heat induced denaturation of fibrous hard

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


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


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


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.


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.


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.


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)


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.


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


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


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.


1. Popescu, C.; Höcker, H. Chemical Society Reviews 2007, 36, 1282-1291.


3. Istrate, D. Unpublished results



6. Spei, M.; Holzem, R. Melliand Textilber 1989, 70, 786-787.


8. Hocker, H. In Wool: science and technology; Simpson, W.; Crawshaw, G., Eds.; Woodhead Publishing:

Cambridge, UK, 2002, p 60-79.


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.


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


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.


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


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

.


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


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

.


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


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


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,


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.


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.


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.


15. Hearle, J. W. S.; Susutoglu, M. 7th Int. Wool Text. Res. Conf., Tokyo, 1985, pp 214-223.


17. Feughelman, M. J. Macromol. Sci. B 1979, 16, 155-162.


19. Feughelman, M. J. Appl. Polym. Sci. 2002, 83, 489-507.

20. Feughelman, M.; Haly, A. R. Kolloid Z. 1960, 168, 107-115.


22. Privalov, P. L.; Potekhin, S. A. Methods Enzymol. 1986, 131, 4-51.










217–270.



33. Schwenker, R. F.; Dusenbury, J. H. Text. Res. J. 1960, 30, 800-801.


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.



39. Popescu, C.; Wortmann, F. J. J. Therm. Anal. Cal. submitted 2007.


41. Wortmann, F. J.; De Jong, S. Text. Res. J. 1985, 55, 750-756.



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.





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.



68. Friedman, H. L. Polym. Lett. 1966, 4, 323–328.


70. Vasile, C.; Calugaru, E. M.; Stoleriu, A. S. M.; Mihai, E. Comportarea termica a polimerilor (Thermal

Behaviour of Polymers): Bucuresti, Romania, 1980.


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.


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.



Nonisothermal kinetics of hard α-keratin thermal denaturation

73


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


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.


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


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


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.








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


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.


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.



2. Popescu, C.; Sendelbach, G.; Wortmann, F. J. 10th Int. Wool Text. Res. Conf., Aachen, Germany, 2000.




6. Istrate, D.; Popescu, C.; Wortmann, F. J.; Möller, M. Biomacromol. submitted 2010.


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.


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.








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.



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.





40. Parry, D. A. Adv. Protein. Chem. 2005, 70, 113-142.


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.








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.



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 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.



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


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.


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.


93

The fibres were eventually dried under hot air blow.

We used acetic acid for adjusting pH from 1 to 7.


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


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


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


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


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


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)


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


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


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.


1. Alberts, B. Essential Cell Biology: An Introduction to the Molecular Biology of the Cell; Garland

Publishing: New York, 1998.


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.






8. Er Rafik, M.; Doucet, J.; Briki, F. Biophys. J. 2004, 86, 3893-3904.


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.


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.






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.



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.


37. Privalov, P. L.; Potekhin, S. A. Methods. Enzymol. 1986, 131, 4-51.









217–270.



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.


107





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.


* 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


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


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:


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


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-


13C, and

129Xe NMR

121

keratin.

120 130 140 150 160 170 18039.0

39.5

45

46

47

hard -keratin

oxidative treat.


linew

idth

[kH

z]

temperature [°C]

rigid

120 130 140 150 160 170 18015

16

17

19

20

21


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


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


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.


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.


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


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]


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.


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


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-


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


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


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.





Eds.; Birkhäuser Verlag: Basel, Switzerland, 1997, p 59-148.

7. Birbeck, M.S.C.; Mercer, E.H.; J. Biophys. Biochem. Cytol. 1957, 3, 203.


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.


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.


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


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


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




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.



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.


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


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


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.


145

B.5. References and notes



3. Istrate, D.; Popescu, C.; Er Rafik, M.; Möller, M. J. Soc. Cosmet. Chem. submitted 2009.


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.



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


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.


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.


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

heat induced denaturation of fibrous hard

Documents