. J. Biol. Macromol. 2000, 27, 123-138.
14. Hearle, J. W. S. J. Mater. Sci. 2007, 42, 8010-8019.
15. Hearle, J. W. S.; Susutoglu, M. 7th Int. Wool Text. Res. Conf., Tokyo, 1985, pp 214-223.
16. Feughelman, M. Text. Res. J. 1959, 29, 223-228.
17. Feughelman, M. J. Macromol. Sci. B 1979, 16, 155-162.
18. Feughelman, M. Text. Res. J. 1994, 64, 236-239.
19. Feughelman, M. J. Appl. Polym. Sci. 2002, 83, 489-507.
20. Feughelman, M.; Haly, A. R. Kolloid Z. 1960, 168, 107-115.
21. Jackson, W. M.; Brandts, J. F. Biochemistry 1970, 9, 2294-2301.
22. Privalov, P. L.; Potekhin, S. A. Methods Enzymol. 1986, 131, 4-51.
23. Sturtevant, J. M. Annu. Rev. Biophys. Bioeng. 1974, 3, 35-51.
24. Sturtevant, J. M. Annu. Rev. Phys. Chem. 1987, 38, 463-488.
25. Ruegg, M.; Moor, U.; Blanc, B. J. Dairy Res. 1977, 44, 509-520.
26. Koshiyama, I.; Hamano, M.; Fukushima, D. Food Chem. 1981, 6, 309–322.
27. Bischof, J. C.; He, X. Ann. NY Acad. Sci. 2005, 1066, 1-22.
28. Eyring, H. In The Theory of Rate Processes in Biology and Medicine; Johnson, H. F.; Eyring, H.; Stover,
B. J., Eds.; John Wiley & Sons Inc: New York USA, 1974.
29. Joly, M. A Physico-chemical Approach to the Denaturation of Proteins; Academic Press: London 1965.
30. Cooper, A. In Protein: A Comprehensive Treatise; Allen, G., Ed.; JAI Press Inc: Stamford CT, 1999, p
217–270.
31. Felix, W. D.; McDowall, M. A.; Eyring, H. Text. Res. J. 1963, 33, 465-471.
32. Haly, A. R.; Snaith, J. W. Text. Res. J. 1967, 37, 898-907.
33. Schwenker, R. F.; Dusenbury, J. H. Text. Res. J. 1960, 30, 800-801.
34. Spei, M.; Holzen, R. Meliand Textiber 1989, 70, 371-376.
35. Crington, J. S. Proc. 8th Int. Wool Text. Res. Conf., Christchurch, New Zealand, 1990, pp 419.
36. Deutz, H.; Wortmann, F. J.; Höcker, H. Proc. Int. Wool Text. Org. Conf., Istanbul, 1993.
37. Wortmann, F. J.; Deutz, H. J. Appl. Polym. Sci. 1993, 48, 137-150.
38. Wortmann, F. J.; Springob, C.; Sendelbach, G. J. Cosmet. Sci. 2002, 53, 219-228.
39. Popescu, C.; Wortmann, F. J. J. Therm. Anal. Cal. submitted 2007.
40. Schmidt, H.; Wortmann, F. J. Text. Res. J. 1994, 64, 690-695.
41. Wortmann, F. J.; De Jong, S. Text. Res. J. 1985, 55, 750-756.
42. Wortmann, F. J.; Zahn, H. Text. Res. J. 1994, 64, 737-743.
Thermal denaturation of fibrous hard α-keratins and the effect of pH
69
43. Riekel, C.; Burghammer, M.; Mueller, M. J. Appl. Crystallogr. 2000, 33, 421-423.
44. Briki, F.; Doucet, J.; Etchebest, C. Biophys. J. 2002, 83, 1774-1783.
45. Busson, B.; Briki, F.; Doucet, J. J. Struct. Biol. 1999, 125, 1-10.
46. Busson, B.; Doucet, J. J. Struct. Biol. 1999, 127, 16-21.
47. Briki, F.; Busson, B.; Doucet, J. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 1998, 1429, 57-68.
48. Fraser, R. D. B.; MacRae, T. P.; Miller, A. J. Mol. Biol. 1964, 10, 147-156.
49. Fraser, R. D. B.; MacRae, T. P.; Miller, A.; Suzuki, E. J. Mol. Biol. 1964, 9, 250-252.
50. Kuzuhara, A. Biopolymers 2005, 79, 173-184.
51. Kuzuhara, A. Biopolymers 2007, 85, 274-283.
52. Kuzuhara, A. Biopolymers 2006, 81, 506-514.
53. Lyubarev, A. E.; Kurganov, B. I. J. Therm. Anal. Cal. 2000, 62, 51-62.
54. Spei, M.; Holzem, R. Melliand Textilber 1989, 70, 786-787.
55. Istrate, D.; Popescu, C.; Er Rafik, M.; Möller, M. J. Soc. Cosmet. Chem. submitted 2009.
56. Istrate, M. In Unpublished results.
57. Crewther, W. G.; Fraser, R. D. B.; Lennox, F. G.; Lindley, H. Adv. Protein Chem. 1965, 20, 191-346.
58. Fraser, R. D. B.; MacRae, T. P.; Parry, D. A. D.; Suzuki, E. Proc. Natl. Acad. Sci., USA, 1986, pp 1179-
1183.
59. Kreplak, L.; Franbourg, A.; Briki, F.; Leroy, F.; Dalle, D.; Doucet, J. Biophys. J. 2002, 82, 2265-2274.
60. Steinert, P. M.; Marekov, L. N.; Fraser, R. D. B.; Parry, D. A. D. J. Mol. Biol. 1993, 230, 436-452.
61. Franbourg, A.; Leroy, F.; Lévêque, J. L.; Doucet, J. In 5th International Conference on Biophysics and
Synchrotron Radiation: Grenoble, France, 1995.
62. Rao, D. R.; Gupta, V. B. J. Appl. Polym. Sci. 1992, 46, 1109-1112.
63. Fraser, R. D. B.; MacRae, T. P.; Rogers, G. E.; Filshie, B. K. J. Mol. Biol. 1963, 7, 90-94.
64. Briki, F.; Mérigoux, C.; Sarrot-Reynauld, F.; Salomé, M.; Fayard, B.; Susini, J.; Doucet, J. J. Phys. IV
2003, 104, 337-340.
65. Fraser, R. D. B.; MacRae, T. P.; Millward, G. R.; Parry, D. A. D.; Suzuki, E.; Tulloch, P. A. Appl. Polym.
Symp. 1971, 18, 65-77.
66. Robbins, C. R. Chemical and Physical Behavior of Human Hair; Springer-Verlag: New York, USA, 1994.
67. Crewther, W. G. Text. Res. J. 1965, 35, 867-877.
68. Friedman, H. L. Polym. Lett. 1966, 4, 323–328.
69. Lepock, J. R. Int. J. Hyperthermia 2003, 19, 252-266.
70. Vasile, C.; Calugaru, E. M.; Stoleriu, A. S. M.; Mihai, E. Comportarea termica a polimerilor (Thermal
Behaviour of Polymers): Bucuresti, Romania, 1980.
71. Wortmann, F. J.; Popescu, C.; Sendelbach, G. Biopolymers 2006, 83, 630-635.
72. Braese, S.; Dahmen, S.; Popescu, C.; Schroen, M.; Wortmann, F. J. Chem. Eur. J. 2004, 10, 5285-5296.
* Macromolecular Bioscience, 9(8), p805-812, 2009
Chapter IV : Nonisothermal kinetics of hard α-
keratin thermal denaturation*
4.1. Introduction
Proteins exhibit their active properties within certain ranges of values of temperature and
unfold beyond the limits. The intervals are around the room temperature, although for many
industrial applications of enzymes it would be of interest to have them active at higher values of
temperature. A few proteins withstand temperatures above 100°C and among them there are
some of the fibrous proteins.
Fibrous proteins are distinguished from globular proteins by their filamentous, elongated
form. Most of them play structural roles in animal cells and tissues. Among the most well-known
representatives of this class are the α-keratins in human hair, wool and finger nails, fibroin in
silk, actin and myosin in muscles, and collagen, the most abundant protein in vertebrate bodies.
Work on fibrous proteins is less extensive, with the possible exception of the myosin /
tropomyosin family of α-helical coiled-coil proteins1 because of being poorly soluble and
difficult to purify in sufficient quantities for biophysical studies.
Specific information about the denaturation mechanism of hair material and its activation
energy were searched by means of different methods of non-isothermal solid state reactions
kinetics2.
Assuming the validity of the two-phase Feughelman model and a 2-state transition model
for the denaturation transition as in case of globular-proteins, Wortmann and Deutz suggest that
the endothermal process recorded on the DSC curve of keratins reflects the progress of the helix-
coil transition in the crystalline sections of the intermediate filaments3. Feughelman model
understands keratin fibres as composed of long, water-impenetrable relatively rigid cylindrical
rods set parallel to the fibre axis and embedded in a water absorbing matrix4. According to the
model, the α-helices aggregated in the IFs, form a crystalline, continuous, axially oriented,
elastic filament phase, which is encapsulated in the amorphous matrix phase that comprises the
non helical proteins (IFAPs) and all other noncrystalline, viscoelastic components
(intermacrofibrillar cement, nuclear remnants, cell membrane complex, cuticle, etc). The
Chapter 4
72
absorption of water by the matrix mechanically weakens this phase, whereas the rods are water
impenetrable, thus mechanically unaffected by the presence of water. Consequently, it was
concluded that the denaturation enthalpy ΔH and the peak temperature, Tp, recorded by DSC
reflects the amount and structural integrity of the α-helical material in the intermediate filaments
(IFs), and respectively the cross-link density of the matrix (IFAPs) in which the IFs are
embedded5.
In soluble proteins the helix thermally denatures predominantly at temperatures up to
80°C. There are no data on the denaturation temperature of IFs alone (not surrounded by a
matrix) but one may expect that α-helix from keratins would also unfold at temperatures below
80°C. The fact that keratin proteins show the peak temperature at above 100°C is assumed that is
due to the rigidity of the matrix, whose viscosity impedes the unfolding of the helix. The
viscosity (and crosslink) of the encapsulating matrix governs, therefore, the segmental mobility
of the α-helix and the unfolding reaction (denaturation).
Understanding properly how the keratins protect the Intermediate Filaments against
thermal denaturation until high values of temperature is of a clear interest for the fundamental
knowledge of protein denaturation. The role of matrix in this process may suggest ways for
designing high-temperature stable proteins as new biomaterials.
In spite of all the work, there is still no mechanism proposed for describing how the
thermal denaturation process occurs in hard α-keratins. Using a non-isothermal approach, we aim
at proposing a kinetic mechanism for giving account of this as a model for an encapsulated
protein.
4.2. Materials and methods
The α-keratin fibres used for analysis were of Caucasian dark-brown hair, supplied by
KERLING International Haarfabrik GmbH.
DSC measurements
There are two ways of measuring the DSC of keratins, namely in water and in dry
environment (allowing the moisture to evaporate during heating). The DSC in dry environment
was shown to give misleading information, due to the interference of pyrolysis with the process
of interest6. Consequently the present work deals exclusively with DSC of keratins in water
excess.
Prior to the measurements the samples were cut into fine snippets (~2mm) and stored
under controlled conditions (~ 24 hours, 22°C, 55% relative humidity) to ensure invariant water
Nonisothermal kinetics of hard α-keratin thermal denaturation
73
contents. 7…10mg of each sample snippets were weighted and placed in crucibles.
Prior to sealing a crucible, 50 μL of distilled water (pH 6.7) was added, and the sealed
crucible was stored over night (~14 hours preceding the measurement), to allow the fibres to wet.
The DSC experiments were run in a DSC-7 Perkin Elmer, using pressure resistant stainless
steel large volume capsules. DSC calibration was done with indium and palmitic acid, both of
high purity. We used five heating rates, viz.: 5, 7.5, 10, 15 and 20 K/min for temperature ranging
from 80 to 180°C. Each experiment was repeated three to five times, for ensuring the
reproducibility of data.
4.3. Kinetic modeling: General description of the kinetic method
The goal of this approach is to gain further insights into the process of denaturation of
helical material and into the interactions between filaments and matrix.
When an effect is recorded during a DSC experiment, while the sample is subjected to a
controlled temperature ramp, it is reasonable to assume that it reflects a transformation which
can be represented for the simplest case as:
1Bk
B 0 (4.1)
where B0 is the material before transformation and B1 the material after transformation. Any
conversion is accompanied by an absorption or release of heat. In a quantitative way it is
expressed by means of the enthalpy of the process:7
2
1
2
1
t
t
p
T
T
p dtcdtcH (4.2)
where ΔH is the enthalpy, t is the time, T is the temperature, cp is the heat capacity and β is the
heating rate:
dt
dT (4.3)
The degree of conversion, α, is then calculated from the DSC curve assuming that the area
under the peak up to a given time is proportional to the degree of conversion:
total
t
PA
PA (4.4)
where PAt is the integral of the peak up to time t and PAtotal is the overall peak area.
With this definition of conversion the reaction rate can be written as:
Chapter 4
74
fk(T)dt
d (4.5)
where dα/dt is the reaction rate, k(T) is rate constant as a function of the absolute temperature, T,
and f(α) is an unknown function of conversion.
Assuming that the rate constant obeys Arrhenius equation:
RT
EaexpAk(T) (4.6)
where Ea is the activation energy of the reaction, T the absolute temperature, R the universal gas
constant, and A is the preexponential factor, one may re-write eqn. 4.5 in a expanded form as:
dα/dt = A f(α) exp(-Ea/RT) (4.7)
The function of conversion, f(α), in eqn. 4.7 is chosen according to experimental data and
is assumed to describe the reaction mechanism. There are many different proposed functions for
the function f(α).8 A quite general one is Sesták- Berggren equation:
9,10
p1lnm1nf (4.8)
where the values of n, m, p allows retrieving the particular form of kinetic models of
heterogeneous reactions.
A convenient change of the variable time (t) into temperature (T), with definition of β from
eqn. 4.3 and the general form of eqn. 4.8 turns eqn. 4.7 into:11
RT
EaexpAln
β
1
dT
dα pmn 11 (4.9)
This equation allows obtaining the group (Ea, A, f(α)), called the kinetic triplet which is
considered to characterise kinetically the investigated process. Any approach of solving eqn. 4.9
for calculating the kinetic parameters of thermally initiated chemical or physical processes has to
take into account the particular form of the signal, for instance whether it displays single or
multiple peaks. In the case of a multiple peak pattern, eqn. 4.9 stands only for separate processes.
For complex competitive or reversible reactions sequences or those complicated by diffusion Ea
varies with α. Under such circumstances eqn. 4.9 remains valid only locally and the kinetic
parameters have to be calculated for each experimental point or rather, on small intervals for
which they are considered as constants12
. The alternative is that a complicated pattern is
separated into elementary processes (deconvoluted) to enable further analysis.
There are several methods for calculating the kinetic triplet, Ea, A and f(α), from thermal
analysis data. In some cases, a certain reaction model is assumed, implying a specific analytical
form of the kinetic function, f(α). There are also the model-free methods which allow calculating
Nonisothermal kinetics of hard α-keratin thermal denaturation
75
the value of the activation energy, Ea, without any knowledge of the reaction pathway. The
advantage of analysing kinetic data using model-free methods is that these methods do not
assume any model or mechanism beforehand, and thus they are able to describe the most
complicated reaction behaviour at different temperatures13,14
. The solely assumption involved by
the use of model-free methods is that the reaction mechanism does not change with temperature
and heating rate13
.
4.3.1. The activation energy, Ea
The forced fitting of experimental data to simple reaction-order kinetic models can produce
significant errors when predicting rates outside the experimental range of temperatures15
. For this
reason, model-free methods are the suitable approach. These are also known as the iso-
conversional methods,15
meaning that the data for calculation are acquired for the same degree of
conversion from a series of experiments conducted at different heating rates. Literature considers
them as the best approach for taking into account reaction mechanisms outside the experimental
range of temperatures15,16
.
There are various iso-conversional methods in the literature17-19
. For the present study we
used the differential, or Friedman,18
method based on the following equation derived from eqn.
4.9 after separating the terms and taking the logarithms:
α
iRT
EaflnA
dT
dln
(4.10)
where Tα is the temperature at which the conversion α is recorded on each DSC plot obtained at
heating rate βi. For the constant , a plot of ln[βi(dα/dT)] versus 1/Tα is a straight line whose
slope allows calculating the activation energy for the corresponding conversion degree, Eaα.
Several values of the degree of conversion, α, are selected, ranging from 0.2 to 0.9. The
beginning and the end of the process are omitted (values below 0.2 and beyond 0.9) as the
initiation and completion of a heterogeneous reaction involves always diffusion and other
physical processes.
4.3.2. The kinetic function, f(α) and the pre-exponential factor, A
Providing the value of Eaα keeps a fairly constant value, Ea, over the range of conversion
degrees considered, one may also evaluate the analytical form of the kinetic function f(α)20,21
.
eqn. 4.9 is rewritten as:
Chapter 4
76
fA
RT
Eaexp
dt
dα
(4.11)
Inserting eqn. 4.4 into eqn. 4.11 and dividing both sides with the maximum values in order
to normalise them, one obtain, after eliminating the constants PAtotal and A:
αfmax
αf
TEa/R/exp/dtPAdmax
TEa/R/exp/dtPAdG(t)
t
t
(4.12)
In eqn. 4.12 the left side function, G(t), ranging from 0 to 1, derives only from
experimental data (relative heat flow values on the DSC plot and the temperatures at which the
values were measured) and from the value of Ea. This function has to be fitted by the right-hand
side function, which is normalised kinetic function only, as A vanishes through division.
In order to find the analytical form of the kinetic function f(α) a non-linear estimation
programme can be used for finding the best values of m, n, p from eqn. 4.9 for fitting the curve
G(t), satisfying the statistical criterions (i.e. minimising the sum of residuals).
Furthermore, dividing eqn. 4.11 by the kinetic function obtained by the method described
above one obtain the mean value of parameter A.
4.4. Results and discussions
The DSC recorded at various heating rates for the hair samples exhibit the endothermal
process shown in Figure 4.1.
Figure 4.1 Typical DSC curves of hair material, recorded at various heating rates
Nonisothermal kinetics of hard α-keratin thermal denaturation
77
The DSC plot provides the two important parameters: peak, or denaturation temperature,
Tp, and the enthalpy of the process, ΔH. The denaturation temperature measures the thermal
stability of proteins, and is influenced by the heating rate22
and protein concentration,23
as well as
by the biochemical environment (especially pH),24-26
The ΔH value is the heat uptake for the
unfolding transition, independent of any denaturation model assumption27
.
The activation energy calculated for various conversion degrees, according to eqn. 4.10,
shows the variation given in Figure 4.2.
The overall kinetic parameters are then inferred as:
Ea = 118.8 ± 13.9 kJ mol-1
A = 2.7∙1013
min-1
f(α) = α2/3
∙(1-α)1
In view of the standard deviations, the variation of the activation energy is not very
pronounced (~11%). Despite its small variation, similar to those observed at the denaturation of
collagen,28
Figure 4.2 shows a statistically significant dependency of the effective activation
energy on the extent of conversion (Spearman rank order correlation R: -0.98 at a significance
level p: 0.000033). Revealing the dependence of the activation energy on conversion degree may
well help to disclose the complexity of a process and to identify its kinetic scheme15,28,29
.
Figure 4.2 The dependence of the activation energy Ea on the conversion degree as
determined by Friedman method for native Caucasian hair samples
An increase of the activation energy with conversion generally applies for the thermal
decomposition of many polymers through competing, consecutive, although some independent
reactions15
. The decrease of Ea on α corresponds to the kinetic scheme of an endothermic
reversible reaction followed by an irreversible one 15,28,30
. Such a behaviour is also reported as
applying to processes which proceeds with a change from a kinetic to a diffusional regime15
.
Chapter 4
78
The mean value of Ea as well as its extreme values (135 kJ mol-1
for α = 0.3 respectively
93.2 kJ mol-1
for α = 0.9) are close to the lower limits of the range of activation energy generally
associated with protein denaturation (i.e.104.6-836.8 kJ/mol),24
although some authors31
argue
that the protein denaturation only occurs for activation energy above 400 kJ/mol.
The kinetic function suggests an autocatalytic-like process. It has to be underlined that the
analytical form of this function, as well as the values of the two parameters, should be regarded
merely as fitting parameters than related to a certain mechanism, unless further evidence is
obtained. The aforementioned effects, however, clearly indicate that the mechanism of fibrous
hard α-keratin thermal denaturation is more complex than the first-order kinetics and cannot be
reduced to a single step.
There are many models for describing the protein thermal denaturation. The simplest, but
most widely used kinetic model to express the ―2-state‖ behaviour is the first order irreversible
rate reaction model. For the case of proteins this is to assume that the process of denaturation
may be represented by a transition between two experimentally distinguishable states, native (N)
and denatured (D):
DN k (4.13)
where k is the rate constant. This approximation implies that there are no significant populations
of intermediate states, allowing correlating the fractional heat uptake at any stage in the
transition with the extent of unfolding. The accuracy of the data thus obtained and their
interpretation are critically dependent on the validity of this assumption.
Even if the experimental data are satisfactorily described by this one-step model, the real
mechanism of denaturation can be more complex32
. As it shows the kinetic function calculated
above, this model is unlikely for describing the thermal denaturation of the keratins.
Higher-order kinetic models (i.e., models assuming that one or more intermediate states
exist between the native and denatured states) were also adopted in several studies32-34
with the
Lumry and Eyring model being by far the most popular:35
N U D
k1k2
k-1 (4.14)
(where N, U, and D are native, partially unfolded and denatured protein form; k1, k-1, k2 are the
rate constant for the corresponding reactions). The model describes the irreversible protein
denaturation by at least two steps: (a) reversible unfolding of the native protein (N); (b)
irreversible alteration of the unfolded protein (U) to yield a final state (D) that is unable to fold
Nonisothermal kinetics of hard α-keratin thermal denaturation
79
back to the native one. It is easily noticeable that the one-step model is a particular case of the
Lumry and Eyring model.
There are two main situations when the Lumry and Eyring model (eqn. 4.14) is reduced to
one-step irreversible model (eqn. 4.13). In the first case the value of k2 is much higher than the
values of k1 and k-1, so that the direct reaction of the first step is rate-limiting and the reverse
reaction is practically absent. If this step is fast enough, the DSC transition is entirely determined
by the kinetics of formation of the final state and equilibrium thermodynamics analysis is not
permissible. The second situation is realized when the rates of the direct and reverse reactions of
the first step are much higher than the rate of the second step, but equilibrium for the first step is
shifted toward the form N34
.
This model seems also unlikely to apply to our results.
The properties and interactions of the main morphological components of keratin fibres
(IFs and IFAPs) are still under academic debates for understanding how these are specifically
related to the various aspects of fibre stability and properties. The head and tail domains in
keratin molecules generally contain a multitude of sites that allow keratin IFs to form covalent
bonds with other proteins. Characteristically, the end domains of the keratin fibre IF lack the
extended runs of glycine residues found in epidermal IF and contain many cysteine residues.
This enables them to participate in extensive disulphide bond crosslinking with the abundant
cysteine-rich proteins of the fibre36-38
, the non helical terminal domains of IF chains projecting
into the interfilamentous space and linking with the matrix proteins39
(Figure 4.3). Lysine to
glutamine crosslinks have been also found between the head and the tail domains40
. Besides, the
matrix that fills the spaces between filaments is made of the small keratin proteins of the
cysteine-rich and glycine / tyrosine-rich protein families. The potential interactions of so many
proteins could attain a bewildering complexity37
.
Fibre axis
Amorphous matrix (IFAP)
Crystalline rod (IF)
Figure 4.3 Schematic representation of the molecular model of hair adapted from 41,42
Chapter 4
80
In our view, this scaffolding structure at the IFs surface made by the side-chain interactions
that anchor microfibrils and matrix molecules (interface phase) assists the heat stability and the
primary control over the denaturation of helical material of keratins materials when heated.
The mechanism of thermal denaturation of keratins should therefore follow several steps.
Thermodynamically, denaturation is viewed as the transfer of enough energy to a native
protein such that an alteration in its molecular structure can take place. This energy includes a
kinetic component, as the activation energy required to overcome the barrier, and an enthalpic
part which is the heat absorbed or released24
. We assume, therefore, a multi-step process, with a
rate limiting reaction which is the breaking of the interface IF - matrix (scaffold).
Schematically, the mechanism we propose for the thermal denaturation of keratins is
presented below (Figure 4.4).
SHIFSHIF
SHIFSHIF
MSSSHIFSHMSSIF
ck
unf
unf
k
ko
ok
o
2
1
1
0
Figure 4.4 Proposed reactions sequence for the heat denaturation of hard α-keratins
In Figure 4.4 IFo-SS-M stands for the filament-interface-matrix complex, SH is the sulphur
compound (HS-, or S, according to if cystine degrades following α- or β-elimination route
respectively,43
IF is the intermediate filament, the indexes ―o‖, ―unf‖ and ―c‖ stand for ―initial‖,
―unfolded‖ and ―thermally denaturated‖ respectively, M is the matrix and k are the rate
constants.
The mechanism for the reaction recorded under the DSC in water excess endotherm, as
noticed, is having a multistep character. Due to the temperature at which the first reaction occurs
(above 100°C), the IFs are already in a meta-stable state, practically kept by the scaffold
interface IF-M. This is to say that the values of k2, k1 and k-1 are higher than the value of k0 so
that the direct reaction of the first step is rate-limiting.
When temperature rises the helical domains from the IFs try to unfold. Unfolding
inevitably involves a transition from a relatively compact ordered structure to a more flexible,
disorganized, open polypeptide chain. As the process of denaturation proceeds, the protein
molecule unfolds and the internally directed hydrophobic regions become exposed to the outside
Nonisothermal kinetics of hard α-keratin thermal denaturation
81
of the molecule. Non-polar, hydrophobic groups in water will tend to cluster together because of
their mutual repulsion from water, not necessarily because they have any particular direct
affinity for each other. Therefore, upon unfolding hydrophobic regions on individual protein
molecules will try to associate with hydrophobic regions on other protein molecules. This is only
possible after the polypeptide chains are set free from the proposed scaffolding structure. Since
the covalent S-S bonds control the strength of this interface, the rate determining step of the
process is their scission, and the recorded enthalpy (ΔH) is the consequence of all those
interactions participating to the stability of the IF structure and the unfolding transition of the
helical material. The total heat uptake required for breaking down the scaffold in order to allow
the denaturation of helical material should decrease if the interface is previously damaged.
These allow us speculating that after chemically reducing the disulphide bonds from the
keratin material and assuming that no other interactions that can stabilize the IF structure occur,
the recorded endotherm reflects only the progress of helix-coil transition in the crystalline
sections of the intermediate filaments. The enthalpy measured this way is then related to the
amount of the helical material. Figure 4.5 gives the DSC recorded for hair material for which the
thermal medium (i.e. distilled water) was replaced by a solution of 8% thioglycolic acid and
ammonia (pH 9). This way the hair material is kept in its reduced form during heating and
disulphide bonds are not allowed reforming through oxidation. One may easily notice that by
disrupting the disulphide bonds between IFs and IFAPs (i.e. from nonhelical tail domains) the
endotherm shifts down with approximately 60° C, being identified at a temperature close to those
recorded for soluble proteins.
For each experiment the keratin material introduced into the crucibles was weighted at
ambient room conditions (approx. 22° C and 55% RH). Correcting the denaturation enthalpy
recorded (8.47 J / g) for a ~10 % moisture content of human hair (as determined with Moisture
Sorption Analyser for Caucasian native hair at 22°C, and 55% RH) one obtains ΔH = 9.41 J / g
for dry hair.
Assuming that dry hair has an amount of 21-22% helical crystals,44
and using an average
molecular weight of 114 g / mol for the helical material in keratins3 one estimates the enthalpy of
5.1 kJ / mol for the denaturation of α-helical material in keratin. This value is in good agreement
with the 5 kJ / mol indicated by Privalov for the denaturation enthalpy of one residue in an
isolated α-helix1.
We suppose that the autocatalytic-like character suggested by the kinetic function f(α) is
due to the nascent sulphur compounds from the cystine degradation by α- or β-elimination (see
Chapter 4
82
Figure 4.4). Once parts of the interface are destroyed, the protective role of the matrix disappears
and the secondary structure (α-helix) from the exposed areas of the IFs proceeds to unfold as a
result of high temperature.
Figure 4.5 DSC trace of Caucasian native hair reduced in capsules at a heating rate of 10 K /
min, redrawn from originals
The energy imparted to protein molecules is more than enough to break the relatively weak
forces that hold the protein in its tertiary and secondary configurations and the process advance
very fast. The hydrophobic groups, exposed now to water, tend to cluster together. Ultimately
this becomes the driving force that determines the entire IF structure to collapse and implicitly to
break down the rest of the interface, a process that liberates more compounds that promote the
process to its end. This domino effect continues irreversibly until all of the protein molecules
aggregate into a large insoluble mass in a randomly organised structural framework that contains
also water entrapped molecules.
The activation energy (kinetic barrier), Ea, characterizes the initial rate-limiting step of the
process. As outlined above, its value is hardly to be associated with protein denaturation. On the
other side the calculated value lays down within the range of those recorded for the thermal
decomposition of polysulphurs (105-210 kJ/mol) and polysulphones (155 kJ/mol), through a
homolytic fission of C-S bond in the aliphatic portion of these polymers45
. The dependency of
the Ea on conversion seems as well to be consistent with the proposed model: decomposition
(cleavage of cystine through a homolytic fission of some C-S bonds) followed by the unfolding
(reversible reaction in Figure 4.4) of the helical material and the irreversible denaturation of the
IFs structure.
Nonisothermal kinetics of hard α-keratin thermal denaturation
83
These considerations on activation energy support the view of the damaging of the
scaffold, probable by the scission of the disulphide bonds, as a rate limiting step of the
denaturation.
The system of differential equations associated to the reaction schema of Figure 4.4 is
readily written as:
d(IFo-SS-M)/dt = -k0 · [IFo-SS-M] · [SH]
d[IFoSH]/dt = k0 · [IFo-SS-M] · [SH] + k-1 · [IFunfSH] – k1 · [IFoSH]
d[IFunfSH]/dt = k1 · [IFoSH] – (k2 + k-1) · [IFunfSH]
The system was solved numerically under the constraint of matching the experimental data
and considering that the reaction rates change in time because of temperature changes with a
heating rate of 10 K/min.
Figure 4.6 Simulated (filled diamonds) and DSC obtained (empty circles) conversion degree
change with the temperature advancement when heating the keratin fibres with 10 K/min. From
the simulation we found Ea = 119 kJ/mol and A = 5.6 1013
min-1
for the rate constant of the first
step, while the other rate constants were determined as: k1 = 12 min-1
, k -1= 6 min-1
and k2 = 0.435
min-1
respectively for the temperature of 142°C
Chapter 4
84
As shown in Figure 4.6, the agreement with the recorded DSC curve at 10 K/min is quite
satisfactory. The small differences between the values of Ea and A obtained from the simulated
pathway and those calculated from DSC analysis are within the standard deviations. The lack of
fit in the terminal part of the curves (for conversion degree higher than 0.8 we predict a quicker
end than experimentally observed) is probably due to other mechanisms which were not
considered in our schema from Figure 4.4, like, for example, the reaction of sulphur compound
with other components in matrix and the hydrophobic aggregation of the denaturated proteins
which impede the mobility of the rest of the chains.
Summing up, the model seems to describe satisfactorily the denaturation of proteins for
which unfolding occurs after they free themselves from a scaffolding structure.
4.5. Conclusions
Based on the 3-phase model of hard α-keratins, in which the interface phase scaffolds the
intermediate filaments and controls their thermal stability, we propose a kinetic mechanism for
describing the thermal denaturation pathway of the α-helix. The process occurs only after the
scaffold is damaged irreversibly through a multistep reaction. The non-isothermal investigation
of the kinetics from data acquired at several heating rates on DSC supports this view and
indicates the scission of S-S bonds as the limiting step of the thermal denaturation process. The
theoretical model shows a good agreement with the experimental data and may also describe the
denaturation of other proteins encapsulated in a rigid structure.
4.6. References and notes
1. Privalov, P. L. Adv. Protein Chem. 1982, 35, 1-104.
2. Popescu, C.; Sendelbach, G.; Wortmann, F. J. 10th Int. Wool Text. Res. Conf., Aachen, Germany, 2000.
3. Wortmann, F. J.; Deutz, H. J. Appl. Polym. Sci. 1993, 48, 137-150.
4. Feughelman, M. Text. Res. J. 1959, 29, 223-228.
5. Wortmann, F. J.; Springob, C.; Sendelbach, G. J. Cosmet. Sci. 2002, 53, 219-228.
6. Istrate, D.; Popescu, C.; Wortmann, F. J.; Möller, M. Biomacromol. submitted 2010.
7. Wortmann, F. J.; Popescu, C.; Sendelbach, G. Biopolymers 2006, 83, 630-635.
8. Pielichowski, K.; Czub, P.; Pielichowski, J. Polymer 2000, 41, 4381-4388.
9. Brown, M. E.; Maciejewski, M.; Vyazovkin, S.; Nomen, R.; Sempere, J.; Burnham, A.; Opfermann, J.;
Strey, R.; Anderson, H. L.; Kemmler, A. Thermochim. Acta 2000, 355, 125-143.
10. Sestak, J.; Berggren, G. Thermochim. Acta 1971, 3, 1-12.
11. Popescu, C.; Segal, E. Int. J. Chem. Kinet. 1998, 30, 313-327.
12. Vyazovkin, S. Int. J. Chem. Kinet. 1996, 28, 95-101.
Nonisothermal kinetics of hard α-keratin thermal denaturation
85
13. Fernandez d‘Arlas, B.; Rueda, L.; Stefani, P. M.; de la Caba, K.; Mondragon, I.; Eceiza, A. Thermochim.
Acta 2007, 459, 94-103.
14. Vyazovkin, S.; Sbirrazzuoli, N. Macromol. Rapid Commun. 2006, 27, 1515-1532.
15. Vyazovkin, S.; Wight, C. A. Annu. Rev. Phys. Chem. 1997, 48, 125-149.
16. Suñol, J. J. J. Therm. Anal. Cal. 2003, 72, 25-33.
17. Flynn, J. H.; Wall, L. A. J. Res. Natl. Bur. Stand. 70A 1966, 487–523.
18. Friedman, H. L. Polym. Lett. 1966, 4, 323–328.
19. Kissinger, H. E. Analytical Chemistry 1957, 29, 1702-1706.
20. Braese, S.; Dahmen, S.; Popescu, C.; Schroen, M.; Wortmann, F. J. Polym. Degrad. Stab. 2002, 75, 329-
335.
21. Braese, S.; Dahmen, S.; Popescu, C.; Schroen, M.; Wortmann, F. J. Chem. Eur. J. 2004, 10, 5285-5296.
22. Ruegg, M.; Moor, U.; Blanc, B. J. Dairy Res. 1977, 44, 509-520.
23. Koshiyama, I.; Hamano, M.; Fukushima, D. Food Chem. 1981, 6, 309–322.
24. Bischof, J. C.; He, X. Ann. NY Acad. Sci. 2005, 1066, 1-22.
25. Eyring, H. In The Theory of Rate Processes in Biology and Medicine; Johnson, H. F.; Eyring, H.; Stover,
B. J., Eds.; John Wiley & Sons Inc: New York USA, 1974.
26. Joly, M. A Physico-chemical Approach to the Denaturation of Proteins; Academic Press: London 1965.
27. Cooper, A. In Protein: A Comprehensive Treatise; Allen, G., Ed.; JAI Press Inc: Stamford CT, 1999, p
217–270.
28. Vyazovkin, S.; Vincent, L.; Sbirrazzuoli, N. Macromol. Biosci. 2007, 7, 1181-1186.
29. Vyazovkin, S. V.; Lesnikovich, A. I. Thermochim. Acta 1990, 165, 273-280.
30. Vyazovkin, S.; Linert, W. Int. J. Chem. Kinet. 1995, 27, 73-84.
31. Lepock, J. R. Int. J. Hyperthermia 2003, 19, 252-266.
32. Lyubarev, A. E.; Kurganov, B. I. J. Therm. Anal. Cal. 2000, 62, 51-62.
33. Cravalho, E. G.; Toner, M.; Gaylor, D. C.; Lee, R. C. In Electrical Trauma: The Pathophysiology,
Manifestations and Clinical Management; Lee, R. C.; Cravalho, E. G.; Burke, J. F., Eds.; Cambridge University
Press: Cambridge, 1992, p 281-300.
34. Sanchez-Ruiz, J. M. Biophys. J. 1992, 61, 921-935.
35. Lumry, R.; Eyring, H. J. Phys. Chem. 1954, 58, 110-120.
36. Hearle, J. W. S. Wool Tech. Sheep Breed 2003, 51, 95-117.
37. Powell, B. C.; Rogers, G. E. In Formation and Structure of Human Hair; Jolles, P.; Zahn, H.; Höcker, H.,
Eds.; Birkhäuser Verlag: Basel Switzerland, 1997, p 59-148.
38. Robbins, C. R. Chemical and Physical Behavior of Human Hair; Springer-Verlag: New York, USA, 2002.
39. Zahn, H. Melliand Textilber 1991, 72, 926-931.
40. Parry, D. A. Adv. Protein. Chem. 2005, 70, 113-142.
41. Hearle, J. W. S. J. Mater. Sci. 2007, 42, 8010-8019.
42. Istrate, D.; Popescu, C.; Er Rafik, M.; Möller, M. Polym. Degrad. Stabil. submitted 2010.
43. Florence, T. M. Biochem. J. 1980, 189, 507-520.
44. Cao, J.; Leroy, F. Biopolymers 2005, 77, 38-43.
45. Porter, M.; Oae, S., Ed.; Plenum Press: New York, 1977.
* Journal of the Society of Cosmetic Chemists, submitted 2009
Chapter V : Differential scanning calorimetry (DSC)
analysis of structural changes in bleached, perm-
waved and dyed hard alpha-keratin fibres*
5.1. Introduction
Proteins are macromolecules (polypeptides) with a complex structure that is important to
their function. This structure may be regarded at several levels, namely primary, secondary,
tertiary, and quaternary one 1. The primary structure is associated with the covalent bonds within
the protein molecule; the secondary structure involves the hydrogen bonds (although some
disulfide bonding can also occur), thereby creating the alpha helix or beta sheet structures; the
3D folded structure of a whole globular protein is called the tertiary structure and it is important
to protein function, whereas the quaternary structure usually involves the conformational fitting
of two proteins together associated with specific function 2. When this structure is changed or
altered as a result of thermal, chemical or mechanical modification of the protein environment,
the protein is unable to carry out its specific function. Such a process is identified as denaturation
2. While the heat-induced denaturation of globular proteins is reasonably well understood, those
of structural proteins (e.g. collagen, keratins) remains an area of active research.
The hard alpha-keratin is a filamentous protein found in mammalian epidermal appendages
(hairs, quills, horn, nails, etc.) distinct from beta/feather keratin (beta sheet-based) found in avian
and reptilian tissues. Keratin-containing tissues were first studied for the economic importance of
animal fibres in the textile industry (wool), along with cosmetic related aspects such as hair
growth and epidermis substitutes. Like other filamentous family members, hard alpha-keratin
fibres act mainly as a mechanical support and are the topic of many investigations 3-7
.
The structure of hard alpha-keratin is characterized by three structural hierarchy levels 8.
At high resolution, the intermediate filament (IF) protein is made of a central rod domain of
sequences (lA, lB, 2A, 2B) containing a heptad repeat, and separated by loop links (L1, L12 and
L2) 4,6
. At the extremity of the rod domain are located the globular C- and N-terminal domains
arranged mostly in ß-sheet and formed of rich of sulphur compounds 8,9
. Two α-helices form a
parallel, super helical dimmer with the apolar residues buried to the inside of the coiled-coil
Chapter 5
88
structure as a consequence of hydrophobic effects 10
. At medium resolution, i.e. the intermediate
level arrangement of the heterodimers inside IFs, the molecules are assembled both
longitudinally and laterally in an ensemble called microfibril 11
. Radially, the number of
molecules, across a keratin IF section, is assumed to be 26–34 12
. The dimers are associated as
straight tetramers with a random orientation 8 and this organisation forms a long cylinder-shaped
intermediate filament 11
with uniform density. The terminal domains play a crucial role in
directing molecular and filament aggregation 4. At low organisation, the bundles of parallel
intermediate filaments are organised in distorted crystalline lateral network and embedded in a
sulphur-rich protein matrix of intermediate filament associated proteins (IFAPs) and form a
macrofibril, the main morphological components of hard alpha-keratin fibres 8.
The properties and interactions of the main morphological components of keratin fibres
(IFs and IFAPs) are still under academic debates for understanding how these are specifically
related to the various aspects of fibre stability and properties. The head and tail domains in
keratin molecules generally contain a multitude of sites that allow keratin IFs to form covalent
bonds with other proteins. Characteristically, the end domains of the keratin fibre IF lack the
extended runs of glycine residues found in epidermal IF and contain many cysteine residues.
This enables them to participate in extensive disulphide bond crosslinking with the abundant
cysteine-rich proteins of the fibre 5,13,14
the non helical terminal domains of IF chains projecting
into the interfilamentous space and linking with the matrix proteins 7. Lysine to glutamine
crosslinks have been also found between the head and the tail domains in all keratin chains 9.
Besides, the matrix that fills the spaces between filaments is made of the small keratin proteins
of the cysteine-rich and glycine / tyrosine-rich protein families. The potential interactions of so
many proteins could attain a bewildering complexity 5.
It is generally known that the physical and mechanical properties of keratin fibres are
changed by chemical cosmetic treatments such as bleaching, permanent waving or dyeing 15-20
.
These modifications are regarded as a result of cleavage of disulphide bonds in the constituent
protein of keratin and alteration of the secondary structure, which is unfolding of the alpha-
helical material (crystalline phase).
In a previous investigation we reported that low pH affects the thermal denaturation in hair
protein 21
, and highlighted the importance of the interface between crystalline and matrix phases,
made of nonhelical tail domains of keratin that scaffolds the intermediate filaments and controls
their interaction with chemical reagents as well as their thermal properties. This allowed us
proposing a multistep mechanism for the thermal denaturation of hard α-keratins in water excess
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
89
that relies on the 3-phase model which describes their structure 22
. The total heat uptake (i.e. the
recorded enthalpy) required for breaking down the scaffold in order to allow the denaturation of
helical material should decrease if the interface is previously damaged. We suggested that the
changes recorded in the variation of the experimental DSC parameters (i.e. peak temperature, Tp,
and enthalpy, ΔH) are more likely to occur as a consequence of modifying the immediate
environment of the intermediate filaments (interface phase) rather than due to a significant loss
of the secondary structure of keratin protein. Our approach is consistent with literature data
indicating that a decrease of denaturation enthalpy or of other parameters indicating extensive
damage to the IFs is not necessarily accompanied by a loss of X-Ray crystallinity 23-26
.
The present investigation provides more support for this hypothesis and points out the need
for a careful interpretation of DSC data in respect with the cosmetic formulations that are
designed to change morphological components within the hair cortex.
5.2. Materials and methods
The alpha-keratin fibres used for analysis were of Caucasian dark-brown hair, supplied by
KERLING International Haarfabrik GmbH. The fibres were cleaned with 1% Lauryl ether
sulphate (LES) and dried at room temperature prior to work with them. The pH of their aqueous
extract was found to be 6.5 to 7.
DSC measurements
There are two ways of measuring the DSC of keratins, namely in water and in dry
environment (allowing the moisture to evaporate during heating). The DSC in dry environment
was showed to supply misleading information, due to the interference of pyrolysis with the
process of interest 27
. Consequently the present work deals exclusively with DSC of keratins in
water excess.
Prior to the measurements the samples were cut into fine snippets (~2mm) and stored
under controlled conditions (~ 24 hours, 22°C, 55% relative humidity) to ensure invariant water
contents. 7…10mg of each sample snippets were weighted and placed in crucibles.
Prior to sealing a crucible, 50 μL of distilled water (pH 6.7) was added, and the sealed
crucible was stored over night (~14 hours preceding the measurement), to allow the fibres to wet.
The DSC experiments were run in a DSC-7 Perkin Elmer, using pressure resistant stainless
steel large volume capsules. DSC calibration was done with indium and palmitic acid, both of
high purity. The temperature ranged from 50 to 180°C at a heating rate of 10 K/min. For each
sample we performed 3…5 measurements and the peak temperature, Tp, and enthalpy, H, of
Chapter 5
90
the endothermal effect were reported as mean values and standard deviations.
Tensile measurements
The measurements were performed in wet conditions, considered to reflect best the
changes at the level of intermediate filaments 20,25,28
.
The tensile measurements were performed on the Miniature Tensile Tester Model 675
(MTT675) and the Fibre Dimensional Analysis Unit Model 765 (FDAS765) of Dia-Stron, UK. A
minimum of 35 single fibres were tested for each sample at a stretching rate of 20 mm/min and a
gauge force of 1 gf, as initial condition.
Prior to loading in the cassettes of the carousel, the samples were immersed in distilled
water for 120 minutes to allow them wetting. During the measurements the cassettes were also
filled with distilled water to ensure the 100% humidity content during the measurement.
The stress-strain curve recorded for each fibre allows calculating the Young‘s modulus, the
yield strength and the breaking extension and total work, which characterise numerically the
fibre mechanic.
Amino acids analyses were conducted conventionally on ―Alpha Plus‖ Amino acid
Analyser, manufactured by Pharmacia LKB, Freiburg, Germany. The results are expressed in
molar percentage.
X-ray microdiffraction experiments were performed at the European Synchrotron
Radiation Facility (Grenoble, France) on microfocus Beamline ID13 29
. A high intensity
monochromatic beam (wavelength = 0.961 Å), coming from an undulator and a Si-111 double
crystal monochromator, was focussed with an ellipsoidal mirror (focal spot 20(h) * 40(v) m2)
and then size-limited down to a 5 m diameter circular section by a collimator placed in the focal
plane. A guard aperture (Pt–Ir, 10 m diameter) reduced diffuse scattering from the collimator
exit.
Samples were made of 10 hair fibres from each sample mounted on a frame with the hair
axis perpendicular to the X-ray beam on a computer-controlled gantry coupled with a
microscope which permitted sample positioning with a 0.1 µm resolution.
The experiments were carried out with a 320 mm sample–detector distance, which was
calibrated using silver behenate, the first order spacing of which is 58.38 Å. Using a small beam
stop of 300 µm diameter, two-dimensional X-ray scattering patterns were collected from 0.006
to 0.4 Å. Patterns were recorded with 1 s exposure times on a MAR-CCD camera (16 bit
readout; 130 mm entrance window; 2048 · 2048 pixels; pixel size of 78.94 · 78.94 µm2).
Radiation damage on the structure has been verified to only occur after exposure times longer
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
91
than 10 s and it is indicated by the strong weakening and broadening of the scattering features.
About five patterns were collected along each hair fibre.
Data analysis has been focussed on the scattering regions that provide information at the
various structural organization levels, viz.:
(iv) the meridian arc located in the 5Å region, which is produced by the regular α-helical
coiled-coil packing. The strong intensity of this arc was shown to be related to the fine
configuration of residues 30-32
;
(v) the fine meridian scattering arc at 67Å. This is indicative for the periodic architecture
of the molecules along the IF;
(vi) the equatorial small-angle X-ray scattering zone, which is related to the radial
geometry of the filaments and to their lateral packing organization in the matrix. In particular,
the distorted crystalline lateral organization gives rise to a strong equatorial reflection observed
around 90Å 33-35
.
The analysis was carried out following two complementary procedures. The position and
intensity of the main scattering features were first estimated and compared from a visual
inspection of all patterns. For the most representative patterns, one-dimensional equatorial
profiles passing through the origin, both along the equator and the meridian, were extracted from
the two-dimensional patterns integrating the intensity around the equator on a 10 pixels thick
rectangular strip. These profiles yielded the precise positions and intensities of the main
scattering features and made the comparison between patterns and modelled profiles more
straightforward. In the SAXS zone the huge scattering intensity was shown to proceed from
nonkeratinous zones in hair 33
. This component has been subtracted from the profile.
Damaging treatment
Bleaching, perm-waving, dyeing and pH treatments were used for achieving controlled
modification of the fibres.
Bleaching treatment was done with IGORA VARIO BLOND PLUS bleaching powder and
IGORA ROYAL 20 vol 6% H2O2 bleaching lotion, a commercial products kindly supplied by
Schwarzkopf.
The bleaching procedure followed the instructions of use, being applied for 35 minutes at
room temperature. A bleaching cycle implies the treatment of 1g hair sample with a mixture (pH
10) made-up of 0.6 g bleaching powder and 1.2 ml of bleaching lotion containing 6% H2O2.
Afterwards, the fibres were rinsed under tap water for 3 minutes, 2 times subsequently washed
with a solution of Texapon N70, 0.1 mL/L, (70% Natrium Laurethsulfat, pH~7), warm water 1
Chapter 5
92
minute, tap water 3 minutes and dried under hot air blow. The pH of the aqueous extract of the
fibre was checked to be 7. The process has been resumed up to seven times (at intervals of 24
hours) on the same hair sample, fibres for analysis being sampled after each complete bleaching
cycle.
Permanent waving treatment was performed with the commercial product POLY LOCK-
PERMANENTE FORTE kindly supplied by Schwarzkopf.
A perm-waving cycle consists of immersion of wetted hair tresses in the reduction solution
(pH 8.5-9; liquor ratio of 1.2 g hair to 1 mL solution). The tresses are then covered with plastic
folia and let to react for 30 minutes at room temperature. After rinsing 3 minutes with tap water
the tresses are immersed in the oxidation lotion (pH 4.5) using similar conditions as for the
reductive process. Processing time is of 10 minutes at room temperature, according to the
product recommendation. The fibres were then rinsed with tap water for 3 minutes, 2 times
subsequently washed with a solution of Texapon N70, 0.1 mL/L, (70% Natrium Laurethsulfat,
pH~7), rinsed with warm water 1 minute, tap water 3 minutes and dried under hot air blow. The
pH of the aqueous extract was 7.
The process has been repeated up to 7 times on the same sample, hair fibres for analysis
being sampled after each complete perm-waving cycle.
Dyeing treatment was performed with LIVE colour cream (4% 1:1 molar mixture of p-
toluene diamine and resorcin) and VISION developer lotion (6% H2O2), commercial products
kindly supplied by Henkel.
The colour cream and the developer were mixed 1:1 shortly before application (pH 6.8), 3
grams of the coloration mixture being applied on 1 gram of dried hair. After elapsing of 30
minutes (the requested processing time) at room temperature the fibres were rinsed under tap
water for 3 minutes, 2 times subsequently washed with a solution of Texapon N70, 0.1 mL/L,
(70% Natrium Laurethsulfat, pH~7), warm water 1 minute, tap water 3 minutes and dried under
hot air blow. The pH of the aqueous extract was checked to be also 7.
The process has been repeated up to 7 times on the same hair sample, hair for analysis
being sampled after each complete dyeing cycle.
pH treatment: At a liquor ratio of 1 gram fibres to 200 mL solution, the hair tresses were
immersed in aqueous solutions with different values of pH, for 30 minutes at room temperature.
The fibres were then rinsed with tap water for 3 minutes, 2 times subsequently washed with a
solution of Texapon N70, 0.1 mL/L, (70% Natrium Laurethsulfat,), rinsed with warm water 1
minute, tap water 3 minutes, and immersed over night in distilled water to completely neutralise.
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
93
The fibres were eventually dried under hot air blow.
We used acetic acid for adjusting pH from 1 to 7.
5.3. Results and discussions
Differential scanning calorimetry (DSC) is widely used as a tool for protein studies 36-39
.
The DSC records an endothermal process whose peak temperature (Tp) and enthalpy (ΔH) are
used for characterising the denaturation of protein. The peak temperature measures the thermal
stability of proteins. Its value is influenced by the heating rate 40
and protein concentration 41
, as
well as by the biochemical environment (mainly pH) 2,42,43
. The ΔH value is the heat uptake
during the unfolding transition 44
. Assuming a two-state transition model for proteins
denaturation, the heat uptake is correlated with the content of ordered secondary structure of a
protein 41,44
.
Little systematic work has been done on the thermodynamics of fibrous proteins,
particularly keratins, because of being poorly soluble and difficult to purify in sufficient
quantities for biophysical studies. The cosmetic treatments such as bleaching, permanent waving
and the use of the permanent colorants were shown to cause changes of the hair fibre structure
noticed by consumers as an increase of hair breakage, a reduced shine, etc. During the past years
thermal and / or mechanical properties, as well as X-ray and amino-acids analysis were
employed for understanding the effects of cosmetic processes on the major morphological
components of human hair. The analysis of the recorded endotherms observed when exposing
keratins to a controlled heating relies on models used for describing the behaviour of globular
proteins.
Assuming the validity of the two-phase Feughelman model 45
and a 2-state transition
model for the denaturation transition as in case of globular-proteins, Wortmann and Deutz
suggested that the endothermal process recorded on the DSC curve of keratins reflects the
progress of the helix-coil transition in the crystalline sections of the intermediate filaments 46
.
Feughelman model 45
describes keratin fibres as composed of long, water-impenetrable
relatively rigid cylindrical rods set parallel to the fibre axis and embedded in a water permeable
matrix. According to the model, the α-helices, aggregated in the rods (IFs), form a crystalline,
continuous, axially oriented, elastic filament phase, which is embedded in an amorphous matrix
phase that comprises the non helical proteins (IFAPs) and all other noncrystalline, viscoelastic
components (intermacrofibrillar cement, nuclear remnants, cell membrane complex, cuticle, etc).
The absorption of water by the matrix mechanically softens this phase, whereas the rods
Chapter 5
94
are water impenetrable, thus mechanically unaffected by the presence of water. The viscosity of
matrix phase is believed to control kinetically the thermal stability of the α-helical part.
Consequently, it was assumed that the DSC recorded enthalpy, ΔH, reflects the amount
and structural integrity of the α-helical material in the intermediate filaments (IFs), and the peak
temperature, Tp, reflects the cross-link density of the matrix (IFAPs) in which the IFs are
embedded 19
. Therefore it is expected that if any of the main morphological components (i.e. IFs
or IFAPs) of the hair material is affected by a cosmetic formulation a variation on the DSC
endotherms must be recorded.
Several authors have investigated the effect of multiple bleaching treatments on the heat
denaturation of hair, using the DSC in water excess method and following the variation of
enthalpies and peak positions on the temperature axis. Wortmann et al. showed a steady decrease
in both ΔH and Tp with an increasing of number of treatment cycles 19
. Based on the two-phase
model the effect of a multiple bleaching treatment is understood as the decrease of matrix cross-
link density due to the loss of disulphide crosslinks (Tp variation) and respectively as the
decrease of the amount of native, α-helical material (ΔH variation) that is chemically denatured
17,47.
The alterations produced in hair fibres following oxidative dyeing processes are expected
to be similar to those produced by bleaching, although on a smaller scale. Gel electrophoresis
studies of the dyeing influence on the pattern of human hair proteins revealed that although the
IFs are not significantly modified, the matrix proteins are strongly affected after such a treatment
48.
The perm waving treatment influences also the DSC endotherm. Within the framework of
the two-phase model both phases (intermediate filaments, IFs, and matrix, IFAPs) are reported to
be affected by the treatment, still at a lower extent than the bleaching does 17
. It was also noted
that the perm-waving damage affects more the helical segments of the intermediate filaments
than the surrounding high sulphur cross-linked matrix 19
. The decrease of the peak area with the
perm-waving, reported also by others 49
is in agreement with 13
C CP/MAS NMR studies showing
decrease of helical material in Asian hair after perm-waving 50
.
Earlier investigations showed that despite of recording a significant reducing of the value
of enthalpy for weathered wool, no significant changes of X-ray pattern were observed 25,51
. For
interpreting the conflicting evidences it was proposed that the decrease of enthalpy may not
reflect a genuine destruction of the α-helix, but a decrease of the amount of the ―native‖ helical
fraction due to radiation-induced crosslinking of the material, and this fraction becomes
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
95
undenaturable within the experimental range25
.
Figures 5.1-5.3 detail the effects of the investigated cosmetic treatments on the
denaturation of hairs heated in water excess, respectively the change of both enthalpy (ΔH) and
peak temperature (Tp) vs. the treatment of hair.
5
7
9
11
13
15
17
110
115
120
125
130
135
140
145
150
155
Native 1x 2x 3x 4x 5x 6x 7x
En
thal
py
(J/
g)
Tem
per
ature
(°C
)
Number of bleachings
Tp ΔH
Figure 5.1 Denaturation temperatures Tp (left y-axis) and enthalpies ΔH (right y-axis) recorded
after a multi step bleaching treatment for Caucasian hair. The symbols give the mean values. The
standard deviations are included in the size of the symbol, in the most cases being very small.
Native refers to untreated material
5
7
9
11
13
15
17
110
115
120
125
130
135
140
145
150
155
Native 1x 2x 3x 4x 5x 6x 7x
En
thal
py
(J/
g)
Tem
per
atu
re (
°C)
Number of perm-wavings
Tp ΔH
Figure 5.2 Denaturation temperatures Tp (left y-axis) and enthalpies ΔH (right y-axis) of the
perm-waved Caucasian hair samples, characterized by the arithmetic means (symbol) and
standard deviations. Standard deviations are included in the size of the symbol in most of the
cases, being very small. Native refers to untreated material
Chapter 5
96
5
7
9
11
13
15
17
110
115
120
125
130
135
140
145
150
155
Native 1x 2x 3x 4x 5x 6x 7x
En
thal
py
(J/
g)
Tem
per
atu
re (
°C)
Number of dyeings
Tp ΔH
Figure 5.3 Denaturation temperatures Tp (left y-axis) and enthalpies ΔH (right y-axis) recorded
after a dyeing treatment of Caucasian hair samples. The symbols give the mean values. The
standard deviations are included in the size of the symbol, in the most cases being very small.
Native refers to untreated material
Figure 5.1 summarises the arithmetic means and standard deviations of Tp and ΔH for
bleached samples. The values are in good agreement with other published data 17,19
. The DSC
results are fairly accurate, having values of standard deviations for Tp of less than 1°C and for
denaturation enthalpy of less than 0.8 J/g. One may observe that the first 3-4 bleaching cycles
lead to a roughly linear decrease for both parameters. Beyond this, both Tp and the enthalpy
level off at a temperature that is ~25 °C lower than for the untreated material and at an enthalpy
of ~ 10 J/g that is almost 35% smaller than the initial value (14.7 J/g). Out of these data the two-
phase model suggests that bleaching leads to largely homogeneous damage of IFs and IFAPs 19
.
Figure 5.2 summarises the effect of multiple cycles of perm-waving on the thermal
denaturation of Caucasian hair samples. One notes a consistent decrease of Tp as well as of ΔH
with increasing the number of treatments. The decrease of enthalpy value occurs much faster
than that of the peak temperature. After the last perm-waving cycle a decrease of only ~12.5°C
was recorded for Tp while the enthalpy decreases by 45-50% from 14.7 to about 7.5 J/g. As in
the case of bleaching, the variation of DSC parameters was found to be in good agreement with
literature reports mentioned previously.
Raman Spectroscopy studies of Kuzuhara 52
on Chinese human hair damaged by a
permanent-waving processes revealed that the reduction step do not affect the secondary
structure (i.e. alpha-helical material) but elute random coil structures of some of the proteins
existing throughout the cortex, other than those related to the matrix. A slight decrease of the
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
97
helical material content and an increased disorder degree of the microfibrils-matrix packing
occurs apparently as a result of the oxidation step from a perm-waving treatment.
Other Raman spectroscopy studies 53,54
indicate a high damage induced to the helical
components by bleaching treatments.
The dyed fibers do not exhibit differences from untreated material on DSC curve (Figure
5.3) although optical microscopy reveals the formation of the pigment inside the hair cortex,
eliminating the assumption of a superficial dyeing (Figure 5.4).
Figure 5.4 Cross-sections of Caucasian blond hair showing formation of the pigment inside the
hair cortex, the intensity of the colour rising with increasing the number of treatments
In our previous reports we noticed a strong dependence of keratin fibre thermal stability on
the pH pre-treatment history, as recorded by wet DSC experiments 21,22
. The amino-acid
analysis, tensile measurements and X-ray analysis which were additionally employed did not
show significant changes of chemistry or crystallinity of the keratin to account for the shifts of
peak temperature or enthalpy of the endothermal process. This allowed us proposing a three-
phase model for describing the structure of hard alpha-keratins in which the nonhelical (or
globular) terminal domains promote filament interactions. The interface phase scaffolds the
intermediate filaments and controls their thermal stability. The thermal denaturation process of
the intermediate filaments can occur only after the scaffold is damaged irreversibly. The total
heat uptake (enthalpy, ΔH) required for breaking down the scaffold in order to allow the
denaturation of helical material is the sum of those of all interactions participating to the
stabilisation of the IF structure and of the unfolding transition of the helical material. The S-S
bonds of the interface bring also their contribution to the denaturation process and any damage of
the scaffold required for inducing the denaturation process needs their breaking. This model is
Chapter 5
98
the background for the kinetic mechanism which we proposed for describing the thermal
denaturation pathway of the α-helix22
. The process occurs only after the scaffold is damaged
irreversibly through a multistep reaction. The non-isothermal kinetic calculations based on data
acquired at several heating rates with DSC supports this view and indicates the scission of S-S
bonds as the limiting step of the thermal denaturation process22
.
We predicted therefore that the changes recorded for the experimental DSC parameters are
more likely to occur as a consequence of modifying the immediate environment of the
intermediate filaments (interface phase) rather than of changing the amount of crystalline
material in the IFs. In other words, the structural changes induced by chemical treatments refers
primarily to the higher levels of organization of the hard-alpha keratins proteins rather than to
alterations of the secondary structure that is the unfolding of the alpha-helical material
(crystalline phase). Extending this hypothesis to the cosmetic treatments investigated herein, we
may expect to observe a combination of two effects reflected by the denaturation peak
temperature and enthalpy, viz.: a change due to irreversible scission of the S-S bonds and a
reversible effect due to self- or chemical induced interactions like those generated by pH
variation 21
.
Young Modulus
GPa
Yield Strength
MPa
Break extension
% Strain
Total work
mJ
Native 2 ± 0.2 45.1 ± 2.9 56.5 ± 3.3 6.8 ± 2.2
Bleaching 3x 1.4 ± 0.2 29.4 ± 2.8 64.1 ± 3.9 5.5 ± 1.7
7x 1 ± 0.2 21.1 ± 2.9 65.2 ± 8.5 3.9 ± 1.5
Perm-waving 3x 1.4 ± 0.1 30.9 ± 2.6 65.7 ± 3.3 6.6 ± 2.3
7x 1 ± 0.1 21.7 ± 2.6 63.9 ± 4.9 4.9 ± 1.4
Dyeing
3x 2 ± 0.3 44.5 ± 2.5 54.7 ± 3.7 5.7 ± 1.8
7x 1.9 ± 0.1 42.6 ± 2.4 55.9 ± 2.5 5.9 ± 1.3
Table 5.1 The influence of the investigated treatments on the wet tensile strength properties (±
standard deviation) for Caucasian hairs, sampled after 3 respectively 7 cycles from a resuming
treatment. Native refers to untreated hair material
The data of the tensile strength and the amino-acids composition summarised in Tables 5.1
and 5.2, respectively, confirm that the drop of DSC parameters is a consequence of chemical
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
99
alteration of the hair structure.
Aminoacid Native Bleaching Perm-waving Dyeing
St.dev 3x 7x 3x 7x 3x 7x
Cysteic acid 0.74 0.16 5.15 8.06 3.48 5.55 0.83 1.33
Aspartic acid 5.83 0.25 6.09 5.40 6.30 6.21 5.56 5.69
Threonine 8.47 0.40 7.53 7.36 8.90 8.75 8.22 8.28
Serine 13.63 1.45 10.34 12.06 13.25 13.54 11.40 11.72
Glutamic acid 14.86 0.84 10.53 12.92 14.12 13.44 12.27 12.75
Proline 7.94 0.59 11.43 8.91 7.26 7.56 7.91 7.93
Glycine 5.27 0.66 7.40 5.26 6.45 5.80 7.75 7.25
Alanine 4.52 0.32 6.31 4.96 5.67 4.66 5.33 5.05
Valine 4.74 0.70 6.54 5.77 5.51 4.78 6.19 6.12
Cystine 8.03 0.53 6.51 6.22 7.06 5.30 8.75 8.71
Methionine 0.31 0.12 0.39 0.30 0.41 0.83 0.72 0.62
Isoleucine 2.50 0.44 1.85 2.29 1.79 3.42 3.19 3.43
Leucine 6.80 0.33 5.69 6.68 6.30 7.22 7.42 7.39
Tyrosine 2.34 0.18 1.32 2.44 1.57 2.48 1.73 1.63
Phenylalanine 2.27 0.30 1.59 2.37 2.09 2.31 2.08 2.28
Ornithine 0.49 0.26 0.36 0.00 0.38 0.23 0.44 0.21
Lysine 3.23 0.41 2.32 2.81 2.52 2.08 2.53 2.77
Lanthionine 0.15
Histidine 1.92 0.53 1.00 1.20 0.94 0.65 0.73 1.00
Arginine 6.10 0.89 7.66 4.99 5.99 5.05 6.97 5.85
Table 5.2 The influence of the investigated treatments on the amino-acids composition (mol %)
of Caucasian hairs, sampled after 3 respectively 7 cycles from a resuming treatment. Native
refers to untreated hair material while the experimental error (st.dev) was determined from 3
measurements on the same sample
The regression analysis performed on the amino-acid composition and DSC recorded peak
suggests that, for hair fibres treated differently, the value of enthalpy may be linearly related to
the cystine content (Figure 5.5). Our data yields a value of 1 J / 30.65 µmol cystine for the
decrease of enthalpy with cystine content.
Chapter 5
100
It has to be underlined that these results are acquired for the same material treated at
various intensities. As a consequence, the linear relationship cannot be extended for comparing
keratin fibres of various origins, for which case it was shown that such correlation does not
hold46
.
y = 30.65x + 105.22
R² = 0.860
100
200
300
400
500
600
700
7 8 9 10 11 12 13 14 15 16
Cy
stin
e (µ
mo
l/g)
Enthalpy (J/g)
7x
3x 7x
7x
3x
3x
Figure 5.5 Regression analysis for the enthalpies vs. the cystine content; Symbols: Δ -
permanent waving treatment; □ - bleaching treatment; ○ - dyeing treatment; ◊ - untreated
material; ■ - bleached 3x material followed by a pH 3 treatment 21
The relationship found for cosmetically treated hair supports the view that the enthalpy
gives account of the nonhelical tails of the IF proteins and other matrix materials which contain
cystine, rather than of the amount of crystalline material. Consequently, the strong correlation
noticed for the variation of enthalpy with the cystine content may be attributed to the interaction
between IFs and matrix composite, i.e. to the nonhelical parts of the IFs that extend into the
matrix and link with the matrix proteins.
The X-ray diffraction patterns of hard alpha-keratins is of central importance in structural
studies of this material because the agreement between calculated and observed intensities
provides a searching test of the correctness of any proposed model 55
. Information about the
amorphous region (the cuticle and the matrix) of keratin fibres cannot be obtained from X-ray
diffraction, since it only reflects the state of the highly crystalline structure in the cortex material.
Along the meridian axis (fibre axis), the dimmers characterized by a regular alpha-helical
coiled-coil folding in the rod central domain give rise to the wide-angle X-ray scattering
(WAXS) meridian arc located in the 5.15 Å region. The strong intensity of this arc was shown to
be related to the fine configuration of residues 31,32
. At the small-angle X-ray scattering (SAXS)
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
101
region, the strong and fine 67 Å meridian scattering arc is related to an axial stagger between
molecules or group of molecules along the microfibril 56-58
, its position being almost insensitive
to humidity variations 57,59
. Along the equatorial axis, the X-ray pattern gives poor information
about the intermediate scale arrangement of the chains inside IFs. At WAXS equatorial region,
the broad scattering maximum located at 9.5 Å peak is supposed to be due to interferences
between coiled coil chains 34,35
or chains distance from others structures 60
. In the SAXS
equatorial region three broad peaks corresponding to the distances 90 Å, 45 Å, and 28 Å
(respectively located at S = 0.012, 0.022, and 0.036 Å-1
) are provided from the dense lateral IF
packing. Hair contains crystallized lipids 61
, more precisely soaps 62
, that give rise to a series of
rings, of which the first order is superimposed on the peak at 45 Å. The signals due to soaps are
the only variable scattering signals displayed by hair; the signals due to keratin are fairly sample-
independent. The pioneering X-ray scattering analyses of Fraser have established that the IFs are
located at the nodes of a distorted two-dimensional quasi-crystalline array 34,35
. This model was
later refined using an analytical description of the corresponding small-angle X-ray scattering
(SAXS) equatorial X-ray scattering pattern 33
. So, the dense lateral packing of the microfibrils
embedded in the matrix namely the microfibril-matrix network can be investigated in this region.
The position and the intensities of these peaks are characteristic of microfibril diameter and of
the mean of the centre-to-centre distance between microfibrils 57
; when the hair fibre is
immersed in water, the 90 Å peak position increase indicating a matrix swelling 57,63
.
For bleached 3 times samples 24 exploitable patterns were obtained out of 50 shots for
both meridian and equatorial profiles. From these, only one pattern indicated a significant
decrease of the 5.15 Å reflections so we may assume that the coiled coil configurations remain
virtually unaffected. The calculation of inter-microfibril distances gives 93.121 Å for native hair
and a range between 89.16-91.64 Å for bleached hair, while the diameter of the microfibril was
found to be 37.5 Å in all cases.
For bleached 7 times samples 35 patterns were found to be exploitable for the meridian
profile and 28 for the equatorial from a total of 50 shots. Even after such a harsh treatment, 82 %
of the exploitable patterns reveal no significant disturbance of the coiled-coil structure for the
meridional profile, while a number of 6 patterns only was found to indicate an important
decrease of the 5.15 Å reflection. 33 % of the exploitable equatorial patterns show the distance
of 93.121 Å between microfibrils, with a microfibril diameter of 38.5 Å that is almost identical
with the value recorded for the untreated material (93.121 Å for the inter-microfibril distance,
respectively 37.5-38 Å for the IF diameter). However, proving a heterogeneity of the chemically
Chapter 5
102
induced effects along the fibre, 35 % of the patterns reveal an increased disorder of microfibrils
(inter-microfibril distance found to be 97.033 Å) together with an increase of the IF diameter (39
Å). Another 32 % of the patterns present a inter-microfibril distance of 91.462 Å while the
microfibril radius was impossible to be determined because of the lipids signal interference.
Permanent-waved 7 times samples were also subjected to X-Ray analysis. 49 patterns were
found to be exploitable from 50 shots. From these, 80 % shows no significant influence of the
treatment on the coiled-coil helical structure, only 10 patterns indicating a decrease or no
presence of the 5.15 Å reflections. The overall distance between microfibrils was found to be
90.21 Å while the radius of microfibril increased to 39 Å.
Summing up, the X-Ray diffraction analysis confirms our aforementioned hypothesis
regarding the protective role of the IFs immediate environment against moderate mild action
treatments.
With respect to Kuzuhara´s observations that refer to the elution of some random coil
structures of proteins existing throughout the cortex region other than those related to the matrix
(i.e. non-helical terminal domains), we assume the involvement of distinct intermediates in the
denaturation process of hard alpha-keratins as result of severe chemical treatments to account for
the variations recorded in the inter-microfibrilar distances and IFs radius. These intermediates
may be similar to molten or premolten globules observed in the transient intermediate states
found during the folding of certain proteins, especially globular proteins that undergo
hydrophobic collapse. The molten globules are described as compact intermediate protein
conformations that generally preserve the native-like secondary structure but have a poorly
defined and dynamic tertiary structure 64-66
. This approach is consistent with the kinetic
mechanism proposed for describing the thermal denaturation pathway of hard alpha-keratin 22
.
The variations recorded after a bleaching treatment refers primarily to the weakening of
IFs-IFAP interface and to the damage of the matrix components, due to disulphide bond scission.
This is confirmed by X-ray diffraction that does not indicate the damage of the crystalline
structure after mild action time treatments, demonstrating the high stability of the helical
material. One may easily notice from Table 5.1 and Table 5.2 that while the tensile strength
seems to give good estimations of the oxidative covalent bond cleavage, the DSC data do not
follow the intensity of the treatment, both peak temperature and the enthalpy reaching a plateau
after the 3rd cycle of bleaching (Figure 5.1). The X-ray diffraction seems to confirm that despite
some heterogeneity, subsequent damage reflects mainly in a slight loss of the tertiary structure,
and change of inter-microfibrilar distance and of the IFs radius. The additional enthalpic
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
103
contribution that ―hide‖ the real damage after the 3-4 bleaching steps may come, as it happens in
case of collagen 44,67
, from regular solvation effects, possibly involving extended hydrogen-
bonded chains of water molecules acting as a sort of ―aqueous scaffolding‖ at the surface of the
now exposed IFs. More likely, this is the result of ionic interactions due to the formation of high
amounts of cysteic acid and incorporation of residual formulation components that compete with
the remaining covalent disulphide bonds to the stabilisation of the interface phase.
A low pH (Figure 5.6) induces the protonation of the keratin proteins. This strengthen the
scaffold, which may explain the higher values of peak temperature and enthalpy recorded by
DSC for hair after 3 bleaching steps and treated at acidic pH compared to the same hair treated
with neutral pH.
The data suggest that keratin IFs can modulate their organisation and thermal properties
through chemically induced interactions.
100
110
120
130
140
150
160
170
O pH1 pH2 pH3 pH5 pH7
Tem
pera
ture
(°C
)
Native 3x 7x
0
5
10
15
20
O pH1 pH2 pH3 pH5 pH7
Enth
alp
y (
J/g)
Native 3x 7x
Figure 5.6 The arithmetic means and standard deviations of denaturation temperatures (a) and
enthalpies (b) respectively, for Caucasian hair samples native, bleached 3, and 7 times
respectively indicating the effect of short acid treatments. ―O‖ refers to the initial values
recorded previously to the pH treatment
Chapter 5
104
Similar consideration can be used for interpreting the variation of DSC parameters as a
result of permanent waving treatment. Since X-ray diffraction reveals limited damage of the
crystalline (helical) material, which cannot explain the 50% decrease of the value of enthalpy,
we assume that the weakening of IFs-IFAP interface by the treatment plays the significant role.
The mechanism of permanent waving consists in breaking and reforming of S-S bonds in new
positions after having arranged the hair fibre in another shape. It is expected that the re-oxidation
of the disulphide bond in keratin protein does not reform all the bridges and the link between IFs
and IFAPs is weakened. This allows also the meta-stable intermediates to lose their coiled-coil
structure. According to Kuzuhara‘s work 52
the broken disulphide groups existing in the matrix
as a result of the reduction process do not return to the original conformation after the oxidation
process. Therefore we conclude that the large decrease of the value of enthalpy in the case of
hair material subjected to a permanent-waving process results from the incapacity to reform a
strong interface. The smaller shift of peak temperature, Tp, compared to those measured for
bleaching, is probable a consequence of the lower pH of the oxidation step (pH of 4.5) than of
the bleaching (pH of 10).
The results obtained from dyeing treatments support the importance of disulphide bonds:
the lack of cystine cleavage noticed at amino-acids analysis means that thermal or mechanical
properties are less affected. The diffusion of the components of dye into the hair cortex does
occur, as revealed by dye pigment formation in cortex, observed under optical microscopy, but
their reaction does not involve breaking and reformation of specific bonds from the keratin
structure. Negligible losses of tensile properties of hair dyed from a lighter to a darker shade are
also reported in literature 14
. Reasons for the contradictory behaviour of the matrix proteins with
regard to some literature data 48
could be the effective hydrogen peroxide concentration and the
treatment time. We may, still, conclude that permanent (oxidative) dyes generate limited damage
of hair structure when using low concentration of peroxide, relative short action times and
neutral pH.
Summing up, we consider that the tail domains contribute to the stability of filaments by
tailoring the filament-matrix interactions. While the tensile properties seems to give a good
estimation of the oxidative cystine bond cleavage, the thermal effect measured by DSC is a
cumulative effect of all participating groups mediated by the exposed tail domains. The
experimental recorded value of enthalpy when hard alpha-keratins are heated in water excess
refers to the total heat uptake required for collapsing the entire IFs structures (including the
interface) and not only to the alteration of the secondary structure (thermal denaturation) of the
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
105
helical material. There is good evidence suggesting that electrostatic interactions play a role in
these interactions, which explains why the thermal properties are modulated by changes in pH
during a short time process.
5.4. Conclusions
We have used differential scanning calorimetry to quantify the damage induced by
bleaching, permanent waving and oxidative dyeing cosmetic formulations on hard alpha-keratin
protein. We have evaluated the effects of cosmetic treatments within the framework of a three-
phase model in which the nonhelical (globular) terminal domains of keratin promote filament
interactions and control the thermal properties of keratin intermediate filaments. XRD, chemical
and mechanical data demonstrate that cosmetic treatments, using mild action time and
concentrations, affect primarily the amorphous area (matrix and interface) of the hair structure,
while the crystalline zones remain virtually unchanged. The DSC data suggests that keratin IFs
can modulate their organisation and thermal properties through chemical induced interactions.
The results suggest strongly the need for a careful interpretation of DSC parameters
variations in the context of formulations that are designed to change morphological components
within the hair cortex.
5.5. References and notes
1. Alberts, B. Essential Cell Biology: An Introduction to the Molecular Biology of the Cell; Garland
Publishing: New York, 1998.
2. Bischof, J. C.; He, X. Ann. NY Acad. Sci. 2005, 1066, 1-22.
3. Fraser, R. D. B.; MacRae, T. P.; Sparrow, L. G.; Parry, D. A. D. Int. J. Biol. Macromol. 1988, 10, 106-112.
4. Parry, D. A. D.; Fraser, R. D. B. Int. J. Biol. Macromol. 1985, 7, 203–213.
5. Powell, B. C.; Rogers, G. E. In Formation and Structure of Human Hair; Jolles, P.; Zahn, H.; Höcker, H.,
Eds.; Birkhäuser Verlag: Basel Switzerland, 1997, p 59-148.
6. Steinert, P. M.; Torkia, D. R.; Mack, J. W. In The Biology of Wool and Hair; Rogers, G. E.; Reis, P. J.;
Ward, K. A.; Marshall, R. C., Eds.; Chapman & Hall: London, 1989, p 157-167.
7. Zahn, H. Melliand Textilber 1991, 72, 926-931.
8. Er Rafik, M.; Doucet, J.; Briki, F. Biophys. J. 2004, 86, 3893-3904.
9. Parry, D. A. Adv. Protein. Chem. 2005, 70, 113-142.
10. Zahn, H.; Schaeffer, K.; Popescu, C. Biopolymers 2003, 155–202.
11. Birbeck, M. S. C.; Mercer, E. H. J. Biophys. Biochem. Cytol 1957, 3, 203-214.
12. Engel, A.; Eichner, R.; Aebi, U. J. Ultrastruct. Res. 1985, 90, 323-335.
13. Hearle, J. W. S. Wool Tech. Sheep Breed 2003, 51, 95-117.
14. Robbins, C. R. Chemical and Physical Behavior of Human Hair; Springer-Verlag: New York, USA, 2002.
Chapter 5
106
15. Marsh, J. M.; Clarke, C. J.; Meinert, K.; Dahlgren, R. M. J. Cosmet. Sci. 2007, 58, 319-327.
16. Popescu, C.; Sendelbach, G.; Wortmann, F. J. 10th Int. Wool Text. Res. Conf., Aachen, Germany, 2000.
17. Popescu, C.; Wortmann, F.-J. Revue roumaine de chimie 2003, 48, 981-986.
18. Robbins, C. R. Chemical and Physical Behavior of Human Hair; Springer-Verlag: New York, USA, 1994.
19. Wortmann, F. J.; Springob, C.; Sendelbach, G. J. Cosmet. Sci. 2002, 53, 219-228.
20. Wortmann, F. J.; Zahn, H. Text. Res. J. 1994, 64, 737-743.
21. Istrate, D.; Popescu, C.; Er Rafik, M.; Möller, M. Polym. Degrad. Stabil. submitted 2010.
22. Istrate, D.; Popescu, C.; Möller, M. Macromol. Biosci. 2009, 9(8), 805-812.
23. Bertrand, L.; Doucet, J.; Dumas, P.; Simionovici, A.; Tsoucaris, G.; Walter, P. J. Synchrotron Rad. 2003,
10, 387-392.
24. Gillespie, J. M. Science 1970, 170, 1100-1102.
25. Schmidt, H.; Wortmann, F. J. Text. Res. J. 1994, 64, 690-695.
26. Spei, M.; Holzen, R. Meliand Textiber 1989, 70, 371-376.
27. Istrate, D.; Popescu, C.; Wortmann, F. J.; Möller, M. Biomacromol. submitted 2010.
28. Wortmann, F. J.; De Jong, S. Text. Res. J. 1985, 55, 750-756.
29. Riekel, C.; Burghammer, M.; Mueller, M. J. Appl. Crystallogr. 2000, 33, 421-423.
30. Briki, F.; Doucet, J.; Etchebest, C. Biophys. J. 2002, 83, 1774-1783.
31. Busson, B.; Briki, F.; Doucet, J. J. Struct. Biol. 1999, 125, 1-10.
32. Busson, B.; Doucet, J. J. Struct. Biol. 1999, 127, 16-21.
33. Briki, F.; Busson, B.; Doucet, J. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 1998, 1429, 57-68.
34. Fraser, R. D. B.; MacRae, T. P.; Miller, A. J. Mol. Biol. 1964, 10, 147-156.
35. Fraser, R. D. B.; MacRae, T. P.; Miller, A.; Suzuki, E. J. Mol. Biol. 1964, 9, 250-252.
36. Jackson, W. M.; Brandts, J. F. Biochemistry 1970, 9, 2294-2301.
37. Privalov, P. L.; Potekhin, S. A. Methods. Enzymol. 1986, 131, 4-51.
38. Sturtevant, J. M. Annu. Rev. Biophys. Bioeng. 1974, 3, 35-51.
39. Sturtevant, J. M. Annu. Rev. Phys. Chem. 1987, 38, 463-488.
40. Ruegg, M.; Moor, U.; Blanc, B. J. Dairy Res. 1977, 44, 509-520.
41. Koshiyama, I.; Hamano, M.; Fukushima, D. Food Chem. 1981, 6, 309–322.
42. Eyring, H. In The Theory of Rate Processes in Biology and Medicine; Johnson, H. F.; Eyring, H.; Stover,
B. J., Eds.; John Wiley & Sons Inc: New York USA, 1974.
43. Joly, M. A Physico-chemical Approach to the Denaturation of Proteins; Academic Press: London 1965.
44. Cooper, A. In Protein: A Comprehensive Treatise; Allen, G., Ed.; JAI Press Inc: Stamford CT, 1999, p
217–270.
45. Feughelman, M. Text. Res. J. 1959, 29, 223-228.
46. Wortmann, F. J.; Deutz, H. J. Appl. Polym. Sci. 1993, 48, 137-150.
47. Monteiro, V. F.; Maciel, A. P.; Longo, E. J. Therm. Anal. Calorim. 2005, 79, 289-293.
48. Mitu, A. M.; Doctoral Dissertation, Rheinisch-Westf.
49. Leroy, F.; Franbourg, A. In 8th International Hair Science Symposium; German Wool Research Institute:
Kiel, 1992.
50. Nishikawa, N.; Tanizawa, Y.; Tanaka, S.; Horiguchi, Y.; Asakura, T. Polymer 1998, 39, 3835-3840.
DSC analysis of structural changes in bleached, perm-waved and dyed hard alpha keratin fibre
107
51. Spei, M.; Holzem, R. Melliand Textilber 1989, 70, 786-787.
52. Kuzuhara, A. Biopolymers 2007, 85, 274-283.
53. Kuzuhara, A. Biopolymers 2005, 79, 173-184.
54. Kuzuhara, A. Biopolymers 2006, 81, 506-514.
55. Crewther, W. G.; Fraser, R. D. B.; Lennox, F. G.; Lindley, H. Adv. Protein Chem. 1965, 20, 191-346.
56. Fraser, R. D. B.; MacRae, T. P.; Parry, D. A. D.; Suzuki, E. Proc. Natl. Acad. Sci., USA, 1986, pp 1179-
1183.
57. Kreplak, L.; Franbourg, A.; Briki, F.; Leroy, F.; Dalle, D.; Doucet, J. Biophys. J. 2002, 82, 2265-2274.
58. Steinert, P. M.; Marekov, L. N.; Fraser, R. D. B.; Parry, D. A. D. J. Mol. Biol. 1993, 230, 436-452.
59. Franbourg, A.; Leroy, F.; Lévêque, J. L.; Doucet, J. In 5th International Conference on Biophysics and
Synchrotron Radiation: Grenoble, France, 1995.
60. Rao, D. R.; Gupta, V. B. J. Appl. Polym. Sci. 1992, 46, 1109-1112.
61. Fraser, R. D. B.; MacRae, T. P.; Rogers, G. E.; Filshie, B. K. J. Mol. Biol. 1963, 7, 90-94.
62. Briki, F.; Mérigoux, C.; Sarrot-Reynauld, F.; Salomé, M.; Fayard, B.; Susini, J.; Doucet, J. J. Phys. IV
2003, 104, 337-340.
63. Fraser, R. D. B.; MacRae, T. P.; Millward, G. R.; Parry, D. A. D.; Suzuki, E.; Tulloch, P. A. Appl. Polym.
Symp. 1971, 18, 65-77.
64. Fink, A. L.; Calciano, L. J.; Goto, Y.; Nishimura, M.; Swedberg, S. A. Protein Science 1993, 2, 1155.
65. Lala, A. K.; Kaul, P. J. Biol. Chem. 1992, 267, 19914-19918.
66. Wikipedia. Wikimedia Foundation, Inc, 2008.
67. Privalov, P. L. Adv. Protein Chem. 1982, 35, 1-104.
* Journal of Physical Chemistry B, 113 (35), p 12136–12147, 2009
Appendix A: Morphology and molecular mobility of
fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
*
A.1. Introduction
The hard α–keratin is a filament protein found in mammalian epidermal appendages (hairs,
quills, horn, nails, etc.) distinct from feather β–keratin found in avian and reptilian tissues. The
hair is the most sophisticated biological composite material1. The structure of hard α-keratin is
characterized by three structural hierarchy levels2. At high resolution, the intermediate filament
(IF) protein is made of a central rod domain of amino acid sequences (1A, 1B, 2A, and 2B)
containing an aminoacid heptad repeat unit, and separated by loop links (L1, L12, and L2)3,4
. At
the extremity of the rod domain are located the globular C- and N-terminal domains arranged
mostly in β-sheets and formed of sulphur rich compounds5,6
. Two strands of α-helices are coiled
coil to form a superhelical dimer. At the medium resolution, i.e. the intermediate level
arrangement of the heterodimers inside IFs, the molecules are assembled both longitudinally and
laterally in an ensemble called a microfibril7. The dimers are associated as tetramers, which
group to form a long cylinder-shaped intermediate filament with 32 keratin chains in cross-
section. At lower structural resolution, the bundles of parallel IFs are organized in amorphous
and disordered crystalline lateral network. These are embedded in a sulphur-rich protein matrix
of intermediate filaments associated proteins (IFAPs) and form a macrofibril, the main
morphological components of hard α-keratin fibers1,2
.
Although the above model was proposed for describing the mechanic behaviour of
keratins, it appeared to also be suitable for explaining the high denaturation temperature found in
keratins8,9
. In soluble proteins, the helix denaturates (unfolds) at temperatures up to 80° C. There
are no data on the denaturation temperature of IFs alone (not surrounded by a matrix), but one
may expect that the α-helix from keratins would also unfold at temperatures around 80° C. It is
assumed that the fact that keratin proteins show denaturation at above 100° C is due to the
rigidity of the matrix, whose viscosity impedes the unfolding of the helix. The viscosity (and
cross-link) of the matrix governs, therefore, the segmental mobility of the α-helix and the
unfolding reaction (denaturation).
Appendix A
110
A similar model was proposed for collagen based materials10
. The model suggests that
adding solvents able to decrease the viscosity of the matrix depresses the temperature of
unfolding. This has been indeed noticed in DSC experiments with keratin fibres11,12
and with
collagen based materials (parchments, leathers) in a water environment13-15
. Understanding
properly how the keratins protect the intermediate filaments against thermal denaturation until
high values of temperature is of a clear interest for the fundamental knowledge of protein
denaturation. The role of the matrix in this process may suggest ways for designing high-
temperature stable proteins as new biomaterials.
Multinuclear and multidimensional liquid- and solid-state NMR are important techniques
in structural biology16-18
. Recently, a 13
C and 2H solid-state NMR study of an α-keratin sourced
from equine hoof has revealed a strong dependence of molecular conformation and molecular
dynamics on the degree of hydration of the material19
. In particular, dehydration results in a
much more rigid and ordered structure, with a loss of α-helical components in the structure and
breaking of cysteine disulfide bonds. Moreover, the molecular dynamics and structural
organization of mouse epidermal keratin intermediate filaments (IF) have been studied via 13
C
and 2H spectroscopy and relaxometry on IF labelled with isotopically enriched amino acids
20.
Solid-state 31
P NMR spectroscopy was also applied to the analysis of phosphorylated wool
keratin to investigate the changes induced on the surface of wool keratin21
.
The clarification of the fine structure of fibrous proteins like keratin in the solid state is
important for the understanding of their nature. This may be achieved because 13
C and 15
N
chemical shifts of polypeptides are substantially dependent on their main-chain conformations
such as α-helix and β-sheet forms. Using this method it was confirmed that both right-handed α-
helix and β-sheet forms exist in native wool fiber22-24
.
Spin-diffusion NMR was proved to be a useful method for characterization of
semicrystalline polymer morphology25-29
. The sizes of the rigid, interfacial, and amorphous
fractions can be estimated from such experiments and the results compared well to that from
TEM and X-ray diffraction. Recently, the morphological domain sizes of thermally denaturated
wool keratin were measured by 1H spin-diffusion NMR experiments
30. For the interpretation of
these experiments the solutions of the spin-diffusion equations for two-dimensional square and
cylindrical morphologies were employed. The keratin mobility gradient in the interfacial region
at different denaturation temperatures was also measured from the 1H spin-diffusion data. A
qualitative model describing the denaturation process of hydrated keratin protein was developed
that explains the changes in domain thickness, spin diffusivities, phase composition, and
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
111
thermodynamic parameters.
The aim of this work is to investigate the changes induced by various chemical treatments
on hair keratin and the thermal denaturation process of these materials by 1H NMR wide-line
spectroscopy and 1H spin diffusion. Moreover, cross-polarization magic angle spinning
(CPMAS) 13
C NMR spectra and thermally polarized and hyperpolarized 129
Xe spectra were used
for this purpose. The phase (fraction) composition is measured from 1H wide-line spectra ex situ
for native and chemically treated Caucasian hair sampled at different temperatures during the
denaturation process. Three fractions are detected, i.e., rigid, interfacial, and amorphous. The
rigid domain sizes for the hair samples were measured by 1H spin-diffusion using initial-rate
approximation. The molecular dynamics gradient of the interfacial region was investigated using
the 1H spin-diffusion NMR experiments. The changes in the degree of fibrous hard α-keratin
organization, the amount of different phases, and molecular dynamics are discussed in
correlation with the type of hair chemical treatments and temperatures during the thermal
denaturation process.
A.2. Materials and methods
The hard α-keratin fibers used for this study were of Caucasian dark-brown hair, supplied
by Kerling International Haarfabrik GmbH, Germany. The fibres were cleaned with 1% lauryl
ether sulphate (LES) and dried at room temperature prior to working with them. The pH of their
aqueous extract was found to be 6.5 to 7.
Damaging treatment
The damage of the keratin fibres was induced by oxidative and reductive/oxidative
chemical treatments, respectively.
The oxidative treatment was performed on 1 g keratin fibres with 0.2 g potassium
persulphate mixed with 1.2 mL of 6% hydrogen peroxide solution to form a paste adjusted at pH
8.5-9 with ammonia. The fibres were covered with the paste, massaged gently between the
fingers and left 30 minutes to react at room temperature. The fibres were then rinsed thoroughly
until the pH of the aqueous extract was checked to be 7. The treatment was resumed two more
times.
The reductive/oxidative treatment was performed with thioglycolic acid (TGA) and
hydrogen peroxide. A 1 g portion of keratin fibres, pre-wetted with water, was immersed for 30
seconds in the reducing solution of 8% w/w TGA at pH 8.5-9 adjusted with ammonia, the
solution excess being removed by gently pressing the fibres between the fingers, then covered
Appendix A
112
with plastic folia and let to react for 30 minutes at room temperature. The fibers were then rinsed
with tap water (3 minutes) and immersed in the oxidative (hydrogen peroxide 3%) solution
adjusted at pH 4.5 with phosphoric acid, for 30 seconds. After squeezing between fingers, fibers
were allowed to react with the oxidative solution for 10 minutes, at room temperature. Finally,
the fibres were washed thoroughly under tap water for 3 minutes, shampooed for 1 minute (70%
Natrium Laurethsulfat, pH 7), rinsed 1 minute with warm water, rinsed with tap water for 3
minutes, and dried under hot air blow. The process was repeated two more times.
Separately, we have prepared a sample in which the disulphide bonds were broken and
protected by alkylation. A 1 g portion of fibres was reacted with 8 mL of 0.5 M tris(2-
carboxyethyl)phosphine hydrochloride (TCEP), at pH 7 adjusted with ammonium hydroxide, 48
hours under continuous stirring at room temperature. After removing the TCEP solution excess,
10 mL of iodacetamide (1M, pH 8) was added without previously washing the fibre material (to
avoid reformation of disulphide bonds) and the sample was kept for 48 hours under continuous
stirring at room temperature and in the dark. Eventually, the fibres were rinsed under tap water
for 3 minutes, two times subsequently washed with a solution of Texapon N70, 0.1 mL/L, (70%
natrium laurethsulfat), rinsed again with warm water 1 minute and then tap water 3 minutes, and
dried in air.
DSC measurements
The DSC experiments were run in a DSC-7 Perkin Elmer instrument calibrated with
indium and palmitic acid, both of high purity, using pressure resistant stainless steel large
volume capsules. DSC calibration was done with indium and palmitic acid, both of high purity.
We used a heating rate of 10 K/min for temperature ranging from 60 to 180°C. Each experiment
was repeated three to five times, for ensuring the reproducibility of data.
Prior to the DSC measurements the samples were cut into fine snippets (about 2 mm) and
stored under controlled conditions (about 24 hours at 22 0C and 55% relative humidity) to ensure
invariant water contents. The amount of 7-10 mg of each sample snippets were weighted and
placed in crucible for the DSC measurements. Prior to sealing a crucible, 50 L of distilled water
(pH 6.7) was added, and the sealed crucible was stored over night for about 14 hours, to allow
the fibres to wet. The samples for NMR measurements were gathered from DSC experiments by
taking the pans at various moments linked to thermal events as disclosed by DSC. Three
different samples, which will be reported below, were collected this way at various temperatures
including the denaturation temperature.
Proton and 13
C NMR Measurements
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
113
Proton solid-state NMR spectra, 1H double-quantum (DQ) build-up curves,
1H spin-
diffusion, and 13
C CPMAS spectra were measured on a Bruker DSX-500 spectrometer operating
at 500.45 and 125.84 MHz for 1H and
13C, respectively. Proton NMR data were collected at
room temperature for non-spinning samples. The dead time of the spectrometer is 5.5 s. The
length of a /2 pulse was about 5.5 s, the dwell time was 2 s, and the recycle delay was 3 s for
all measurements.
Figure A.1 Scheme for the spin-diffusion experiment with a DQ filter. The first two pulses
excite DQ coherences that evolve for a short time tDQ. These coherences are converted by the
following two pulses into z-magnetization. The spin diffusion takes place during the time
interval of duration td. The last pulse readout is the distribution of magnetization between
different keratin phases
Proton spin-diffusion measurements were performed using the general scheme consisting
of a double-quantum (DQ) dipolar filter, a spin-diffusion period, and an acquisition period as
presented in Figure A.1. The gradient of magnetization was created by the dipolar filter that
excites DQ coherences (Figure A.1) and selects mainly the magnetization of the rigid phase
(fraction)28-30
. The pulse sequence is based on the two pulses acting during the excitation and
reconversion periods. The value of the excitation/reconversion times used in the spin-diffusion
experiments is = 7 s. It corresponds to the rising region of the DQ build-up curve for each
sample (see below).
The experimental wide-line spectra were decomposed in three components using the
DMFIT program. The broad component describing the rigid fraction of the spectra was
approximated by a Gaussian function. A Lorentzian line shape was used to describe the narrow
component of the spectra corresponding to the mobile phase. A combination of Gaussian and
Lorenzian functions was used to describe the intermediate line corresponding to the interface.
Appendix A
114
The proton NMR DQ build-up curves were recorded for setting the optimum parameters of
the DQ dipolar filter. They were measured on a Bruker DSX-500 spectrometer at a proton
resonance frequency of 500.45 MHz. The duration of the applied 90° pulses was 5.5 s. The DQ
evolution time and the z-filter delay were fixed to tDQ = td = 5 s (Figure A.1).
13C NMR spectra were measured using cross-polarization (CP) magic-angle sample
spinning (MAS) with power decoupling by the two-pulse phase modulation (TPPM) method at a
rotor frequency of 5 kHz. The contact pulse for CP has duration of 2 ms. All NMR
measurements were made at room temperatures.
Hyperpolarized and Thermally Polarized 129
Xe NMR Measurement
The Rb-Xe gas hyperpolarizer working in the continuous flow mode was build at the
Research Centre Jülich, Germany by the group of S. Appelt. The gas mixture used for
hyperpolarization consists of 98% helium, 1% nitrogen and 1% xenon at a pressure of 7 bar. The
gas flow through the pumping cell and the flow can be regulated by a needle valve and
controlled by a flowmeter. The typical flow rate used was about 300 cm3/min. The total degree
of polarization which was achieved by this hyperpolarizer varied in the range of 20-35%. The
xenon pressure was 5 bar during the NMR measurements. The hyperpolarized 129
Xe gas flows
through a 7 meter plastic tube into the sample cell which is positioned in a 200 MHz Bruker
spectrometer. During the transit time to the fringe field of the NMR spectrometer, the
hyperpolarized 129
Xe gas experienced the stray magnetic field of the superconductive magnet.
Due to short transit time of about 50 s the depolarization was assumed to be negligible.
The 129
Xe NMR spectra were measured at the resonance frequency of 55.3 MHz and room
temperature. The length of the radio-frequency pulse was 50 s and the recycle delay was 30 s.
The partial dehydration of the hair samples was obtained by keeping the samples for several
hours under a vacuum until the pressure in the system reached 6 x 10-4
mm Hg.
The NMR spectra measured with thermally polarized 129
Xe used a homemade sapphire
tube with a volume of 4.89 cm3 sealed by a titanium valve which was approved for pressures up
to 50 bar. For the NMR measurements, the tube was loaded with 129
Xe gas at a natural
abundance of 26.4 % and a pressure of 20 bar. The 129
Xe NMR spectra were measured at room
temperature with a Bruker 500 MHz NMR spectrometer with a recycle delay of 120 s.
A.3. Theory of NMR spin diffusion
Spin-diffusion observables. The transport of z-magnetization oriented along the static
magnetic field in an NMR experiment can be described by the diffusion equation in the
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
115
continuum approximation. The concentration trm ,
of nuclear z-magnetization at the position
r
in the sample from the centre of symmetry of different morphologies (cf. Figure A.2) at the
moment of time t is defined by
rVr
trMtrm z
,, , (A.1)
where trM z ,
is the total z-magnetization and rV
is the infinitesimal volume around the
point defined by the vector r
. The number density of spins is denoted by r
.
In the limit of isotropic spin-diffusion and spatially constant spin diffusivity (D), the spin
diffusion equation has the form
trmDt
trm,
, 2
. (A.2)
The instantaneous NMR observables in a spin-diffusion experiment are represented by the
normalized integral intensity 0/ ItIi of the ith
component of the NMR spectrum with the total
integral intensity 0I . More specific the NMR spin-diffusion observables are defined by
00
,
I
rdtrm
I
tIiV
ii
i
(A.3)
where Vi is the volume of the ith
domain.
Solution of the spin-diffusion equation for a finite source and semi-infinite sink.
The real morphology of keratin in hair can be approximated by a square transverse morphology
(Figure A.2).
We assume that the spin-diffusion takes place in a heterogeneous matrix from a source R with
low segmental mobility represented by the intermediate filaments into a semi-infinite sink M
with larger segmental mobility corresponding to the amorphous phase of the keratin.
The above morphology is valid only for short spin diffusion time t i.e., RR Ddt /2 , where dR is
the size of the rigid domain R (source of magnetization) and DR is the spin diffusivity for the R
domain. The interfacial region is taken together with the amorphous fraction in the following
considerations.
Appendix A
116
Figure A.2 (a) Schematic representation of the three-phase model of keratin fibres. The
intermediate filaments (IFs) are imbedded in an amorphous keratin matrix and stabilized by the
interface. (b) Schematic representation of the square morphologies with finite source and semi-
infinite sink used to approximate the initial regime of the spin-diffusion process. The size of the
rigid domain is doted by dR
The solution of the spin diffusion equation for the composite medium of finite source and
semi-infinite sink can be obtained using the solution for a one-dimensional (1D) composite
medium27,31-33
. For a -dimensional diffusion process with > 1, the solution of the spin diffusion
equation can be written simply as a product of the solutions for the 1D diffusion process27
. For
this, the essential condition is that the initial conditions must be expressible as a product of those
for the one-variable problems taken separately. The space and time evolution of the
concentration of magnetization in the R domain is given by27
1
0000
4
2/)(,
i R
iR
MMRR
RMMM
MMRR
MMMRRR
RtD
xderf
DD
mmD
DD
mDmDtrm
(A.4)
where xi are the coordinates of the vector 321 ,, xxxr
and 2/Ri dx . The error function is
defined as
z
x dxezerf0
22
(A.5)
A highly efficient dipolar filter is characterized by the initial concentration of magnetization:
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
117
00 Rm , and 00 Mm . For such a condition eqn. A.4 has the form
1
00
4
2/,
i R
iR
MMRR
RMM
MMRR
RRR
RtD
xderf
DD
mD
DD
mDtrm
(A.6)
Using the results presented in ref. 27 (eqn. A.30) and the above equations A.3, A.5 and A.6 we
get for the time evolution of the integral intensity of the NMR signal from domain R the
relationship:
tD
dierfc
d
tD
DD
D
DD
D
I
tI
R
R
R
R
MMRR
MM
MMRR
RRR
4
141
0
(A.7)
where the integral error complement function is
z
dxxerfzierfc 1 (A.8)
At the beginning of the spin-diffusion process for short spin diffusion times t, the quantity
1tD
d
R
R and 0ierfc . It is evident from eqn. A.7 that in the initial regime of the spin-
diffusion, i.e., for RR Ddt /2 the time dependence of the NMR observable 0/ ItIR is linear in
(t)1/2
and is given by
R
R
MMRR
MM
MMRR
RRR
d
tD
DD
D
DD
D
I
tI
41
0
(A.9)
The spin-diffusion decay curve described by eqn. A.9 corresponds to an initial slope straight line
that intersects the (t)1/2
axis at (t0)1/2
. The domain thickness dR for a rectangular 1D, 2D, or 3D
morphology is given from eqn. A.9 by
0
4t
DD
DDd
MMRR
MRM
R
(A.10)
In the time regime in which the spin diffusion is not affected by the spin-lattice relaxation, the
theorem of total magnetization conservation leads to
100
I
tI
I
tI MR (A.11)
Hence, from equations A.9 and A.11, the time evolution of the spin-diffusion build-up curve for
the sink domain (M) has the same slope as that of the source domain. The intercept of the tangent
straight line starting from t = 0 with the horizontal line at I(0)/I0 = 1 for the spin-diffusion build-
up curve will lead to the same t0 value as that of the decay curve eqn. A.10. This is a direct
Appendix A
118
consequence of finite and semi-infinite morphology. This morphology is a good approximation
for the morphology with both finite domains in the initial-rate regime. The thickness of the
mobile domain dM can be obtained by the same procedure discussed above using a Goldman-
Shen dipolar filter34
.
We can also note that the derivation of the relationship for dR (eqn. A.10) employs only the
solution of the spin diffusion equation with corresponding initial and boundary conditions.
Moreover, the time evolution is considered for the normalized magnetization. This is not the case
for the intercept spin diffusion time, reported in refs. 26 and 35, where the phase structure
considerations and magnetization at equilibrium were taken into account.
A.4. Results and discussions
Thermal denaturation by DSC. Typical DSC plots in D2O for the temperature range of
denaturation of hard α-keratin in the untreated state, after oxidative, and after reductive/oxidative
treatment, respectively, are shown in Figure A.3. The endothermal process recorded around
154°C for the untreated sample is attributed to the thermal denaturation of keratin by melting of
the α-helix crystalline structure26
. The scenario for the thermal denaturation of hard α-keratin in
the native and after chemical treatments as reflected in the DSC and NMR data will be discussed
below.
Figure A.3 DSC signals of hard α-keratin, in the native state, after oxidative, and
reductive/oxidative treatments in D2O for the denaturation temperatures. The arrows mark the
temperatures at which the samples were used in the NMR measurements
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
119
Proton NMR Spectra, Phase Composition and Molecular Dynamics. The proton NMR
spectrum of hard α-keratin, recorded under static conditions at room temperature, is presented in
Figure A.4. The best fitting parameters have been found by decomposing the spectra in three
lines described by a Gaussian, a Lorenzian, and a combination of Gaussian and Lorenzian
functions, respectively. The broad component, associated with the Gaussian line, corresponds to
the rigid phase. The Lorenzian line associated with the narrow component of the spectra
describes the mobile phase. The intermediate line, described by the combination of Gaussian and
Lorenzian functions is associated with the interface.
The phase composition for hard α-keratin, in the native state, after oxidative, and
reductive/oxidative treatments in D2O for the temperature range where denaturation takes place
is shown in Figure A.5. The denaturation temperature of hard a-keratin is 154°C, for the hair
samples after the oxidative treatment is 122°C and for the sample subjected to
reductive/oxidative treatment is 144°C.
The measurements reveal a slight increase in the relative amount of rigid phase and a
decrease of the interface for native hard α-keratin. We can note an opposite behaviour for the
sample after the oxidative and reductive/oxidative treatments.
Figure A.4 Proton wide-line NMR spectrum of hard α-keratin. All NMR spectra were
decomposed into three components corresponding to the rigid, semi-rigid (interface) and mobile
fractions
Appendix A
120
145 150 155 160 165 170 175 180
10
20
30
40
50
60
70
hard -keratin
temperature [°C]
phase fra
ction [%
]
rigid
interface
mobile
120 130 140 150 160 170 180
10
15
20
25
60
65
oxidative treat.
temperature [°C]
ph
ase
fra
ctio
n [
%] rigid
interface
mobile
140 150 160 170 1800
10
20
70
80
reductive & oxidative treat.
rigid
interface
mobile
pha
se f
raction
[%
]
temperature [°C]
Figure A.5 Phase composition of hard α-keratin, in the native state, after oxidative, and
reductive/oxidative treatments for different temperature ranges
The amount of mobile (amorphous) fraction is not essentially affected by the denaturation
temperature. The reductive/oxidative treatment increases the relative amount of rigid fraction as
the expense of interface compared to the hard α-keratin.
The molecular dynamics of hard α-keratin, after oxidative, and reductive/oxidative
treatments for the temperature range where denaturation occurs reflected in the line width of the
1H spectral components is shown for the rigid, interface, and mobile fractions in Figure A.6. In
general denaturation at 180 0C induces a greater disorganization in the nanostructured keratin
and hence a large molecular mobility.
An exception is the mobile fraction of the sample after reductive/oxidative treatment. The
molecular motion is strongly hindered by the matrix disorganization induced by this chemical
treatment. The molecular motions are more hindered for the rigid phase and interface of hard α-
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
121
keratin.
120 130 140 150 160 170 18039.0
39.5
45
46
47
hard -keratin
oxidative treat.
reductive & oxidative treat.
linew
idth
[kH
z]
temperature [°C]
rigid
120 130 140 150 160 170 18015
16
17
19
20
21
reductive & oxidative treat.
hard -keratin
oxidative treat.
linew
idth
[kH
z]
temperature [°C]
interface
120 130 140 150 160 170 1803.0
3.2
3.4
3.6
3.8
4.0
4.2
4.4
4.6
4.8
reductive & oxidative treat.
hard -keratin
oxidative treat.
linew
idth
[kH
z]
temperature [°C]
mobile
Figure A.6 Full line width at half-intensity of the NMR spectral components corresponding to
rigid, interface, and amorphous fractions of hard α-keratin, in the native state and after oxidative,
and reductive/oxidative treatments for different temperature intervals
Double-Quantum Dipolar Filter for 1H Spin Diffusion. The spin diffusion experiments
observe the equilibration of spatially heterogeneous magnetization over the sample. A
magnetization gradient can be created, for example, with a dipolar filter which excites double-
quantum (DQ) coherences24
. This type of filter is more advantageous than a dipolar filter for
mobile domains because it allows a more accurate detection of the narrow signals on the top of
the broad component as compared to the detection of a broad component under a narrow signal.
This is valid especially at short diffusion times when the magnetization of one of the component
is very small.
The DQ filter can be set such to select the magnetization only from the most rigid part of a
heterogeneous sample. By choosing appropriate excitation/reconversion times (Figure A.1) of
Appendix A
122
the double-quantum coherences, the magnetization corresponding to the stronger dipolar
couplings will pass through the filter and that of the weaker dipolar couplings is filtered out. The
optimum value of can be chosen by recording 1H DQ build-up curves. The maxima of the DQ
build-up curves appear at very short excitation/reconversion times of about 10-12 s for all
investigated samples. In this range of values, the mobile component is completely filtered out
as shown below.
The DQ filtered NMR spectra recorded for different values of the excitation/reconversion
times are shown in Figure A.7 for the hard α-keratin sample. For short values the DQ filtered
1H spectrum edits mainly the spin-pairs of aminoacids with the strongest dipolar couplings
(Figure A.7, top). In the region of the maximum of the DQ build-up curves the pulse sequence
edits a dipolar network of many spins corresponding to the crystalline and partially the interface
fraction (Figure A.7, middle). The 1H spectrum in Figure A.7 (bottom) filtered only the mobile
keratin from the amorphous fraction. The value of =5 s has been chosen for the dipolar filter
of the rigid domain, which still keeps the filter efficiency close to unity with a reasonable value
of the signal–to–noise ratio.
-200 -100 0 100 200
35 s
2 s
ppm
17 s
a)
b)
c)
Figure A.7 Proton DQ filtered NMR spectra recorded for hard -keratin at different values of
the excitation/reconversion times (a) sb s, and (c) s
Proton Spin Diffusivities. An accurate analysis of the domains thickness by NMR spin
diffusion experiments requires three steps. These are as follows: (i) an optimization of a dipolar
filter to obtain the highest selectivity to the different phases, (ii) knowledge of the spin diffusion
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
123
coefficients for modelling the experimental data, and (iii) proper choice of a model that describes
the morphology of the material studied.
The values of the spin-diffusion coefficients DR, and DM for the rigid and mobile fractions,
respectively, can be determined by approximating, the NMR line shapes of the rigid and the
mobile fractions by Gaussian and Lorentzian functions, respectively. The equations describing
the spin-diffusion coefficients for the rigid (Gaussian line) and mobile (Lorentzian line) regions
are given by20
212
R2ln212
1
rD (A.12)
and
2121
2M
6
1rD (A.13)
where is the cutoff parameter of the Lorentzian line, 1/2 is the full line width at half height,
and r2 is the mean square distance between the nearest spins. An estimation of
2/12r 0.22
nm was given for keratin taken into account the amino acid composition13
.
The calculated spin-diffusion coefficients using equations A.12 and A.13 are shown in Figure
A.8. For each denaturation temperature and type of sample the specific values of DR, and DM are
used for domain size evaluation. The largest value for DR, showing the highest organization and
packing corresponds to hard α-keratin (Figure A.8a). The morphology dezorganization due to
reductive/oxidative treatment leads to a reduction of DR. This trend is also valid for DM (Figure
A.8b).
120 130 140 150 160 170 1800.21
0.22
0.23
0.24
0.25
0.26
0.27
diffu
sio
n c
oeffic
ient [n
m2/m
s]
hard -keratin
oxidative treat.
reductive & oxidative treat.
temperature [°C]
rigid
120 130 140 150 160 170 1800.400
0.405
0.460
0.465
0.470
0.475
0.480
0.485
temperature [°C]
mobile
d
iffu
sio
n c
oe
ffic
ien
t [n
m2/m
s]
hard -keratin
oxidative treat.
reductive & oxidative treat.
a) b)
Figure A.8 Effective 1H spin diffusivities DR (a) and DM (b) evaluated from eqn. A.12 and eqn.
A.13 for different hard α-keratin as a function of the denaturation temperature
Appendix A
124
Morphology and Domain Sizes. The spin-diffusion experiments were performed on native
hard α-keratin and chemically treated samples after they were heated at the temperatures shown
in Figure A.3. Proton wide-line NMR spectra recorded at three different diffusion times td are
shown in Figure A.9. In all cases, the flow of magnetization from the rigid domain into the
mobile domain is observed with increasing diffusion times. At short diffusion times, mainly the
rigid fraction of keratin composed of the α-helical conformation of the intermediate filament is
observed, and can be seen in Figure A.9 for td=40 s. Upon increasing the spin diffusion time,
for example, at td=250 s and td=600 ms, the relative intensity of the rigid fraction in the spectra
decreases, and the intensity of the narrow line that originates from the soft amorphous fraction
represented by the keratin matrix surrounding the intermediate filament increases (Figure A.9).
-100000 -50000 0 50000 100000
600 ms
250 s
40 s
[Hz]
Figure A.9 Proton wide-line NMR spectra recorded at three different spin-diffusion times td of
40 s, 250 s, 600 ms after the action of the DQ dipolar filter (Figure A.1)
The presence of the highly mobile amorphous regions complicates the interpretation of the
spin diffusion data. Due to the fact that it is less than 10% for all samples and that the flow of
magnetization is reaching it only after longer spin diffusion times, our approach will mainly
focus on the transfer of magnetization between the crystalline and the less-mobile amorphous
regions. Therefore, a renormalization of the integral intensities corresponding to these two
phases was made by adding the signal of the amorphous phase to the signal of the interface. The
time evolution of NMR observables for the reductive/oxidative treated hard α-keratin sample
with increasing spin-diffusion time is shown in Figure A.10 for the rigid and less rigid fractions.
The quasi-equilibrium is reached after about 4 ms, which is less than the longitudinal relaxation
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
125
times.
0 5 10 15 20
0.0
0.1
0.8
0.9
1.0
d0t
0 2 4 60.0
0.2
0.4
0.6
0.8
1.0
no
rma
lize
d s
ign
al in
ten
sity
td
1/2 [ms
1/2]
reductive & oxidative treat.
T = 180o C
no
rma
lize
d s
ign
al in
ten
sity
td
1/2 [ms
1/2]
rigid
mobile
Figure A.10. Normalized signal intensities for the 1H spin diffusion experiment (Figure A.1) for
a hard α-keratin after the reductive/oxidative treatment and denaturated at 180 0C. The initial
slope and the intercept at (td0)1/2
is shown in the insert
140 145 150 155 160 165 170 175 180 185
12
14
16
18
20
22
24
26
hard -keratin
temperature [° C]
t d0
1/2 [m
s1/2]
a)
120 130 140 150 160 170 18017
18
19
20
21
22
23
24
25
26oxidative treat.
temperature [° C]
t d0
1/2 [m
s1/2]
b)
140 150 160 170 180
5.95
6.00
6.05
6.10
6.15
6.20
6.25
6.30reductive & oxidative treat.
t d0
1/2 [m
s1/2]
temperature [° C]
c)
Figure A.11 The denaturation temperature dependence of the intercept (td0)1/2
for hard α-keratin,
in the native state, after oxidative and reductive/oxidative treatments
Appendix A
126
To estimate the domain sizes for the rigid and less-mobile amorphous domains based on
the analysis of spin diffusion data using initial rate approximation. The symmetric display of
(td0)1/2
is shown in the insert of Figure A.10, and the two-dimensional (2D) morphology ( 2 )
is considered in eqn. A.10. The values of (td0)1/2
for hard α-keratin, in the native state, after
oxidative, and reductive/oxidative treatments of the denaturation temperatures are given in
Figure A.11.
The rigid domain thickness of hard α-keratin samples for different denaturation
temperatures are presented in Figure A.12. It is evident that in the case of reductive/oxidative
treatment of hard α-keratin the rigid domains do not change in the limit of experimental errors
with the denaturation temperatures. Moreover, the disorganization in the intermediate filaments
will reduce with about 50% the domain size as compared to hard α-keratin and the sample
submitted to the oxidative treatment. For the last two samples, dR decreases for the largest
denaturation temperature
120 130 140 150 160 170 180
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
dom
ain
siz
e [nm
]
temperature [°C]
hard -keratin
oxidative treat.
reductive & oxidative treat.
Figure A.12 Rigid domain sizes for rigid and mobile + interface fractions of hard α-keratin with
different treatments as a function of denaturation temperature
Dynamic Heterogeneity of Hard α-Keratin Fibre Interface. The 1H spin-diffusion
experiment using a DQ filter was discussed above for the hard α-keratin at different denaturation
temperatures. The evolution of the z-magnetization front can be measured from the spectral
component decomposition. At the beginning of the spin diffusion experiment after the action of
the dipolar filter, the magnetization is stored only in the rigid domain. For short spin diffusion
times, the magnetization is present only in the interface and at the longer diffusion times it will
reach the mobile region. The average distance 2/1
2z travelled by the 1H z-magnetization can be
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
127
estimated using the Einstein relationship in one dimension, i.e.,
dMR t
DDz
22
2/12 (A.24)
where 2
MR DD is the average spin diffusivity, and td is the spin-diffusion time. Therefore, in
the initial time approximation 2/1
2z is proportional with 2/1
dt and also with the average spin
diffusivity.
The interfacial region has a gradient in molecular mobility, and therefore a change in the
line width at half-intensity of the spectral component can be detected. At smaller spin diffusion
times, the experiment edits the part of the interface closer to rigid region and at the longer
diffusion times the most mobile part of the interface connected to the mobile region. The
changes in the line-width at the half-intensity Δν1/2 of the interface spectral component are
shown in Figure A.13 for hard α-keratin at different temperatures before and after denaturation
temperature of 1540
C.
0 2 4 6 8 10 12 14 16 18 20
20
22
24
26
28
30
32
34hard -keratin
td
1/2 [ms
1/2]
1/2 [kH
z]
145o C
154o C
160o C
180o C
Figure A.13 Full line width at the half-intensity of the interfacial component of the 1H NMR
spectrum for hard α-keratin measured at different denaturation temperatures
It is interesting that the dynamic heterogeneity of the interface shows a maximum in the
molecular mobility. This is more pronounced for the hard α-keratin denaturated at T=154 0C.
The width of the molecular mobility heterogeneity measured in the units of spin diffusion time is
almost the same for different temperatures around DSC denaturation peak (Figure A.3).
Moreover, the heterogeneity of the molecular dynamics at the interfacial region in hard α-keratin
Appendix A
128
at the extreme temperatures of 145 0C and 180
0C before and after the DSC peak is almost the
same, showing a reorganization of the morphology after the denaturation occurs. Nevertheless,
this is not complete because the molecular dynamics at 180 0C is slightly faster compared to that
at 145 0C, as it is evident from the right hand side of Figure A.13.
13C CPMAS Spectra of Chemically Treated Hard α-Keratin.
Cross-polarization (CP) MAS 13
C spectra of hard α-keratin, in the native state, after
oxidative, reductive/oxidative and disulfide bonds scission treatments at room temperature are
shown in Figures A.14 and A.15.
These spectra are similar with those reported for α-keratin of equine hoof under hydrated
condition and thermally denaturated wool keratin19,30
. Many spinning-sidebands are present in
these spectra due to the large values of the 13
C chemical shielding anisotropy (cf. Figure A.14).
The 13
C spectrum show several distinct regions: (i) a broader signal due to the α-carbons at
54 ppm, (ii) the peak at 40 ppm which has contributions from β-carbon in leucine residues and
the β-carbon in cross-linked cystine residues, (iii) a complex line shape in the 10-35 ppm region
due to alkyl components of the side-chains (Figure A.15a) and (iv) a carbonyl region with a
maximum at 173 ppm (Figure A.15b). The assignment was made following the 13
C isotropic
chemical shifts for the common amino acid residues as reported in ref. 19.
200 150 100 50 0
13C CP MAS
hard -keratin
T = 25 °C*
*
* *****
**
**
[ppm]
Figure A.14 13
C CPMAS spectrum of hard α-keratin at a rotor frequency of 5 kHz where the
spinning sidebands are marked by stars
In Figure A.15a is a zoom of the alkyl and α-carbon regions. The α-carbon shows an
increase of intensity, expressed by the shoulder at around 64 ppm, noticeable for the hard α-
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
129
keratin subjected to oxidative treatment. The peak at 40 ppm is related to β-carbon in cross-
linked cystine residues. These participate to the -S-S- (disulfide) bonds between neighbouring
keratin molecules. The intermolecular disulfide links between cysteine residues confer some
degree of rigidity to the intermediate filaments in the amorphous matrix component of α-keratin.
The intensity and the broadening of the resonance at 40 ppm 13
C resonance does not change
significantly with the chemical treatment with the exception of the sample where the sulphur of
the broken S-S bond was acetylated for arresting its reactivity. This sample has a lower intensity
of 13
C resonance at 40 ppm compared to the other samples. Moreover, resulting reduced cysteine
residues would be expected to have a β-carbon resonance between 25 and 29 ppm that is indeed
revealed in Figure A.15a.
Figure A.15 Enlarged version of the alkyl (a) and α-carbon (b) regions of the 13
C CPMAS
spectra at the rotor frequency of 5 kHz for hard α-keratin, in the native state, after oxidative, and
reductive/oxidative treatments. The dashed line at 40 ppm marks the 13
C resonance of the
cysteine engaged in the disulfide links. The 13
C CPMAS spectrum of –S-S- bond free sample is
also shown
The 13
C CPMAS spectra from Figure A.15b show that the line shape of the carbonyl signal
has not changed significantly with chemical treatments. This indicates that the amino acid chains
(the keratin molecules) are not significantly damaged (hydrolysed) by our treatments. A larger
degree of disorder of the intermediate filaments would lead to a broadening of the resonances.
This disorder will induce 13
C chemical shielding tensors with different orientations of the
75 50 25 0 200 150 100
hard a - keratin oxidative reductive & oxidative S - S free
T= 25°C
[ppm] [ppm]
a) b) 13
C CPMAS
Appendix A
130
principal reference frames relative to the laboratory frame leading finally to the line broadening.
The broadening of the 13
C carbonyl resonance would suggest a shift in the conformation of the α-
helix components. For the investigated chemical treatments the amino acid residue composition
does not change and therefore, the lack of the changing in the carbonyl line shape shows
basically the same conformation of the hard α-keratin. This does not stand for the case of hard α-
keratin with acetylated sulphur where the line width of the carbonyl signal is slightly larger than
the others, suggesting that the acetylation of sulphur after breaking the disulphide bond induced a
certain degree of disorder in the hard α-keratin. This supports our view of a three-phase model
for the hard α-keratins, where the interface, mainly cystine based, scaffolds the intermediate
filaments. The breaking of the cystine and the arresting of the reactive formed thiols by
acetylation fragments the scaffold and deprives the intermediate filaments of their mechanical
support.
Thermally Polarized and Laser Hyperpolarized 129
Xe Spectra of Chemically Treated Hard
α-Keratin. The 129
Xe spectra of thermally polarized xenon at room temperature and p = 20 bar
for hard α-keratin, in the native state, after oxidative, and reductive/oxidative treatments are
shown in Figure A.16a. The sharp weak signal at 40 ppm is a spectrometer artifact. The base of
the free gas resonance set as reference at 0 ppm shows an asymmetry.
Figure A.16 Thermally polarized (a) and laser hyperpolarized (b) 129
Xe NMR spectra for hard α-
keratin, in the native state, after oxidative and reductive/oxidative treatments
hard α -keratin
oxidative
reductive & oxidative hard α-keratin
oxidative
reductive & oxidative
b) Hyperpolarised 129
Xe a) Thermallypolarised 129
Xe
[ppm] [ppm]
200 100 0 15 10 5 0 - 5
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
131
This distribution of the 129
Xe chemical shift in the range of 0–15 ppm reflects most
probably the xenon atoms trapped between the hair fibres and the defects and the scales at the
surface of the fibres. This effect is even better shown by the 129
Xe spectra measured using laser
hyperpolarized xenon (Figure A.16b). The hydrophobic xenon atoms do not penetrate the hard α-
keratin and oxidative treated fibres because no 129
Xe resonance is detected at larger values of
chemical shift. This is not the case for the reductive/oxidative treated fibres where a weak,
relatively broad resonance is present in the range 175–190 ppm. We could interpret this
resonance as due to the 129
Xe atoms trapped in the voids of the amorphous keratin.
The reductive/oxidative treatment produced hydrophobic voids in the keratin amorphous
matrix during breaking (reduction) and reformation (oxidation) of the disulfide bonds. The size
of these voids can be estimated using the chemical shift ( s ) of 129
Xe atoms originating from the
collisions between xenon atoms and the wall of the cavity. This is given by the relationship37,38
ppmd pore
s0145.05.0
912.49
(A.25)
valid for low xenon pressure. For the chemical shift around 180 ppm (Figure A.16a) from the
above equation, we yield pored 0.6 nm larger than 0.44 nm, the diameter of the xenon atoms.
The same result is obtained from the calibration curve presented in ref. 39. The size, which is a
little bit larger than the usual bonds (disulfide, ionic, hydrogen bonds) between chains in keratin,
is also small enough to speculate that the voids formed from washing out protein material are
from medulla. Therefore we believe that the voids appeared as the result of the incomplete re-
formation of the disulfide bonds which allowed the chains to arrange less compact than before.
The experiments with laser hyperpolarized 129
Xe does not allow us to investigate the bulk
of the hard α-keratin and the effects of the chemical treatments due to the lost of the xenon atom
polarization induced by the interaction with the matrix. As we already mentioned before, an
asymmetric NMR resonance is detected with a good signal-to-noise ratio. The hyperpolarized
129Xe atoms could be confined between the fibres and between the scales of the fibre surface.
The distribution of the chemical shift and its small values that is almost the same for all hard α-
keratin samples favours the detection of the xenon atoms trapped by the keratin fibres surface.
Morphological Changes Induced by Chemical and Thermal Treatments as Seen by DSC
and NMR Data. The model in Figure A.2a gives a simplified schema of a microfibril with
protofibrils showing the α-helical rods and the non-helical terminal domains projecting into the
interfilamentous space and linking with the matrix proteins through disulfide bonds. The
Appendix A
132
terminal domains contain, besides cystine, glycine, threonine, valine, alanine and serine, acidic
sites as glutamic and aspartic acid. This scaffolding structure at the IFs surface made by the side-
chain interactions that anchor the microfibrils to the matrix (interface phase) assists the thermal
stability and the primary control over the denaturation of helical structure of keratin materials
when heated. It has a protective role and the capacity to participate in the formation of a solid
interface.
The mechanism of thermal denaturation of keratins, as it has been described by ref. 36,
follows several steps. Beyond a certain temperature (the peak on DSC), the temperature rise
leads to the breaking of the scaffold structure of IFs. At that temperature, the IFs are in a
metastable state. The α-helix denaturates at around 80° C in soluble proteins and it is only the
interface that still keeps it organised. Once set free, the IFs (α-helices) denaturate. This involves
a transition from a relatively compact ordered structure to a more flexible, disorganized, opened
polypeptide chain. As the process of denaturation proceeds the protein molecules unfold and the
intern hydrophobic regions expose to the outside of the molecules. The hydrophobic groups in
water tend to cluster, leading to associations of molecules.
The chemical modifications we used for the keratin material were focused on attacking the
disulphide bonds. The oxidative modification aims at breaking the S-S bonds and oxidise them
into cysteic acid (see Figure A.17a). Under the reaction conditions not all the bonds will be
broken, but overall, it is expected that both the scaffold and the matrix are crumbled. As a result
the DSC peak corresponding to the denaturation of protein shifts towards lower temperature
(around 130°C, see Figure A.3) and the enthalpy decreases compared to the original keratin
material. The interface amount for oxidative modification is reduced as compared to the hard α-
keratin as shown in Figures A.5a and A.5b.
Moreover, the molecular dynamics of side chains are intermediate between that of hard α-
keratin and the sample subjected to the reductive/oxidative treatment (cf. Figure A.6). The rigid
domain thickness is not essentially affected by the oxidative treatment (Figure A.12).
The reductive/oxidative modification occurs in two steps. Firstly the S-S bonds are broken
by the action of the reductive reagent, thioglycolic acid (TGA), and then are reformed by the
oxidative reagent (see Figure A.17b). During this sequence of reactions, not all the bonds are
broken and not all the broken bonds reform; besides, not all the reformations occur at the same
places.
In other words we expect to reform the material but with a more hindered molecular
mobility of the interface and matrix as shown in Figure A.6 obtained from 1H NMR spectra
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
133
deconvolution. The rigid fraction increases slightly compared with hard α-keratin (Figure A.5a
and A.5c), but a reorganization process takes place that reduces the rigid domain sizes (Figure
A.12). Consequently, the DSC peak is recorded at a temperature between those of not-treated
and of oxidative-treated material and has also an intermediary value of the enthalpy (Figure A.3).
Figure A.17 Schematic representation of the intermediate filament (α-helices) imbedded in the
keratin amorphous matrix (blue islands). The chemical changes induce by the oxidative (a) and
reductive/oxidative (b) treatments (see text). RS-H is the abbreviation for thioglycolic acid
Appendix A
134
A.5. Conclusions
Proton, 13
C and 129
Xe NMR spectroscopy, and 1H spin diffusion were used for
characterization of phase composition, dynamics of amino acid side chains, domain sizes, the
presence of voids at the fibre surface, and in the bulk for hard α-keratin under various chemical
treatments and in a range of temperatures including the temperature of denaturation. Proton
NMR spin diffusion offers quantitative information about the side chain mobility heterogeneity
of the interfacial region. The side chain motions play a very important role in the mechanical
deformation of keratin.
These reported NMR results support the thermal denaturation pathway described above
according to which concomitantly with the collapse of the scaffolds, the α-helices go from a
relatively compact ordered structure to a more flexible, disorganized, open polypeptide chain.
This is shown by an increase of the mobile phase at the expense of rigid and interphase. Next,
the protein molecules unfold and the intern hydrophobic regions expose to the outside of the
molecules. The hydrophobic groups tend to cluster in the deuterated water, leading to
associations of molecules and rebuilding the amount of rigid phase from the mobile phase. The
interphase amount remains, however, at the same value, as no other reorganisation occurs.
A.6. References and notes
1. Popescu, C.; Höcker, H. Chem. Soc. Rev. 2007, 36, 1282 and references therein.
2. Er Rafik, M.; Doucet, J.; Briki, F. Biophys. J. 2004, 86, 3893.
3. Parry, D. A. D.; Fraser, R. D. B. Int. J Biol. Macromol. 1985, 7, 203–213.
4. Steinert, P. M.; Torkia, D. R.; Mack, J. W. In The Biology of Wool and Hair; Rogers, G. E.; Reis, P. J.;
Ward, K. A.; Marshall, R. C., Eds.; Chapman & Hall: London, 1989, p 157-167.
5. Parry, D. A. Adv. Protein. Chem. 2005, 70, 113-142.
6. Powell, B. C.; Rogers, G. E. In Formation and Structure of Human Hair; Jolles, P.; Zahn, H.; Höcker, H.,
Eds.; Birkhäuser Verlag: Basel, Switzerland, 1997, p 59-148.
7. Birbeck, M.S.C.; Mercer, E.H.; J. Biophys. Biochem. Cytol. 1957, 3, 203.
8. Wortmann, F. J.; Popescu, C.; Sendelbach, G. Biopolymers 2006, 83, 630-635.
9. Wortmann, F. J.; Popescu, C.; Sendelbach, G. Biopolymers 2008, 89(7), 600-605.
10. Miles, C.; Ghelashvili, M. Biophys. J. 1999, 76, 3243.
11. Haly, A.R.; Snaith, J.W.; Text. Res. J. 1967, 37, 898.
12. Leroy, M.; Cao, J.; Biopolymer 2005, 77, 38,
13. Budrugeac, P.; Miu, L.; Bocu, V.; Wortmann, F.-J.; Popescu, C. J. Thermal Anal. Cal. 2003, 72, 1057.
14. Budrugeac, P.; Miu, L.; Bocu, V.; Wortmann, F.-J.; Popescu, C. J. Thermal Anal. Cal. 2004, 77, 975.
15. Popescu, C.; Budrugeac, P.; Wortmann, F.-J.; Miu, L.; Demco, D.E.; Baias, M. Polym. Degrad. Stab.
2008, 93, 976.
Morphology and molecular mobility of fibrous hard α-keratins by 1H,
13C, and
129Xe NMR
135
16. Ernst, R.R.; Bodenhausen, G.; Wokaun, A. Principles of Nuclear Magnetic Resonance in One and Two
Dimensions, Oxford Univ. Press: Oxford, 1990.
17. Wüthrich, K., NMR of Proteins and Nucleic Acids, John Wiley & Sons, Inc., 1986.
18. Duer, M.J., Solid-State NMR: Theory and Applications, Blackwell Science Ltd.: Oxford, 2001.
19. Duer, M.J.; McDougal, N.; Murray, R.C. Phys. Chem. Chem. Phys. 2003, 5, 2894.
20. Mack, J.W.; Torchia, D.A.; Steinert, P.M. Biochemistry 1988, 27, 5418.
21. Perich, J.W.; Johns, R.B.; Thompson, A.R. Aust. J. Chem. 1995, 48, 1925.
22. Yoshimizu, H.; Ando, I. Macromolecules, 1990, 23, 2908.
23. Nishikawa, N.; Horiguchi, Y.; Asakura, T.; Ando, I.; Polymer 1999, 40, 2139.
24. Nishikawa, N.; Tanizawa, Y.; Tanaka, S.; Horiguchi, Y.; Asakura, T. Polymer 1998, 39, 3835.
25. Schmidt-Rohr, K.; Spiess, H.W. Multidimensional Solid-State NMR and Polymers, Academic Press:
London, 1994.
26. VanderHart, D.L.; McFadden, G.B.; Solid State Nucl. Magn. Reson. 1996, 7, 45.
27. Demco, D.E.; Johansson, A.; Tegenfeldt, J. Solid State Nucl. Magn. Reson. 1995, 4, 13.
28. Buda, A.; Demco, D.E.; Bertmer, M.; Blümich, B.; Litvinov, V.M.; Penning, J.-P. J. Phys. Chem. B 2003,
107, 5357.
29. Buda, A.; Demco, D.E.; Bertmer, M.; Blümich, B.; Litvinov, V.M.; Penning, J.-P. Chem. Phys. Chem.
2004, 5, 876.
30. Baias, M.; Demco, D.E.; Popescu, C.; Fechete, R.; Melian, C.; Blümich, B.; Möller, M. J. Phys. Chem. B
2009, 113(7), 2184.
31. Wang, J. J. Chem. Phys. 1996, 104, 4850.
32. Buda, A.; Demco, D.E.; Bertmer, M.; Blümich, B; Reining, B.; Keul, H.; Höcker, H. Solid State Nucl.
Magn. Reson. 2003, 24, 39.
33. Crank, J., The Mathematics of Diffusion, Clarendon Press: Oxford, 1975.
34. Goldman, M.; Shen, L. Phys. Rev. 1966, 144, 321.
35. Clauss, J.; Schmidt-Rohr, K.; Spiess, H.W. Acta Polymer 1993, 44, 1.
36. Istrate, D.; Popescu, C.; Möller, M. Macromol. Biosci. 2009, 9(8), 805-812.
37. Ito, T; Fraissard, J. J. Chem. Phys. 1982, 76, 5225.
38. Demarquay, J.; Fraissard, J. Chem. Phys. Lett. 1987, 136, 314.
39. Ripmeester, J.A.; Ratcliffe, C.I.; Tse, J.S. J. Chem. Soc. Faraday Trans. 1988, 84, 3731.
Appendix B: Nonisothermal kinetics of chemically
damaged hard α-keratin thermal denaturation
B.1. Introduction
In a recent study1 the differential scanning calorimetry (DSC) measurements carried out at
different heating rates were used for the kinetic analysis of the endothermic process assigned to
the denaturation of the helical material from native (untreated) human hair in water excess. We
found that the kinetic mechanism is autocatalytic-like and that the value of the activation energy
is rather close to disulphide bond scission than to protein denaturation. This allowed us
proposing a multistep mechanism for the thermal denaturation pathway of hard α-keratins in
water excess that relies on the 3-phase model2. This describes the fibrous hard α-keratins
structure in terms of an interface phase that scaffolds the intermediate filaments and controls
their thermal stability. The limiting step of the thermal denaturation process was found to be the
scission of disulphide bonds between the main morphological components, namely intermediate
filaments (IF) and matrix (IFAP). The theoretical proposed model has been shown to be in good
agreement with the experimental recorded.
DSC analysis of structural changes in hard alpha-keratin fibres as a result of chemical
treatments (i.e. bleaching, permanent waving and oxidative dyeing), corroborated with XRD,
chemical and mechanical data demonstrated that cosmetic treatments, using mild action time and
concentrations, affect primarily the amorphous area (matrix and interface) of the hair structure,
while the crystalline zones remain virtually unchanged3. It was shown that the changes recorded
in the variation of the experimental DSC parameters (i.e. peak temperature, Tp, and enthalpy,
ΔH) are more likely to occur as a consequence of modifying the immediate environment of the
intermediate filaments (interface phase) rather than due to a significant loss of the secondary
structure of keratin protein, that is of the α-helical material.
For this stage of the investigation, to further our understanding the treatment-specific
effects, certain information about the denaturation mechanism of previously chemically damaged
hair material by oxidative and reductive treatments and its activation energy were searched by
means of different methods of non-isothermal solid state reactions kinetics.
Appendix B
138
The nonisothermal kinetic analysis allows acquiring further insights into the process of
thermal denaturation of the hard -keratins and the changes due to chemical processing.
B.2. Material and methods
The alpha-keratin fibres used for analysis were of Caucasian dark-brown hair, supplied by
KERLING International Haarfabrik GmbH. The fibres were cleaned with 1% Lauryl ether sulphate
(LES) and dried at room temperature prior to work with them. The pH of their aqueous extract was
found to be 6.5 to 7.
DSC measurements
Prior to the measurements the samples were cut into fine snippets (~2mm) and stored
under controlled conditions (~ 24 hours, 22°C, 55% relative humidity) to ensure invariant water
contents. 7…10mg of each sample snippets were weighted and placed in crucibles.
Before sealing a crucible, 50 μL of distilled water (pH 6.7) was added, and the sealed
crucible was stored over night (~14 hours preceding the measurement), to allow the snippets to
wet.
The DSC experiments were run in a DSC-7 Perkin Elmer, using pressure resistant stainless
steel large volume capsules. DSC calibration was done with indium and palmitic acid, both of
high purity. We used five heating rates, viz.: 5, 7.5, 10, 15 and 20 K/min for temperature ranging
from 80 to 180°C. Each experiment was repeated three to five times, for ensuring the
reproducibility of data.
Damaging treatment
Bleaching, perm-waving, dyeing and pH treatments were used for achieving controlled
modification of the fibres.
Bleaching treatment was done with IGORA VARIO BLOND PLUS bleaching powder and
IGORA ROYAL 20 vol 6% H2O2 bleaching lotion, a commercial products kindly supplied by
Schwarzkopf.
The bleaching procedure followed the instructions of use, being applied for 35 minutes at
room temperature. A bleaching cycle implies the treatment of 1g hair sample with a mixture (pH
10) made-up of 0.6 g bleaching powder and 1.2 ml of bleaching lotion containing 6% H2O2.
Afterwards, the fibres were rinsed under tap water for 3 minutes, 2 times subsequently washed
with a solution of Texapon N70, 0.1 mL/L, (70% Natrium Laurethsulfat, pH~7), warm water 1
minute, tap water 3 minutes and dried under hot air blow. The pH of the aqueous extract of the
fibre was checked to be 7. The process has been resumed up to seven times (at intervals of 24
Nonisothermal kinetics of chemically damaged hard α-keratin thermal denaturation
139
hours) on the same hair sample, fibres for analysis being sampled after each complete bleaching
cycle.
Permanent waving treatment was performed with the commercial product POLY LOCK-
PERMANENTE FORTE kindly supplied by Schwarzkopf.
A perm-waving cycle consists of immersion of wetted hair tresses in the reduction solution
(pH 8.5-9; liquor ratio of 1.2 g hair to 1 mL solution). The tresses are then covered with plastic
folia and let to react for 30 minutes at room temperature. After rinsing 3 minutes with tap water
the tresses are immersed in the oxidation lotion (pH 4.5) using similar conditions as for the
reductive process. Processing time is of 10 minutes at room temperature, according to the
product recommendation. The fibres were then rinsed with tap water for 3 minutes, 2 times
subsequently washed with a solution of Texapon N70, 0.1 mL/L, (70% Natrium Laurethsulfat,
pH~7), rinsed with warm water 1 minute, tap water 3 minutes and dried under hot air blow. The
pH of the aqueous extract was 7.
The process was repeated up to 7 times on the same sample, hair fibres for analysis being
sampled after each complete perm-waving cycle.
B.3. Results and discussions
For oxidative (i.e. bleaching) treated hair, the course of denaturation was investigated by
kinetic analysis of DSC-curves4,5
. Oxidation was chosen since it represents comparatively
straightforward case, where the treatment affects both morphological components (i.e. IF and
IFAP) to very similar extents5. Using curves recorded at one heating rate, it was found that the
denaturation process remain largely unchanged after oxidation and that the reaction rate constant
at denaturation temperature increases with cumulative chemical modifications. It was concluded
that the kinetic hindrance of the unfolding of the α-helix by the matrix in the IF/IFAP-composite
is the primary controlling mechanism of the onset of the denaturation process. Once the
temperature rise, in combination with the natural composition or the chemical change, induce a
suitable drop of the viscosity of the matrix around the IFs, their denaturation occurring along a
pathway independent of temperature and treatment history. This emphasized the kinetic control
of the matrix over the denaturation process of the helical segments in the filament/matrix
composite4.
The effects of the reduction (i.e. permanent waving treatments) on the denaturation kinetics
of human hair were also subject of recent investigations6. Generally, similar considerations as in
case of the oxidative treatments were drawn.
Appendix B
140
The chemical modifications we used for the keratin material were focused on attacking the
disulphide bonds. The oxidative modification aims at breaking the S-S bonds and oxidising them
to cysteic acid. Under the reaction conditions not all the disulphide bonds are broken but, overall,
it is expected that both the scaffold (i.e. interface) and the matrix are crumbled. As a result the
DSC peak corresponding to the denaturation of protein shifts towards lower temperature (around
125°C, see Table B.1) and the enthalpy decreases compared to those of the original keratin
material3.
The reductive/oxidative modification occurs in two steps. Firstly the S-S bonds are broken
by the action of the reductive reagent, thioglycolic acid (TGA) and then are reformed by the
oxidative reagent. During this sequence of reactions not all the bonds are broken and not all the
broken bonds reform; besides, not all the reformations occur at the same places3,7
.
The activation energy calculated for conversion degrees ranging from 0.1 to 0.9, according
to Friedman´s method (see chapter IV), shows the variation given in Figure B.1.
80
100
120
140
160
180
200
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Act
ivat
ion
ener
gy,
Ea
(kJ/
mol)
Conversion degree, α
Bleached 3x Bleached 7x
80
100
120
140
160
180
200
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Act
ivat
ion
en
ergy
, Ea
(kJ/
mo
l)
Conversion degree,α
Perm-waved 3x Perm-waved 7x
Figure B.1 The dependence of the activation energy Ea on the conversion degree α as
determined by Friedman method for bleached and permanent-waved Caucasian hair samples
Nonisothermal kinetics of chemically damaged hard α-keratin thermal denaturation
141
In view of the standard deviations, the variation of the activation energy is not very
pronounced (~ 12% and 7% for bleached 3, respectively 7 times hair material, while for the
permanent waving a variation of approximately 8% was recorded independent of the intensity of
the treatment). Despite its small variation, similar to those observed at the denaturation of
collagen8, Figure B.1 shows a statistically significant dependency of the effective activation
energy on the extent of conversion. Revealing the dependence of the activation energy on
conversion degree may well help to disclose the complexity of a process and to identify its
kinetic scheme8-10
. An increase of the activation energy with conversion generally applies for the
thermal decomposition of many polymers through competing, consecutive, although some
independent reactions9. The decrease of Ea on α may correspond to the kinetic scheme of an
endothermic reversible reaction followed by an irreversible one 8,9,11
. Such a behaviour is also
reported for processes which proceeds with a change from a kinetic to a diffusional regime9. One
may easily notice the similarity with the behaviour of native hair material already reported1.
The overall kinetic parameters are inferred and summarised in Table B.1, the pre-
exponential factor being given in its usual form as ln(A).
Focusing on the values recorded for the activation energies one may notice that remain
generally quite low, close to the lower limits of the range of activation energy generally
associated with protein denaturation (i.e.104.6-836.8 kJ/mol)12
.
The kinetic function suggests as well autocatalytic-like processes (Table B.2). It has to be
underlined that the analytical form of this function, as well as the values of the two parameters,
should be regarded merely as fitting parameters than related to a certain mechanism, unless
further evidence is obtained.
Sample Tp10°C/min
(°C)
ΔH10°C/min
(J/g)
Ea ± St.dev
(kJ mol-1
)
lnA ± Stdev
(min-1
)
Native 151.6 ± 0.8 14.3 ± 0.1 118.77 ± 13.89 30.91 ± 0.07
Bleached 3x 128.4 ± 0.7 9.6 ± 0.4 158.13 ± 19.11 44.54 ± 0.06
Bleached 7x 125.4 ± 1.1 10.7 ± 0.3 142.95 ± 11.12 40.18 ± 0.07
Perm-waved 3x 145.4 ± 1.0 10.9 ± 0.7 124.94 ± 10.19 33.19 ± 0.08
Perm-waved 7x 138.3 ± 1.0 7.8 ± 0.7 149.06 ± 12.89 40.64 ± 0.08
Table B.1 Activation energies, Ea and pre-exponential factors, as ln(A), calculated by
Friedman´s method for Caucasian hairs
Appendix B
142
Summing up, the thermal denaturation pathway of previously chemically damaged hair
material seems to be independent of temperature and treatment history, in good agreement with
the aforementioned results4,6
.
Sample f(α)
Native α2/3
(1-α)1
Bleached 3x α1/2
(1-α)3/4
Bleached 7x α1/2
(1-α) 3/4
Perm-waved 3x α2/3
(1-α)1
Perm-waved 7x α1/2
(1-α)1
Table B.2 Kinetic function f(α) for chemically damaged Caucasian hairs
The results support our view regarding the denaturation pathway schematically represented
in Figure B.2, respectively the way of attack of chemical reagents on the keratins morphology.
When temperature rises the helical domains from the IFs try to unfold. Unfolding inevitably
involves a transition from a relatively compact ordered structure to a more flexible, disorganized,
open polypeptide chain. As the process of denaturation proceeds, the protein molecule unfolds
and the internally directed hydrophobic regions become exposed to the outside of the molecule.
Non-polar, hydrophobic groups in water will tend to cluster together because of their mutual
repulsion from water, not necessarily because they have any particular direct affinity for each
other. Therefore, upon unfolding hydrophobic regions on individual protein molecules will try to
associate with hydrophobic regions on other protein molecules.
This is only possible after the polypeptide chains are set free from the proposed scaffolding
structure. Since the covalent S-S bonds control the strength of this interface, the rate determining
step of the process is their scission. Once set free, the helical material will proceed to unfold and
consequently the hydrophobic interactions will play further a significant role for the
irreversibility of the denaturation process.
The dependency of the Ea on conversion validates our hypothesis1: decomposition
(cleavage of cystine through a homolytic fission of some C-S bonds) followed by the unfolding
of the helical material (reversible reaction) and the irreversible denaturation of the IFs structure.
One may easily notice from Figure B.1 that the decomposition step is virtually absent for harsh
reductive treatments (i.e. permanent waved 7 times hair material). The mechanism of permanent
Nonisothermal kinetics of chemically damaged hard α-keratin thermal denaturation
143
waving consists, as above mentioned, in breaking and reforming of S-S bonds in new positions
after having arranged the hair fiber in another shape. It is expected that the re-oxidation of the
disulphide bond in keratin protein does not reform all the bridges and the link between IFs and
IFAPs is weakened. According to Kuzuhara‘s work 13
the broken disulphide groups existing in
the matrix as a result of the reduction process do not return to the original conformation after the
oxidation process. Therefore, as shown additionally by the large decrease of the value of
enthalpy (Table B.1) in the case of hair material subjected to a harsh permanent-waving process,
the interface has its minimum number of disulphide cross-links (if any). The interface being
destroyed, the protective role of the matrix disappears and the secondary structure (α-helix) from
the exposed areas of the IFs proceeds easily to unfold as a result of high temperature.
Figure B.2. The denaturation pathway of one residue of helical material from the intermediate
filaments. The helical rod is flanked by nonhelical head and tail domains at the NH2- and
COOH- termini that extend into the matrix through cystine bonds and link with the matrix
proteins. Together with other linkages, the interface, as the scaffolding structure at the surface of
IFs, controls and enhances the thermal properties of keratin filaments
The autocatalytic-like character suggested by the kinetic function f(α) is due to the nascent
Appendix B
144
sulphur compounds from the cystine degradation by α- or β-elimination1.
As previously mentioned, the changes of the activation energies are not pronounced,
independent of the treatment applied in good agreement with studies of the effects of oxidation
on nonisothermal denaturation kinetics of human hair that reports low variation of activation
energies after multiple time bleached samples. The slight increase of the activation energies after
chemical treatments may be the consequence of the involvement of distinct intermediates in the
denaturation process of hard alpha-keratins as result of severe chemical treatments that were
suggested to account for the variations recorded in the inter-microfibrilar distances and IFs
radius3. These intermediates may be similar to molten or premolten globules observed in the
transient intermediate states found during the folding of certain proteins, especially globular
proteins that undergo hydrophobic collapse. The molten globules are described as compact
intermediate protein conformations that generally preserve the native-like secondary structure
but have a poorly defined and dynamic tertiary structure 14-16
. It is possible that in the
―intermediate‖ to denaturated state transition the hydrophobic interactions in the protein interior
resist this disruption adding their influence to the initiation of the process. This approach is also
consistent with the kinetic mechanism proposed for describing the thermal denaturation pathway
of hard alpha-keratin1.
B.4. Conclusions
The differential scanning calorimetry measurements carried out at different heating rates
were used for the kinetic analysis of the endothermic process assigned to the denaturation of the
helical material from previously chemically damaged human hair by oxidative and reductive
processes. We found that despite the fact that pronounced decreases of denaturation temperature
as well as of enthalpy occur, the kinetic parameters of the denaturation process remain largely
unchanged, independent of the treatment applied.
The kinetic mechanism is autocatalytic-like and the value of the activation energy is rather
close to disulphide bond scission than to protein denaturation. This result is in line with our
previously proposed multistep mechanism for the thermal denaturation of hard α-keratins in
water excess that relies on a 3-phase model which describes their structure1-3
. The transient
intermediate states of the denaturation process of hard alpha-keratins which result because of
chemical treatments and were suggested by the variations recorded in the inter-microfibrilar
distances and IFs radius3, are assumed to account also for the slight increase of the activation
energy of the process recorded under the endothermal peak, as a result of chemical damaging.
Nonisothermal kinetics of chemically damaged hard α-keratin thermal denaturation
145
B.5. References and notes
1. Istrate, D.; Popescu, C.; Möller, M. Macromol. Biosci. 2009, 9(8), 805-812.
2. Istrate, D.; Popescu, C.; Er Rafik, M.; Möller, M. Polym. Degrad. Stabil. submitted 2010.
3. Istrate, D.; Popescu, C.; Er Rafik, M.; Möller, M. J. Soc. Cosmet. Chem. submitted 2009.
4. Wortmann, F. J.; Popescu, C.; Sendelbach, G. Biopolymers 2006, 83, 630-635.
5. Wortmann, F. J.; Sendelbach, G.; Popescu, C. J. Cosmet. Sci. 2007, 58, 311-317.
6. Wortmann, F. J.; Popescu, C.; Sendelbach, G. Biopolymers 2008, 89(7), 600-605.
7. Baias, M.; Demco, D. E.; Istrate, D.; Popescu, C.; Blümich, B.; Möller, M. J. Phys. Chem. B. 2009,
113(35) 12136–12147.
8. Vyazovkin, S.; Vincent, L.; Sbirrazzuoli, N. Macromol. Biosci. 2007, 7, 1181-1186.
9. Vyazovkin, S.; Wight, C. A. Annu Rev. Phys. Chem. 1997, 48, 125-149.
10. Vyazovkin, S. V.; Lesnikovich, A. I. Thermochim. Acta 1990, 165, 273-280.
11. Vyazovkin, S.; Linert, W. Int. J. Chem. Kinet. 1995, 27, 73-84.
12. Bischof, J. C.; He, X. Ann. NY Acad. Sci. 2005, 1066, 1-22.
13. Kuzuhara, A. Biopolymers 2007, 85, 274-283.
14. Fink, A. L.; Calciano, L. J.; Goto, Y.; Nishimura, M.; Swedberg, S. A. Protein Sci. 1993, 2, 1155.
15. Lala, A. K.; Kaul, P. J. Biol. Chem. 1992, 267, 19914-19918.
16. Wikipedia. Wikimedia Foundation, Inc, 2008.
Appendix C: Factors influencing the DSC
thermogram of hard alpha-keratin proteins and the
reproducibility of the experimental results
C.1. Introduction
To check the efficiency of DSC in water excess to reflect and adequately quantify the
effects of various chemical treatments on fibrous keratin proteins, additional investigations are
required to facilitate the removal of possible interferences. These interferences may be related
with factors depending on the DSC methodology as such or may relate to morphological
subcomponents of fibrous proteins that could undergo thermal destruction in parallel with the
effect of interest (the unfolding of the alpha-helical material). Consequently some emphasis in
this study was given to additional experiments meant to reveal the accuracy and the correctness
of the recorded data.
C.2. Factors relating to the DSC methodology
C.2.1. Instrument baseline
For differential scanning calorimetry it is recommended that one scan the analyzer before
analyzing samples under the conditions that will be used further for samples in order to check the
baseline curvature and noise level. This is done by placing empty sample pans in the sample and
reference holders and performing a run using the method that will be used for the samples. The
baseline subtraction process, automatically performed by the Pyris Software for Windows-
v.3.80, offers the advantage of avoiding any additional errors that could be introduced if the
instrument's baseline is curved or too noisy.
C.2.2. Analysis of the experimental thermograms
The type of baseline to be used in the calculation of the peaks characteristics was chosen to
be ―Standard‖ for all the calculations. The ―Standard‖ option allows drawing a baseline parallel
to the X axis, within the limits selected.
Appendix C
148
C.2.3. Sample pans and crucibles
The form and dimensions of the capsules used for subjecting the sample to a controlled
heating program, may as well induce errors to experimental data if varies during the
experiments. Pressure resistant stainless steel, large volume capsules (Art.:0319-0218) with the
following characteristics were used within this work:
Material
• Capsule: 0.178mm (0.007"); corrosion resistant, stainless steel; 0-ring: Viton Rubber
• Capacity: 60 microliters
Dimensions
• 7.54 mm Diameter, 2.79 mm Height, 0.33 g Weight
Temperature range
• -40°C (limited by O-ring glass transition) …+300°C (or temperature of pressure limit)
Pressure limit
• 350 psi (~ 24 bar); or equilibrium water vapor pressure at 225°C (assumes proper sealing)
The DSC calibration was done with indium (melting point Tm= 156.60°C, enthalpy ΔH=
28.45J/g) and palmitic acid (melting point Tm= 63°C), both of high purity.
C.2.4. Pressure influence
The pressure developed in the pan is not well controllable. It may be estimated from the
Pascal's law that in the area of interest (130-160°C) the pressure ranges from 3 to 9 bar. This can
induce an estimated error of 1-2 °C to peak position 1,2
.
C.2.5. pH influence
The pH of the thermal medium in which the protein samples are controlled heated was
shown to significantly influence the DSC thermogram, as discussed in Chapter III.
Consequently, if not otherwise specified, the pH of the thermal medium (i.e. distilled water) was
checked to be 7.
C.3. Factors relating to fibrous protein structure
C.3.1. Cortex-cuticle assembly
Cuticle isolation: Approximately 4-5 g of hair fibres were cut in 3-5 mm snippets and
swelled over night in a tumbler with 200 ml water. Next day, the fibre-water solution was
Factors influencing the DSC thermogram of hard alpha-keratin proteins and the reproducibility of the experimental
149
transferred in a mixer and hackled 10 times, 1 minute each, between two mixing steps the system
being cooled down in an ice bath. The dispersion of water-cuticle was further separated by the
remaining fibers on a suction filter and centrifuged twice, 30 min 12000rpm. Before analysis, the
cuticle residue was dried over night in an exsiccator.
Cortex isolation: ~ 100 mg of hair fibres (Caucasian brown hair cut in snippets of ~ 0.5 - 1
cm) + 2.5 g corundum were weighted in four 50 ml plastic bottles with a screw cap. A few
thymol crystals were added to prevent bacterial growth during the experiment and the bottles
filled up with distilled water (pH ~7). The bottles were clamped into a high-frequency ellipsoid
shaker. Shaking proceeded in 16 steps, one our each. After each shaking step the system is
cooled down in an ice bath for 1 hour to avoid heating of the sample. The suspension it is then
passed through metal sieves: the cuticle and the corundum passed the sieves, leaving the stripped
fiber snippets that are the cortex. The cortex snippets were washed repeatedly with distilled
water and dried in air. After drying the progress of cuticle removal is checked by SEM.
Figure C.1 DSC traces in water excess of mechanically isolated a)-cortex and b)-cuticle
The thermal behaviour of the cuticle observed when heating hair material in perforated
crucibles is similar to those of using DSC in water excess. The lack of any thermal effect on
DSC trace of mechanically isolated cuticle is expected in view of the knowledge that proteins of
cuticle are of a predominantly amorphous in nature3. This also indicates that the endothermal
effect is due to the cortical cells only. The result is in fairly good agreement with the current
understanding that relate the endothermal with the denaturation of the helical material from the
intermediate filaments, which together with the matrix material are the basic constituents of the
Appendix C
150
cortical cells.
C.3.2. Melanin pigment
The colour of the human hair is due to the melanin granules in the hair fibre included
during keratinisation. The aim of bleaching treatments is to eliminate or tone down the hair
colour, this being accomplished by oxidation. During bleaching the melanin pigment undergoes
irreversible physicochemical changes and the colour of the hair fibres is modified, but the fact
that the pigment granules are distributed within the cortex of the fibre also leads to the oxidation
of the keratin matrix4. The principal oxidising agent used in bleaching composition is hydrogen
peroxide, and salts of persulphate are often added as ―accelerators‖.
The reactivity of melanin towards hydrogen peroxide is much higher than that of keratin;
although the amount of melanin is usually 2%, it is important to note if the pigment has any
influence on the DSC endotherms exhibited when heating hair samples in water excess in order
to be able to discriminate the effects of the cosmetic treatments on the main morphological
components.
14
7.9
14
8.4
14
5.4
14
5.5
14
7.3
14
7.7
14
5.3
14
5.0
100
110
120
130
140
150
Untreated 1x 2x 3x
Tem
pera
ture
(°C
)
14
.4
10
.2
10
.6
10
.4
14
.1
10
.9
9.2
9.6
0
2
4
6
8
10
12
14
16
Untreated 1x 2x 3x
En
thal
py
(J/
g)
Figure C.2 Denaturation temperatures (top) and enthalpies (bottom) recorded for pigmented
(grey bars) vs. unpigmented (white bars) hair material. 1x, 2x, 3x refers to the number of steps
Factors influencing the DSC thermogram of hard alpha-keratin proteins and the reproducibility of the experimental
151
from the bleaching treatment
Pigmented and unpigmented human hair fibres, sampled from the same head, were
thermally investigated, and respectively subjected to bleaching multi treatment- 3 subsequent
steps (Figure C.2).
It can be easily noticed that the melanin pigment presence do not influence significantly
neither the peak position, nor the enthalpy recorded for any of the samples. We therefore exclude
a possible contribution of melanin destruction to the denaturation process of the helical material.
C.3.3. Ethnic differences
The ethnic differences are important for cosmetic industry some treatments being
especially designed for one of the three major racial types of hair: Afro, Asian and Caucasian
hair respectively. The distinction between these hair types is particularly related to diameter,
geometry, crimp and colour. These differences are known to have an influence on the degree of
change and damage after a treatment.
Figure C.3 DSC traces of Caucasian (a) and Asian (b) hair material
Figure C.3 illustrate the major differences that allow discriminating between Caucasian
and Asian hair material by DSC in water excess. Generally, we noticed a difference in peak
position of 4-5°C between the two types of hairs, and almost identical enthalpy values.
Despite the initial distinction in the peak position for native samples, Asian hair behaviour was
Appendix C
152
found similar with those of Caucasian hair when chemically treated (Figure C.4).
Figure C.4 3D representation of the DSC parameters variation for Caucasian (full symbols) and
Asian (empty symbols) hairs as a result of persulphate bleaching (circle), permanent waving
(square) and dyeing (triangle) treatments; 0 refers to native hair material; a line was drawn
between the points as an eye indication
C.4. Reproducibility of the experimental results
The reproducibility (Figure C.5) of the DSC analysis method is evaluated using duplicate
determinations. 30 days after the original experiments, treatments using the same formulations
and conditions were repeated. The DSC method shows a good precision in terms of Tp and ΔH
variation.
Factors influencing the DSC thermogram of hard alpha-keratin proteins and the reproducibility of the experimental
153
Figure C.5 Reproducibility (empty symbols) of the DSC parameters variation as a result of
bleaching (circle), permanent waving (square) and dyeing (triangle) treatments
C.5. References and notes
1. Popescu, C., personal communication.
2. Wortmann, F. J.; Deutz, H. J. Appl. Polym. Sci. 1993, 48, 137-150.
3. Bradbury, J. H.; Chapman, G. V.; Hambly, A. N.; King, N. L. R. Nature 1966, 210, 1333-1334.
4. Wolfram, L. J.; Hall, K.; Hui, I. J. Soc. Cosmet. Chem. 1970, 21, 875-900.
General Conclusions
The results reported in this work are related to the thermal behaviour of fibrous proteins
encapsulated in rigid structures, among the most well-known representatives of this class being
the α-keratins in human hair.
The assessment of the damages induced by cosmetic processes is the core of the claim support in
cosmetic industry. Human hair is a reactive substrate whose structure and physico-chemical
properties are of interest in relation to environmental factors and chemical reagents applied to it.
Consequently, many studies investigate ways to evaluate the degree of hair damage or to develop
anti-damaging products. Among them, thermal analysis has been developed as a powerful
analytical tool able to reveal the morphological transitions of the fibre with respect to the
treatments applied.
Our results strongly suggest the need for a careful interpretation of the DSC results within the
frame of formulations. The key findings of this work are summarised below, underlining the
advantages and the limitations of classical DSC to properly reflect and quantify the existing
condition of the hair.
1. We showed that the DSC of keratin fibres under a gaseous draft (i.e. in open pan)
supplies misleading information, due to the interference of pyrolysis with the process of
interest. For the scope of acquiring information about the influence of a treatment on
keratin fibre the DSC experiments must be conducted in an aqueous surrounding that do
not allow water to evaporate.
2. The endothermal effect recorded by DSC on keratin fibres relates to the cortical cells
only and the peak temperature and the enthalpy of the process offers hints about the state
of alpha-helix. This means that DSC cannot be of use for assessing the effects of
treatments which do not affect the hair cortex.
3. The presence of melanin pigments in cortical cells does not influence significantly neither
the peak temperature, nor the enthalpy, i.e. the two parameters that are experimentally
directly accessible.
General conclusions
156
4. We show that DSC allows discriminating among major racial types of hair. This is to say
that the value of either the peak temperature, or of the enthalpy, or of both, differ for hair
of Caucasian, African or Asian sources.
5. Based on a kinetic study, I proposed a mechanism for describing the endothermal effect
recorded by DSC. It accounts on how thermal denaturation process occurs in hard α-
keratins. The mechanism comprises a sequence of reactions and has a self-catalytic
nature being more complex than the first-order kinetics used as a first approximation in
literature. According to the proposed mechanism the changes of the recorded DSC
parameters are more likely to occur as a consequence of modifying the immediate
environment of the intermediate filaments (interface phase) rather than due to a
significant loss of the secondary structure of keratin protein.
6. The pH of the thermal medium in which the hair samples are controlled heated was
shown to significantly influence the DSC results, in accordance with the mechanism
proposed.
7. The study of the influence of pH, particularly strong acid values, on the thermal
behaviour of hard alpha-keratins, indicates limits of the two-phase model used so far to
describe the fibrous alpha-keratins. We propose a three-phase model for explaining the
high thermal stability of fibrous hard alpha-keratins and their response to chemical
treatments. The third phase, the interface between crystalline and matrix phases, made of
nonhelical tail domains of keratin, scaffolds the intermediate filaments and controls their
interaction with chemical reagents as well as their thermal properties.
8. Despite pronounced decreases of peak temperature, as well as of enthalpy, which may
occur, the kinetic parameters of the alpha-helix thermal denaturation process and its
pathway remain virtually unchanged, independently of the treatment applied.
9. The work indicates strongly that the DSC parameters should be evaluated carefully,
always within the context of the cosmetic formulations and against reference of similar
origin. In all the cases, for a better understanding and assessment of the treatment effect,
one should consider supplementary methods like amino-acid analysis and tensile
measurements.
About Me
Personal Information
Daniel Istrate
Date of birth: 10.10.1978
Place of birth: Brasov, Romania
Family status: Married
Work experience, Education and Training
Jannuary 2011 to present: Scientist Morphology- Thermal Analysis, DSM
Resolve, The Netherlands
February 2005 to December 2009: PhD student @ DWI Interactive Materials
Research, RWTH Aachen, Germany
April 2004 to January 2005: Production Manager @ ―S.C. Textila Unirea S.A.‖,
Bucharest, Romania
July 2003 to April 2004: Engineer @ ―S.C. Textila Unirea S.A.‖, Bucharest,
Romania
April 2004 to June 2004: Training @ DWI Interactive Materials Research, RWTH
Aachen, Germany
September 1998 to July 2003: Diplomat Engineer @ Faculty of Textiles and
Leather Engineering, Textile Chemical Technology Department, University
―Gh.Asachi‖, Iasi, Romania