In vitro mechanical characterization of
human skin layers: stratum corneum,
epidermis and hypodermis.
ii
Individual skin layer mechanics:
Stratum corneum, epidermis, and hypodermis
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de
Technische Universiteit Eindhoven, op gezag van de
Rector Magnificus, prof.dr.ir. C.J. van Duijn, voor een
commissie aangewezen door het College voor
Promoties in het openbaar te verdedigen
op donderdag 21 januari 2006 om 16.00 uur
door
Marion Geerligs
geboren te Hoogezand-Sappemeer
iv
Dit proefschrift (De documentatie van het proefontwerp) is goedgekeurd door de
promotoren:
prof.dr.ir. F.P.T. Baaijens
Copromotoren:
dr.ir. C.W.J. Oomens
en
dr.ir. G.W.M. Peters
Contents
Summary .............................................................................................. viii
1 General introduction........................................................................... 1
1.1 Introduction............................................................................................................... 2
1.2 A mechanical view of skin anatomy and physiology ............................................... 4
1.2.1 Skin relief........................................................................................................... 5
1.2.2 Stratum corneum................................................................................................ 5
1.2.3 Viable epidermis ................................................................................................ 6
1.2.4 Dermal-epidermal junction ................................................................................ 8
1.2.5 Dermis................................................................................................................ 8
1.2.6 Hypodermis........................................................................................................ 8
1.3 State-of-the-art in skin layer mechanics ................................................................. 10
1.3.1 In vivo vs in vitro experiments ........................................................................ 10
1.3.2 Mechanical behavior of the stratum corneum ................................................. 10
1.3.3 Mechanical behavior of the viable epidermis .................................................. 12
1.3.4 Hypodermis...................................................................................................... 12
1.4 Aim and Outline ..................................................................................................... 13
2 Isolation and preservation methods for the epidermis and stratum
corneum ............................................................................................. 15
2.1 Introduction............................................................................................................. 16
2.2 Skin preparation and analyses ................................................................................ 18
2.2.1 Skin preparation.......................................................................................................... 18
2.2.2 Histological examination ............................................................................................ 18
2.2.3 Analyses of skin viability ........................................................................................... 19
2.3 Epidermal isolation techniques ............................................................................... 19
2.3.1 Mechanical separation ................................................................................................ 19
2.3.2 Ionic change................................................................................................................ 20
2.3.3 Heat............................................................................................................................. 21
2.3.4 Enzymatic digestion ................................................................................................... 22
2.3.5 Microwave irradiation ................................................................................................ 23
2.4 Isolation techniques for the stratum corneum......................................................... 24
2.4.1 Mechanical separation ................................................................................................ 24
vi
2.4.2 Chemical separation ................................................................................................... 25
2.4.3 Enzymatic digestion ................................................................................................... 26
2.5 Preservation of the upper skin layers ...................................................................... 27
2.5.1 Short-term storage ...................................................................................................... 27
2.5.2 Long-term storage ...................................................................................................... 28
2.6 Discussion ............................................................................................................... 30
3 Linear shear response of the upper skin layers ............................. 35
3.1 Introduction............................................................................................................. 36
3.2 Methods .................................................................................................................. 37
3.2.1 Sample preparation .......................................................................................... 37
3.2.2 Experimental set-up ......................................................................................... 38
3.2.3 Rheological methods ....................................................................................... 41
3.2.4 Experimental procedures ................................................................................. 42
3.2.5 Histological examination ................................................................................. 43
3.3 Results..................................................................................................................... 44
3.4 Discussion ............................................................................................................... 48
4 A new indentation method to determine the mechanical properties
of epidermis ....................................................................................... 51
4.1 Introduction............................................................................................................. 52
4.1.1 Sample preparation .......................................................................................... 53
4.1.2 Experimental procedure ................................................................................... 54
4.1.3 Determination of the Young‟s modulus .......................................................... 55
4.2 Results..................................................................................................................... 57
4.3 Discussion ............................................................................................................... 59
5 Linear viscoelastic behavior of subcutaneous adipose tissue ....... 63
5.1 Introduction............................................................................................................. 64
5.2 Methods and Materials ........................................................................................... 66
5.2.1 Sample preparation .......................................................................................... 66
5.2.2 Rheological methods ....................................................................................... 66
5.2.3 Testing procedure ............................................................................................ 68
5.2.4 Statistics ........................................................................................................... 69
5.3 Results..................................................................................................................... 69
5.3.1 Small oscillatory strain behavior ..................................................................... 69
5.3.2 Model application ............................................................................................ 70
5.3.3 Time-Temperature Superposition .................................................................... 71
5.3.4 Freezing effects................................................................................................ 72
5.4 Discussion ............................................................................................................... 73
6 Does subcutaneous adipose tissue behave as an (anti-)thyxotropic
material? ............................................................................................ 77
6.1 Introduction............................................................................................................. 78
6.2 Materials & Methods .............................................................................................. 79
6.2.1 Sample preparation ..................................................................................................... 79
6.2.2 Rheological methods .................................................................................................. 80
6.3 Results..................................................................................................................... 82
6.3.1 Long term small strain behavior ................................................................................. 82
6.3.2 Large strain experiments ............................................................................................ 83
6.4 Discussion ............................................................................................................... 85
7 General discussion............................................................................. 89
7.1 Introductory remarks .............................................................................................. 90
7.2 In vitro model ......................................................................................................... 91
7.3 Mechanical methods ............................................................................................... 92
7.4 Main findings .......................................................................................................... 94
7.4.1 Small strain behavior of the epidermal layers ............................................................ 94
7.4.2 Mechanical behavior of the subcutaneous adipose tissue .......................................... 95
7.5 Implications for clinical and cosmetic applications ............................................... 95
7.6 Recommendations................................................................................................... 96
7.7 General conclusion ................................................................................................. 97
Samenvatting ........................................................................................ 99
Dankwoord.......................................................................................... 101
Curriculum Vitae ............................................................................... 103
References ........................................................................................... 104
Summary
In vitro mechanical characterization of human skin layers
Stratum corneum, epidermis, and hypodermis
The human skin is composed of several layers, each with an unique structure and
function. Knowledge about the mechanical behavior of these skin layers is important for
clinical and cosmetic research, such as the development of personal care products and
the understanding of skin diseases. Until today, most research studies were performed in
vivo and focused on the mid-layer, the dermis. However, clinical and cosmetic
applications require more detailed knowledge about the skin layers at the skin surface,
viable epidermis and stratum corneum, and the deeper lying hypodermis. Studying these
layers in an in vivo set up is much more challenging. The different length scales, ranging
from μm for the stratum corneum to cm for the hypodermis, the interwoven layered
structure and the inverse relation between penetration depth and resolution of noninvasive measurement techniques form major problems. As a consequence, hardly any
data are available for the viable epidermis and hypodermis and reported data for stratum
corneum are inconsistent.
The objective of this thesis was therefore to characterize the mechanical behavior of
individual skin layers in vitro and, for that, to develop the required experimental
procedures. It was considered essential to perform experiments with samples of
consistent quality in an accurate measurement set-up in a well-controlled environment.
To obtain samples of consistent quality, the integrity and viability of a skin layer needs
to be maintained. Therefore, various isolation and preservation methods were
investigated on tissue performance, reproducibility, and ease of handling.
Because of the inhomogeneous layered structure of the upper skin layers, mechanical
properties of the stratum corneum and viable epidermis were determined for various
loading directions. First, the stratum corneum and epidermis were subjected to shear
over a wide frequency range and with varying temperature and humidity. The typical
geometry of the upper skin layers required preliminary testing series in order to define
right experimental conditions to ensure reliable results. Subsequently, micro-indentation
experiments were applied using a spherical tip with a relatively large diameter. The
Young‟s moduli was derived via an analytical and numerical method. Because of the
complexity of measuring those skin layers, it was decided to focus on small
deformations first.
For both types of loading, result were highly reproducible. The shear tests demonstrated
that the shear modulus is influenced by humidity but not by temperature in the measured
range. The indentation tests showed that analytical methods are not appropriate to assess
the Young‟s modulus, such that finite element are required. If the skin is loaded
perpendicularly, the stiffness of the epidermis and stratum corneum, which is about 1-2
MPa, is about a factor 100 higher than for shear. No significant differences in stiffness
between the stratum corneum and viable epidermis were observed per loading type. The
results of these tests prove that it is essential to taken into account the highly anisotropy
of the tissue in numerical models.
Rheological methods were developed to study the mechanical response of the
subcutaneous adipose tissue. In the small linear viscoelastic strain regime, the shear
modulus showed a frequency- and temperature-dependent behavior and is about 7.5 kPa
at 10 rad/s and 37°C. Time-Temperature Superposition is applicable through shifting the
shear modulus horizontally. A power-law function model was able to describe the
frequency dependent behavior at constant temperature as well as the measured stress
relaxation behavior.
Prolonged loading of small strain results into a dramatic stiffening of the material.
Loading-unloading cycles showed that this behavior is reversible. In addition, various
large strain history sequences showed that stress-strain responses are reproducible up to
0.15 strain. When the strain further increases, the stress is decreasing for subsequent
loading cycles and, above 0.3 strain, the stress response become stationary. These
contrary results due to time effects and strain effects indicate that adipose tissue likely
behaves as an (anti-)thixotropic material, meaning that a constitutive model should
contain parameters to describe the build-up and breakdown of material structure.
However, further experimental research is needed to fully understand the thixotropic
behavior before such a model can be worked out in detail.
In conclusion, this thesis evaluates the mechanical behavior of stratum corneum,
epidermis and hypodermis using various in vitro set-ups. It was proven that for all skin
layers reproducible results can be obtained. The research was aimed at developing
reliable methods to determine the mechanical behavior of individual human skin layers.
Future work should be focused on the relationship between tissue deformation using
imaging techniques and heading to the determination of the skin‟s failure behavior in
relation to clinical and cosmetic treatments.
Chapter 1
General introduction
2 Chapter 1
1.1 Introduction
The largest organ of the human body, the skin, has a major role in providing a barrier
against the hostile external environment. The skin prevents excessive water loss from the
aqueous interior, the ingress of foreign chemicals and micro-organisms and provides
strength and resistance to mechanical loading. Other functions include insulation,
temperature regulation and sensation. To fulfill these functions, mechanical stability is
as important as mechanical flexibility. However, the mechanical balance of skin can be
threatened by diseases or medical or cosmetic treatments. In order to understand the skin
behavior during treatments or diseases, knowledge of the mechanical behavior of healthy
skin in normal conditions is essential.
Human skin is composed of several layers, each with a unique structure and function, but
most research on its mechanical properties have ignored this non-uniform layered
structure. For many clinical and cosmetic applications, however, knowledge of the
mechanical behavior of the various skin layers is indispensible (Figure 1.1). For
example, the benefit of transdermal drug delivery is that the microneedles exclusively
damage the pain-free outer skin layer, the epidermis. Its mechanical response is therefore
of particular interest. For needle insertion procedures going deeper into the skin or
diseases such as pressure ulcers, the mechanical properties of all individual skin layers
play a role. Although often not recognized, this is also the case during skin adhesive
removal or the use of consumer products such as shavers. For all these applications, the
subcutaneous fat layer considerably contributes by attenuating or dispersing the external
pressures, even when those are very small [1]. In addition, mechanical properties of the
distinct skin layers are needed to grow them artificially, serving a wide application field;
artificial outer skin can substitute animal and clinical testing in evaluating drugs,
cosmetics and other consumer products, while engineered fatty tissue facilitates large
volume soft tissue augmentation in plastic surgery. Furthermore, the mechanical
behavior of subcutaneous fat is crucial for many other clinical treatments outside the
scope of this thesis, such as liposuction surgery and cellulite treatments.
As the top layer of the epidermis, the stratum corneum, is the first barrier between the
human body and its environment, it is obvious that the mechanical response of this layer
needs to be understood. The significance of a proper understanding of the mechanical
behavior of the other part of the epidermis, the viable epidermis, and the subcutaneous
fat tissue is not yet commonly felt. Until today, research on skin mechanics has mainly
focused on full-thickness skin, the mid-layer (dermis) and the stratum corneum. Hardly
any experimental data are available for the other skin layers, i.e. the viable epidermis and
the subcuateneous fat. In addition, there is no consistency in data for the stratum
corneum. Hence, accurate numerical models including the mechanical behavior of
individual skin layers have not yet been developed for any of the applications mentioned
above. This thesis therefore focuses on the mechanical characterization of individual skin
General introduction 3
(a)
(b)
(c)
(d)
(e)
(f)
Figure 1.1 Clinical and cosmetic applications where the mechanical properties of separate
skin layers are important: (a) transdermal drug delivery; (b) skin-device contact such as
during shaving; (c) removal of adhesives such as ECG electrodes; (d) decubitus; (e) needle
insertion procedures; (f) tissue engineering.
4 Chapter 1
layers. Before the scope and outline of the thesis is given, the anatomy of the skin and
the state of the art on skin layer mechanics is shortly discussed.
1.2 A mechanical view of skin anatomy and physiology
Mechanical properties of skin are very diverse and depend on body site, age, race and
gender. Individual factors like exposure to UV irradiation, the use of creams and the
person‟s health and nutritional status may also modify the mechanical properties.
From the skin surface inwards, skin is composed of epidermis, dermis and hypodermis
(Figure 1.2 ). The epidermis is mainly composed of cells migrating to the skin surface.
The stratum corneum is considered as a separate layer because of its specific barrier
properties. It consists of non-viable cells and is very firm but pliable and wrinkled. The
other part of the epidermis, the viable epidermis, is also wrinkled. The underlying layer,
the dermis, is largely composed of a very dense fiber network dominating the
mechanical behavior of the total skin. The deepest skin layer, the hypodermis or
subcutaneous adipose tissue, is composed of loose fatty connective tissue.
All skin layers contain microstructures like blood vessels, lymph vessels, nerve endings,
sweat glands and hair follicles. The influence of these structures on the mechanical
properties are consideration to be ignorable, because the interest is on the bulk
mechanical behavior caused by the main components of the skin layer.
Of all skin layers, the dermis is the layer that is studied most. Consequently, data on
dermal properties are readily available. This thesis therefore focuses on the mechanical
behavior of the other layers, i.e. stratum corneum, viable epidermis and hypodermis.
Consequently, the anatomy and physiology of these skin layers are of particular interest.
Figure 1.2 Schematic representation of the different skin layers.
General introduction 5
1.2.1 Skin relief
The relief of the skin surface is formed by the association of furrows, follicular orifices
and sweat pores, and slightly protruding corneocytes. On most body sites, the main
furrows, called primary lines, are 70-200 μm deep, follow at least two directions and
delimit plateaus of variable shapes. The follicular orifices are located in the junction of
the furrows, whereas the sweat pores are mainly found on the plateaus or in more
superficial furrows, called secondary lines, being 20-70 μm deep. The third type of
furrows separate groups of corneocytes. The network of furrows varies with age and
gender.
The main function of the furrows is considered to be mechanical. By (partially)
smoothing out, the skin surface and the epidermis can extend without loading the cells.
Their anatomical distribution at a body site reflects the direction of the mechanical
constraints sustained by the skin. These furrows cannot be ignored when methods are
developed to mechanically characterize the stratum corneum and the epidermis.
1.2.2 Stratum corneum
The stratum corneum is composed of corneocytes, which are hexagonal flat cells without
a nucleus, held together by lipids and desmosomes in what is commonly referred to as a
brick-and-mortar structure (Figure 1.3). The diameter and thickness are ranging from 25
to 45 μm and approximately 0.3-0.7 μm, respectively [2,3]. The stratum corneum
consists of 15-25 [3,4] layers of corneocytes, resulting in a total layer thickness of about
10-25 μm [5]. The lipids are arranged in lamellar sheets, which consist of membrane-like
bilayers of ceramides, cholesterol, and fatty acids together with small amounts of
phospholipids and glucosylceramides. The intercellular spaces, i.e. the distance between
neighboring corneocytes, are about 0.1-0.3 μm [6]. Desmosomes, also called
corneosomes, are specialized inter-corneocyte linkages formed by proteins and, together
with the lipids, they maintain the integrity of the stratum corneum [7]. The lipids form
the major permeability barrier to the loss of water from the underlying epidermis.
The stratum corneum is continuously renewed. Cells are shed from the outside and
replaced by new ones. Changes in structure, composition and function of the corneocytes
occur as they move toward the outer skin surface. Cells of the deeper layers of the
stratum corneum are thicker and have more densely packed arrays of keratins, a more
fragile cornified cell envelope and a greater variety of modifications for cell attachment
as compared to cells of the outer stratum corneum. Consequently, the deeper part of the
stratum corneum has a major influence on its overall mechanical behavior. The outer
stratum corneum cells have less capacity to bind water. The cells in the outermost
stratum corneum have a rigid cornified envelope and in the same area, the desmosomes
undergo proteolytic degradation. These changes contribute to the continuous shedding of
the cells at the surface of the stratum corneum. Renewal time of the stratum corneum and
viable epidermis under normal conditions varies from 6 to 30 days [8].
6 Chapter 1
(a)
(b)
Figure 1.3: The morphology of the stratum corneum. (a) schematic drawing (b) cryostat
section of normal human skin treated with Sorensen’s alkaline buffer and methylene blue
to show the brick-and-mortar structure of the stratum corneum. Obtained from Marks
[9].
Although the corneocytes are non-viable, the stratum corneum is considered to be fully
functional, particularly in terms of barrier properties, and retains metabolic functions
[10].
The mechanical properties of both stratum corneum and viable epidermis are influenced
by environmental conditions such as relative humidity (RH) and temperature. In
addition, topical application either of pure water, moisturizers or emollients changes the
hydration state of the stratum corneum, significantly modifying some of its mechanical
properties. Although the hydration level depends on those factors, the hydration in the
stratum corneum under normal conditions varies from 5-10% near the surface up to 30%
near the transition to the viable epidermis. Bound water as component of proteins and
lipids, accounts for 20-30% of the total water volume. The total water content varies
little between 30% and 60% RH but considerably increases thereafter [11]. When fully
hydrated, the stratum corneum is able to become as twice as thick. In an in vitro
situation, however, the stratum corneum can increase to 400% of its original thickness
[10].
The stratum corneum matches the creases forming the skin surface. The deeper the
furrows and the steeper their sides, the higher their physiological range of extension. The
direction of the higher extensibility is perpendicular to the direction of the main furrows.
As a consequence, the stratum corneum in vivo hardly experience elongation stresses,
but only unfolding. This unfolding is an important feature of the overall skin resistance
to stretching.
1.2.3 Viable epidermis
The viable epidermis is a layered structure, consisting of three layers or „strata‟. The bulk
of epidermal cells are the keratinocytes, which migrate to the skin surface and become
non-viable in the stratum corneum. Other cell types within the viable epidermis include
melanocytes, Langerhans cells and Merkel cells.
General introduction 7
Keratinocytes change their shape, size and physical properties when migrating to the
skin surface. The structure of an individual keratinocyte correlates with its position
within the epidermis and its state of differentiation, which is reflected by the different
strata: the stratum basale, the stratum spinosum and the stratum granulosum (Figure 1.4).
The deepest layer is the stratum basale in which cell division occurs. It consists of 1 to 3
layers of small cubic cells. As the cells move towards the surface, they become larger
and polyhedral. The next layer is the stratum spinosum. The keratinocytes have become
polyhedral and are connected by desmosomes, which are symmetrical laminated
structures. The shape of the polyhedral cells becomes more flattened as they move
further outward. In the uppermost layers of the stratum spinosum so-called lamellar
granules appear. Those lamellar granules are lipid-synthesizing organelles that migrate
toward the periphery of the cell and eventually become extruded into the intercellular
compartment in the next layer, the stratum granulosum. At this stage of differentiation,
the degradation of mitochondria and nuclei starts and the cytoplasm of the flattened cells
become almost filled by keratohyalin masses and filaments. Furthermore, the cell
membrane becomes gradually thicker.
The thickness of the viable epidermis varies roughly from 30-100 μm [12]. The number
of cell layers differs from 5 up to 10. The cells are communicating by very strong
desmosomes in the very compact tissue; the intercellular spaces occupy less than 2% of
the volume [5,13]. Consequently, the viable epidermis is considered to be more rigid
than other soft tissues.
Because of its non-vascular structure, the epidermal cells are nourished from plasma that
originates in the dermal blood vessels and then transits through the epidermal-dermal
junction.
stratum
corneum
granulous
layer
LG
spinous
layer
N
KF
D
basal
layer
Figure 1.4: A schematic drawing and histological cross-section to show the structure of the
epidermis. In the schematic drawing the nucleus (N), the keratin filaments (KF), the
desmosomes (D) and the lamellar granules (LG) are depicted. The histological section is
taken from the skin of a young woman, obtained from Montagna et al. [14].
8 Chapter 1
1.2.4 Dermal-epidermal junction
The boundary between the dermis and epidermis is called the dermal-epidermal junction,
which provides a physical barrier for cells and large molecules. Four distinctive zones in
this strong junction can be identified: 1) the plasma membrane and hemidesmosomes of
the basal keratinocytes adhered to the junction, 2) the lamina lucida zone with anchoring
filaments, 3) the lamina densa, and 4) the amorphous sublamina densa fibrillar zone (see
also Figure 2.1). The firmness of the attachment is enhanced by parts of the epidermis
penetrating the papillary dermis resulting in large cones called rete ridges or papillae
[15]. The major point of weakness is considered to be the lamina lucida [16]. The
dermal-epidermal junction length over a straight line ranges from 1.1 to 1.3 [5].
1.2.5 Dermis
The dermis can be divided into two anatomical regions: the papillary and reticular
dermis. The papillary dermis is the thinner outermost portion of the dermis, constituting
approximately 10% of the 1-4 mm thick dermis. It contains smaller and more loosely
distributed elastic and collagen fibrils within a greater amount of substance than the
underlying reticular dermis. Its content in water and vascular volume show physiological
variations that can alter the mechanical behavior of skin as a whole. In addition, collagen
and elastin fibers are mostly vertically oriented in the papillary region and connect to the
dermal-epidermal junction, whereas in the reticular dermis they are horizontally oriented.
The amorphous ground substance acts as a viscous gel-like material, which does not leak
out of the dermis, even under high pressure. The reticular dermis forms a solid structure
with a permanent tension.
The dermis has a mainly mechanical function. The reticular dermis is able to extend up
to about 25% by undulating the collagen fibers, whereas it can be squeezed due to the
capacity to displace the ground substance laterally. The elastic fiber network ensures full
recovery of tissue shape and architecture after deformation. The permanent tension in the
reticular dermis generates the folding of the nonelastic overlying structures and hence,
the skin surface relief. The fiber network in the papillary dermis contributes to the
protection of vessels and cells against mechanical insults.
The dermis nourishes the epidermis. In the papillary dermis, therefore, the
microvasculature consists of papillary loops exchanging with extravascular elements and
a horizontal plexus in which the loops emerge. Although the vascularization throughout
the dermis seems sparse, the supply of the papillary loops is ensured by arterioles
irrigated from the deep dermis.
1.2.6 Hypodermis
The hypodermis is defined as the adipose tissue layer found between the dermis and the
aponeurosis and fasciae of the muscles. Its thickness varies with anatomical site, age,
sex, race, endocrine and nutritional status of the individual. The subcutaneous adipose
General introduction 9
tissue is structurally and functionally well integrated with the dermis through nerve and
vascular networks and the continuity of epidermal appendages such as hairs and nerve
endings.
The bulk of subcutaneous adipose tissue is a loose association of lipid-filled cells called
white adipocytes, which are held in a framework of collagen fibers. Only one third of
adipose tissue contains mature adipocytes [17]. In addition to the adipocytes, the
remaining two third contains stromal-vascular cells including fibroblastic connective
tissue cells, leukocytes, macrophages, and pre-adipocytes [18]. Adipose tissue has little
extracellular matrix compared to other connective tissues.
Stored fat is the predominant component of the adipocytes: the size of the lipid droplet
can exceed 50 μm. The cytoplasm and nucleus appears as a thin rim at the periphery of
the cell (Figure 1.5). The diameter of the entire white adipocyte is variable, ranging
between 30 and 70 μm [17]. Collections of white adipocytes comprise fat lobules, each
of which is supplied by an arteriole and surrounded by connective tissue septae. Each
adipocyte is in contact with at least one capillary. The good blood supply is necessary for
the exchange of metabolites and allows the adipocytes to function effectively. The
subcutaneous adipose tissue of the lower trunk and the gluteal thigh region has a thin
fascial plane dividing it into superficial and deep portions. Morphological differences are
observed between these two adipose tissue layers [19].
N
A
W
A
V
(a)
(b)
Figure 1.5: Schematic drawing and histological section of subcutaneous adipose tissue
showing white adipocytes (WA) with the nucleus (N) at the periphery. The adipocytes are
in good contact with the blood circulation via the arterioles (not visible here) which
branches the larger arteries (A) and veins (V).
The mechanical function of the subcutaneous adipose tissue has a double purpose: first
to allow the overlying skin to move as a whole, both horizontally and vertically, and
second, to attenuate and disperse spells of external pressure.
10
Chapter 1
1.3 State-of-the-art in skin layer mechanics
Measurement methods and mechanical properties of skin have already been extensively
reviewed in the literature [5,20,21]. Therefore, given the focus of this thesis, this review
is limited to studies on the behavior of stratum corneum, viable epidermis and
hypodermis. More specifically, mainly force-elongation studies, either in vivo or in vitro,
and currently available constitutive models are discussed.
1.3.1 In vivo vs in vitro experiments
When measurements regarding skin mechanics are carried out in vivo, the human skin
has its natural pre-stress and skin relief. The number of in vivo measurement methods is,
however, limited [21] and a numerical-experimental approach is usually required. In any
in vivo study, it is difficult to determine the contribution of each individual skin layer to
the overall skin response, whereas in vitro measurement methods offer the potential to
perform well-controlled experiments on individual skin layers. Another benefit is that all
kinds of mechanical testing can be applied and a wide range of reliable direct
measurement methods becomes available. However, due to the limited availability of
skin grafts, the number of experiments, the variety of skins, and the variety of body sites
might be limited.
The appropriateness of in vitro experiments on the stratum corneum should be carefully
considered. In vivo, the stratum corneum partly unfolds when the total skin is stretched,
but does not elongate. Therefore, in vitro mechanical characterization is only relevant to
the mechanical function of the skin under normal in vivo conditions, when its hardness
and capacity to absorb mechanical shocks are of concern. Stretching out of the stratum
corneum exclusively occurs in critical, extra-physiological situations.
1.3.2 Mechanical behavior of the stratum corneum
Force-elongation curves at constant elongation rate show one, two or three phases
depending on the hydration level in an in vitro situation (Figure 1.6) [22]. The first
phase, up to a 10% extension, is considered to be purely elastic. The next phase, absent
in low RH, is an irreversible elongation with a low slope, with strains ranging from 20125%. Only almost fully hydrated stratum corneum show the last phase before rupture,
where strain hardening is observed before rupture, at approximately 200% extension.
The slope becomes steeper if the extension rate rises, confirming the viscoelastic nature
of the material.Although the corneocytes are very elongated in tensile testing, the final
rupture is always extracellular and most likely at the desmosomes [8].
From the 1970s, various authors performed tensile tests [22,23,8,24,25,26]. Thereafter,
torsional techniques were developed to measure the stratum corneum behavior in vivo
[27,28,29,30]. From the nineties, nano-indentation techniques have been introduced to
determine the Young‟s modulus in vitro [31,32], whereas an in vivo indentation
technique had been developed already some years before [33]. Furthermore, imaging
General introduction 11
Load [g]
40
30
in vivo range
techniques such as ultrasound and magnetic resonance elastography have been used to
estimate mechanical properties [32,21].
32% RH
76% RH
20
98% RH
10
I
II
III
30 Elongation [%] 120
Figure 1.6 Typical force elongation curves for the stratum corneum at different RH
showing different phases: the elastic phase (I), the plastic phase (II) and the strain
hardening phase (III). Obtained from [22].
Reported Young‟s moduli vary over more than three decades, from a few MPa to GPa
[34,24,35,23]. The range of tensile moduli for various RH is shown in Figure 1.7. As
indicated in this figure, the stiffness of the stratum corneum varies from rubber-like to
nylon-like over the RH range. The differences may be due to a combination of several
reasons, such as regional differences, anisotropy, differences between species but also
test conditions such as sample preparation, difficulties in determining sample dimensions
and controlling the environmental conditions. A general trend, however, is a more
pronounced decrease of the elastic modulus beyond 60% RH. At a constant RH, the
stratum corneum hydration increases by 50% when the temperature rises from 20°C to
30°C. At higher RH, however, the temperature dependence of the modulus decreases and
declines to a minimum above 90% RH. More common trends due to an increase in RH
or temperature include an increase of the maximum extension and rupture work, and a
reduction of the force at rupture [22,24,36].
Preconditioning effects does not exist with stratum corneum, which is an important
difference with the whole skin, indicating the absence of mobile components in the
material [35]. In the same study, it was shown that stratum corneum behaves
isotropically in transversal plane only.
Current constitutive models of the stratum corneum are based on traction, relaxation and
creep tests [5]. From the tests, it was concluded that the model should include elasticity,
12
Chapter 1
non-linear viscosity and strain hardening parameters. The link between the defined
parameters and the anatomical components is yet to be determined.
Figure 1.7 An overview of Young’s moduli as function of the RH derived from in vitro
tensile tests on stratum corneum.
1.3.3 Mechanical behavior of the viable epidermis
Only recently, few studies have focused on this part of the epidermis. From an
indentation study, a local Young‟s modulus for the viable epidermis of murine ear skin
has been reported to be a few MPa [37,38]. However, murine skin has more dense hair, a
higher hair follicle density and a very thin epidermis compared to human skin. A
combined experimental-numerical approach on in vivo human skin led to an estimated
value of about 0.5 kPa for the Young‟s modulus of the upper human skin layers
including the papillar dermis [38,1]. The authors hypothesized that this low value is due
to the fact that the influence of the stratum corneum is negligble on the overall
mechanical response of the skin when suction is performed with small aperture sizes.
Because experimental data is limited, a constitutive model describing the mechanical
behavior of viable epidermis is not yet available.
1.3.4 Hypodermis
A limited number of studies is available regarding the mechanical behavior of
subcutaneous adipose tissue, either applying shear [39], compression [40,39], indentation
[41,42] or suction [38,1,39,43,40]. Only the suction experiments were performed in
vitro. Measured Young‟s moduli vary from a few kPa up to more than 100 kPa.
All studies give limited descriptions of the mechanical behavior as they were developed
for very specific applications. Consequently, a proper constitutive model based on
experimental data is not available yet. Current models are either limited to small strain
behavior [38,44] or based on other soft connective tissues.
General introduction 13
1.4 Aim and Outline
The objective of this thesis is to characterize the mechanical behavior of individual skin
layers in vitro and, for that, to develop the required experimental procedures. The focus
is on those skin layers for which hardly any data are available, i.e. the viable epidermis
and hypodermis, and those with inconsistent data, i.e. the stratum corneum. The results
should provide insight into the relationship between the mechanical responses of the
various skin layers to their structure and, hence, provide better understanding of the way
a treatment or disease affects the skin. Furthermore, the experimental data should be
suitable as input for constitutive models.
Previous studies, such as the various in vitro tensile tests on the stratum corneum, have
indicated that differences in mechanical properties of the epidermis and stratum corneum
cannot be caused by variations in humidity and temperature only, but also by test
conditions, anisotropy, sample preparation, and so on. It is therefore essential to perform
experiments with samples of consistent quality in an accurate measurement set-up in a
well-controlled environment. This will be initially done for relatively simple small strain
experiments in various directions under different environmental conditions. If this small
strain behavior is reproducible and well-understood, it has become meaningful to explore
the non-linear behavior.
In order to obtain in vitro samples of consistent quality, various isolation and
preservation treatments are first thoroughly investigated for both skin layers (Chapter 2).
Subsequently, a rheological measurement set-up has been designed to measure the shear
response of thin, soft tissues in a controlled environment (Chapter 3). A microindentation method has been adapted to enable the measurement of loading
perpendicular to the skin surface (Chapter 4). Because viable epidermis cannot be
isolated as a single layer, a numerical model is introduced to derive its properties from
the shear and indentation experiments on stratum corneum and whole epidermis.
Rheological methods are developed to study the shear response of subcutaneous adipose
tissue (Chapter 5). In order to study the large deformation behavior, it is essential to
understand its small strain behavior first. From those results, a constitutive model
describing the linear viscoelastic behavior of subcutaneous adipose tissue has been built.
Then, a set of experiments were designed to study the large deformation behavior and, in
relation to that, the time-dependent behavior (Chapter 6).
Finally (Chapter 7), a general discussion contemplates the chosen measurement methods
for the skin layers and the measurement outcomes as well as the significance of the
findings of this study for the various application fields.
Chapter 2
Isolation and preservation methods for
the epidermis and stratum corneum
16
Chapter 2
2.1 Introduction
Ex vivo human skin grafts provide a cost-effective alternative to animal and clinical
testing. Various companies, such as cosmetic, household product and pharmaceutical,
could benefit from in vitro studies to evaluate drugs, cosmetics and other consumer
products. Skin models are already used in many transdermal drug delivery and
percutaneous absorption studies, as well as in irritancy and toxicology studies. Studies
on ex vivo skin increase fundamental knowledge on the structural as well as mechanical
properties of skin. In addition, studies on isolated skin layers, such as the epidermis or
stratum corneum, could provide an insight into the specific contribution of each skin
layer to the overall skin response. Skin models enable improved control of experimental
conditions (i.e. temperature, hydration level) and offer the potential to perform wellcontrolled in vitro experiments. In order to obtain significant results, it is of utmost
importance that the structural integrity and viability of the skin are maintained.
The epidermis, the outermost skin layer, is directly contiguous to the external
environment and acts as a permeable barrier. It prevents excessive water loss from the
aqueous interior and protects the internal tissue against mechanical insults, UV
irradiation and the ingress of foreign chemicals and micro-organisms. Due to the
extraordinary nature of the epidermis, it is a challenge to completely isolate this skin
layer, while maintaining its structural integrity. The keratinocytes are surrounded by a
poor extracellular matrix and lack the support of a fiber structure, which usually provides
the strength and elasticity in a biological tissue. Within the epidermis, the mechanical
properties are determined by the rigid tonofilament cytoskeleton and the numerous
desmosomes to which the filaments are anchored at the cell periphery of the
keratinocytes. At the epidermal-dermal junction hemidesmosomes anchor the epidermis
to the dermis (Figure 2.1). These hemidesmosomes or the adjacent anchoring filaments
need to be disrupted to fully separate the epidermis from the dermis.
During isolation, to maintain the complex structure of the top layer, the stratum
corneum, the curvature of the skin surface needs to be followed. The architecture of the
stratum corneum is widely known as a solid brick-and-mortar structure, with flat
corneocytes surrounded by a matrix of lipid enriched membranes strongly held together
by desmosomes.
Due to the high number of plastic and cosmetic surgery procedures, such as
abdominoplasty and breast reduction, the availability of ex vivo human skin is high.
Whether a skin graft can be successfully used as skin model during in vitro experiments
depends on the nature of the tissue. The integrity of the skin tissue mainly depends on
the age of the subject as well as on the body site from which a graft is obtained.
Furthermore, within one skin graft structure changes might be as a result of disease or
treatments. These factors are usually reflected in tissue changes such as convolutions of
the epidermal-dermal junction, thickness of epidermal strata, cell shape and surface
folding, but may also lead to qualitative and quantitative differences in the various
Isolation and preservation methods for the epidermis and stratum corneum
17
epidermal components [45]. To obtain the best experimental outcome from in vitro
studies, it is important to use structurally and functionally intact models.
In order to use the available intact skin grafts as efficient as possible, factors such as
cleaning, preservation, and storage should be properly addressed, next to isolation
techniques. In various studies, such as transdermal drug delivery, percutaneous
absorption studies, irritancy and toxicology studies, an intact skin barrier is essential.
Furthermore, proper preservation is crucial for maintaining the viability and integrity of
the skin tissue. Tissue damage such as the creation of vacuoles are easily induced and the
selection of a proper tissue storage method is therefore important.
Evaluation techniques to assess skin viability during storage are numerous and have been
extensively discussed [46,47,48]. Common methods to assess viability include Trypan
blue dye exclusion, tetrazolium reductase activity, oxygen consumption rates, lactate and
glucose levels, and NMR spectroscopy. Structural integrity is usually assessed by
histological routines or imaging techniques.
Figure 2.1: Ultrastructure of the dermal-epidermal junction.
This study aims to critically review various isolation methods for the epidermis and
stratum corneum and preservation methods useful for in vitro research on split-thickness
skin, epidermis and stratum corneum. Some methods have already been reviewed today
[49,50,51,52], but none of the reviews are up to date. No standards exist yet,
complicating comparison between studies. In addition, much of the outcome of already
published work may have been influenced by the used preparation technique. Studies
performed in our own laboratories are added to this paper for completion. The present
paper describes mechanical, ionic change, heat, enzymatic digestion and irradiation
18
Chapter 2
techniques for isolation of the skin layers. The advantages and disadvantages of each
technique are discussed in terms of maintaining the skin integrity and ease of handling.
In addition, the influence of various storage conditions on the skin structure and viability
are discussed.
2.2 Skin preparation and analyses
General steps in the preparation of skin samples used for our own experiments, are
described below as well as the analysis techniques used to study the skin structure and
viability.
2.2.1 Skin preparation
For our own studies, human skin is obtained from female patients undergoing
abdominoplasty. The research proposal for our studies was approved by the Medical
Ethics Committee of the Catharina Hospital, Eindhoven, the Netherlands. Immediately
after excision, the skin is brought to the laboratory for further processing. Here, the skin
is placed on a stainless steel plate covered with paper towels to absorb body fluids. The
skin surface is cleaned with pure water. Using multiple forceps, the skin graft is
stretched and fixed to the stainless steel plate. Subsequently, split-thickness skin
samples, varying in thickness from 100-400 µm, are generated using a dermatome (D42,
Humeca, The Netherlands) (Figure 2.2).
(a)
(b)
Figure 2.2: Skin is stretched using forceps (a) and dermatomed (b).
2.2.2 Histological examination
In order to examine tissue structure in our laboratories, samples were fixated in 10%
phosphate-buffered formalin and processed for conventional paraffin embedding. The
sections were cut into 5 μm slices and stained with aldehyde-fuchsin and yellow green
SF (Merckx) or standard heamotoxilyn and eosin (H&E) staining. The tissue
morphology was studied by light microscopy. The aldehyde-fuchsin staining is used to
clearly identify the different skin layers: stratum corneum, viable epidermis, papillar
dermis and reticular dermis (Figure 2.3a). The structural integrity is examined by using
the H&E staining.
Isolation and preservation methods for the epidermis and stratum corneum
19
2.2.3 Analyses of skin viability
Skin viability was studied by using a colorimetric MTT (Thiazolyl Blue Tetrazolium
Bromide) assay. Skin samples with a diameter of 8 mm were placed in a 24 wells-plate
containing 300 µl 1 mg/ml MTT solution in PBS (Phosphate Buffered Saline). The
plates were incubated at 37C and 5% CO2 for a period of 3 hours. After incubation, the
skin samples were removed and gently blotted with tissue paper, before completely
submerging them in 2 ml 2-propanol per well. The extraction plates were placed in
sealed bags to reduce evaporation and were gently shaken overnight at room temperature
to extract the reduced MTT. The absorption of the extractant was measured at 570 nm
using plain extractant as blank.
2.3 Epidermal isolation techniques
Isolation techniques for the epidermis can be divided into the following categories:
mechanical, ionic change, heat, enzymatic digestion and irradiation techniques. These
techniques are discussed in this section. The success rate of the various methods are
summarized at the end of the section in terms of actual cleavage plane, retaining viability
and maintenance of integrity (Table 2.1).
2.3.1 Mechanical separation
Cutting by using a dermatome
Van Scott et al. [53] recommended a stretching method for separating the epidermis
from the dermis. In this method, the skin is manually stretched to its limit over a slightly
convex wooden surface, and is anchored in place by means of thumbtacks. A razor blade
or scalpel is used to scrape off the epidermis. Subsequently, the epidermis is grasped by
tweezers to gently detach a continuous sheet. However, damage is easily induced in the
epidermis using this rough stretching technique. The severity of this damage depends on
the vigour of scraping and the degree of stretching. The development of keratomes,
either handheld devices or as part of a mechanical device, has improved the
reproducibility of this stretching technique.
In our study, we used a cordless, battery operated dermatome. As indicated in section
2.2, ex vivo skin was mounted on a stainless steel plate to facilitate the cutting process.
When the dermatome was set to 100 μm, samples of the epidermis could be obtained. In
some cases, however, some papillar dermis is still attached (Figure 2.3). Due to the
presence of rete ridges, it is highly unlikely that the cutting plane is going through the
dermal-epidermal junction only. The number of skin layers present in the separated
tissue can be assessed visually; the yellowish translucent epidermis is easily
distinguishable from the white opaque dermis. A MTT-test demonstrated that the
dermatomed skin retained its viability, which is in agreement with Wester et al.[54].
The obtained geometric shape is very convenient for assessing its mechanical properties.
It is assumed that the mechanical properties of the present papillary dermis are similar to
20
Chapter 2
the surrounding epidermal tissue, because no differences in shear properties were found
between 100 and 200 μm thick split-skin samples (Chapter 3).
SC
SC
VE
VE
(b)
PD
SC
VE
PD
RD
(a)
(c)
Figure 2.3. (a) Full thickness skin stained with aldehyde-fuchsin to visualize the stratum
corneum (SC), viable epidermis (VE), papillar dermis (PD) and reticular dermis (RD); (b)
Dermatomed skin with a set thickness of 100 μm consists of the epidermal layer only; (c)
In some cases, however, some papillar dermis is still attached.
Suction device
Suction blisters can be produced by applying suction cups on the skin, in vivo and in
vitro. In vivo separation of the human epidermis was first accomplished in 1964 [55].
Kiistala et al.(1968) found that within 130 minutes a blister with a suction gap of 25 mm
can be incited. The diameter of a suction cup may vary from 15-50 mm depending on
body site. To avoid tissue damage, the pressure within the cup has to be maintained at
200 mm Hg or more. The cleavage occurs in the plane through the lamina lucida, leaving
the basement membrane on the dermis and retaining an intact, viable basal cell layer.
However, enlargement of intercellular spaces due to considerable stretching might cause
large vacuoles in keratinocytic cytoplasm [56,50].
Suction blister time depends on factors such as suction pressure, individual variation and
regional differences as well as temperature, but does not depend on cup size. Because of
the low reproducibility caused by individual variations that cannot be controlled, this
method is considered to be unfavourable.
2.3.2 Ionic change
One of the first methods to isolate the epidermis was by maceration in dilute acetic acid
in order to perform mitotic counts. Cowdry [57] described that dilute acetic acid causes
swelling of collagen fibers which decreases their cohesive strength and, therefore, the
binding of epidermis to dermis. In addition, it was found that collagen fibers also swell
Isolation and preservation methods for the epidermis and stratum corneum
21
in an alkaline environment. These methods, however, kill epidermal cells and are
therefore no longer used [58].
In addition, EDTA (ethylenediamine tetraacetic acid) has been used to obtain epidermal
sheets [59]. The location of the split depends on the duration of the treatment. After 30
min incubation in 0.01 M EDTA at pH 7.4 the split occurred in the lower granular layer,
after 45 min it was in a spinous-suprabasilar location and after 60 min or more at the
dermal–epidermal junction. Besides this, intracellular oedema is increasing with time. So
this is not a favourable method for epidermal separation either.
After prolonged incubation in 1 M NaCl at 4°C, the epidermis can also be easily
removed from the dermis with forceps. The split occurs through the lamina lucida.
Nevertheless, mitochondrial swelling was noted within the keratinocytes [50]. Although
no other degenerative features have been reported, epidermal components may be
diminished or modified during the long incubation times of 24 to 96 hours [60].
Prolonged incubation in PBS is also known to separate the epidermis from the dermis.
After 72-96 h at 37°C, the epidermis can be readily peeled off [61]. In contrast to the
above techniques, where the split occurs through the lamina lucida, the split is closer to
the epidermal site of the dermal-epidermal junction [61].
Since no intact viable epidermal sheets can be obtained using techniques based on ionic
change, all are considered not to be suitable for epidermal isolation.
2.3.3 Heat
Separating the epidermis from the dermis using a hot plate is a simple and rapid method
[58]. The skin is heated up to 50 to 60C for 30 s. To maintain enzyme activity, mild
heat treatment at 52C for 30 s is required. Separation occurs at the basal cell layer.
Depending on the exact conditions, release of enzymes, cytolysis and cell separation
may occur. However, it has been claimed that heat does not modify fibrous proteins
within isolated epidermis [62]. Heating can easily cause tissue dehydration. This
problem can, however, be circumvented by increasing the humidity of the environment
or by placing the skin in a sealed bag in hot water instead of using a hot plate. After
heating, the epidermis can be gently peeled from the dermis.
In our studies, human skin samples were heated on a hot plate and in a sealed bag.
Heating the skin on a hot plate seemed to flatten the undulating epidermal structure,
while the papillae remained intact after heating in a sealed bag in hot water. The
epidermis could be peeled from the dermis after more than 5 minutes.
For both heat separation techniques, structural tissue damage occured in terms of
vacuoles and a disrupted basal layer (Figure 2.4). In addition, Wester et al.[63]
demonstrated that heat treated skin (60°C for 1 minute) and heat-separated epidermis and
dermis significantly lose viability. Furthermore, some practical problems arose when
using a hot plate, such as curling of the dermal tissue and uneven separation of the
epidermis over the complete skin surface due to gradual thermal diffusion. Lastly, it
should be noted that much longer heating times were needed than mentioned in
literature.
22
Chapter 2
(a)
(b)
Figure 2.4. Histological cross sections of epidermis isolated using heat by means of a hot
plate (a) or placing the epidermis in a sealed bag in hot water (b). A standard H&E
staining has been used.
2.3.4 Enzymatic digestion
Trypsin
Epidermal separation by means of trypsin has been widely used, although some
conflicting results have been published. Briggeman et al. [64] reported that the epidermis
is isolated by the cleaving effect of trypsin, whereas other authors reported that many
basal cells remain loosly attached to the basement membrane after trypsin treatment
[65,66].
The epidermis can be easily peeled from the dermis using 0.1-0.3% trypsin in a saline
solution supplemented with calcium and magnesium at 4°C. However, these conditions
also induce a high level intra-epidermal split at the spinous-granular interface [45].
Inconsistencies within the reported findings seem to be related to various factors such as
size and thickness of the skin sample, enzymatic concentration and its solvent,
incubation time and temperature. In addition some side-effects are not yet expressed
immediately after trypsinization and post-trypsinization recovery may take up to a few
days [45].
The result of epidermal isolation using trypsin depends on the specific treatment
conditions in relation to the donor skin and hence, is less suitable for obtaining intact
epidermal sheets. Other enzymatic methods for epidermal separation have been proven
to be more consistent.
Thermolysin
The epidermis can easily be separated from the dermis following incubation at 4C for 1
h in a solution containing 250-500 g/ml thermolysin, a proteolytic enzyme hitherto
mostly used for protein analysis [65]. Thermolysin can be dissolved in sterile magnesium
free PBS containing 1 mM CaCl2 at pH 7.8. It is strongly advised to remove at least the
subcutaneous fat and the lower dermis to enable the penetration of this enzyme. Light
and electron microscopy revealed that the separation occurred at the lamina lucida and
that the hemidesmosomes were selectively disrupted, whereas Willsteed et al.[50]
noticed an intraepidermal split, without any lamina lucida separation. Since the
Isolation and preservation methods for the epidermis and stratum corneum
23
introduction of this relatively new treatment, it has not been widely used. This is likely
due to the gentle treatment by another enzyme called dispase.
Dispase
Dispase II (Roche Diagnostics) has been proven to be a rapid, effective, but gentle agent
for separating intact epidermis from the dermis. This proteolytic enzyme is able to cleave
the basement membrane zone region while preserving the viability of the epithelial cells
[67]. Kitano and Okado were the first authors who described the seperation process [68].
Based on recommendations from the supplier, 2.4 U/ml dispase in 50 mM Hepes/KOH
buffer pH 7.4 with 150 mM NaCL was used in our studies to separate the epidermis from
the dermis. Fresh skin samples of various sizes were placed on top of sterile gauzes in
petridishes (diameter = 6 cm) containing 5 ml of 2.4 U/ml Dispase II. The stratum
corneum of the skin samples was not exposed to the enzymatic solution during the
separation process to prevent loss of the skin barrier integrity. After overnight incubation
at 4C and thereafter 10 min at 37C, the epidermis was gently peeled from the dermis
using tweezers. It was demonstrated that the bottom surface of the separated epidermal
sheet retained its rete-ridges and hair follicles with sebaceous glands and the eccrine
sweat glands retained their undistorted shape [68] (Figure 2.5). The cleavage occurred in
the lamina densa [69]
This isolation method is very suitable for generating intact epidermal sheets. The best
results were obtained when split-thickness skin samples of roughly 300 µm were used to
facilitate the diffusion of this enzyme. Therefore, it is recommended to dermatome skin
grafts prior to performing the enzyme treatment.
Figure 2.5. H&E staining of epidermis separated with Dispase.
2.3.5 Microwave irradiation
Sanchez et al. [70] explored the effects of microwave irradiation on epidermal-dermal
separation. Epidermal samples were obtained after incubation in 0.02 M EDTA in PBS
and microwave irradiation with 4 pulses of 420 watts for 5 sec, with a total incubation
period of 4 min. The hemidesmosomal junctions are then disrupted, whereas an
additional incubation time may affect keratinocyte junctions. Microwave irradiation has
been widely used for tissue fixation and immunostaining.
Care should be taken to avoid damage to the tissue integrity. It is important to use the
prescribed buffer and specifically adhere to the recommended microwave exposure
24
Chapter 2
times. Nevertheless, microwave irradiation seems to be a rapid method for separation of
the epidermis from the dermis.
Table 2.1: Overview of effectiveness of isolation techniques for the epidermis.
Treatment
duration
< 1 hr
< 2hrs
Suction
Heat
5 min
Ionic
24-96 hrs
NaCl
change
> 1 hr
EDTA
72-96 hrs
PBS
Enzymatic Trypsin
1-24 hr
digestion
1 hr
Thermolysin
24 hrs
Dispase
Irradiation Microwave
5 min
Type
Method
Mechanical Dermatome
Tissue Tissue
Cleavage plane integrity viability Reproducibility
variable
+
+
+
lamina lucida
basal layer
lamina lucida
n.a.
hemidesmosomes
variable
0
0
-
0
n.a.*
n.a.*
0
0
0
0
0
-
hemidesmosomes
lamina densa
hemidesmosomes
+
+
0
+
+
n.a.*
n.a.*
+
+
*n.a. = not available
2.4 Isolation techniques for the stratum corneum
Isolation techniques for the stratum corneum can be divided into the following
categories: mechanical, chemical and enzymatic digestion techniques. Again the benefits
and drawbacks of the various techniques are discussed. At the end of the section, the
succes rate of the different techniques are summarized in Table 2.2.
2.4.1 Mechanical separation
Stratum corneum separating by cutting techniques is complicated because of the skin
curvature. Howevr, the thickness of the stratum corneum has little variation. So when
there are means to decrease the skin curvature, mechanical separation through cutting
might become possible. It was already shown that the skin relief dramatically decreases
when a microscope slide is placed on top of it [71]. We performed topography
measurements on unloaded and loaded skin using a PRIMOS (GFM, Germany), using
light profilometry to assess the surface roughness. A piece of skin of 20x20 mm was
placed on a microscope slide after removal of the subcutaneous fat layer. First, the initial
surface roughness parameters were measured. Then, another microscopic glass was
placed on top and pushed down by two weighs of 100 g on each side. Again the
roughness parameters were determined. Preliminary testing showed that the microscopic
slide on top was neglected by the system and did not influence the measurement output.
A significant decrease in skin surface roughness average was measured: 42 μm in a
loaded configuration versus 85 μm when unloaded. The latter is comparable to what can
be found in literature [5]. Unfortunately, the surface roughness was still at least three
times the thickness of the stratum corneum.
Isolation and preservation methods for the epidermis and stratum corneum
25
Following the topography measurement, the sample was kept between the two plates and
stored at -80°C. In order to retain the flattened state of the skin sample, the sample was
cut by use of a cryotome. The surface of the stratum corneum was aligned with the
cutting system to obtain stratum corneum with one single cut with a thickness of 20 μm.
The stratum corneum sheets have some other epidermal strata attached and cavities
(Figure 2.6).
(a)
(b)
Figure 2.6. Stratum corneum isolated from flattened skin. Due to the skin curvature, other
epidermal strata and cavities are still present. Transversal sections of the obtained sheets
are depicted with 5x (a) and 40x (b) enlargement.
2.4.2 Chemical separation
Cantharidin blister procedure
This method, however, has only been reported up to the early seventies [22,8].
Cantharidin was impregnated into 1 cm diameter disks of filter paper and placed under
occlusive patches rather than applied directly to the skin surface in a volatile solvent.
The disks were removed after 4 hours and protective caps were placed over the forming
blisters to prevent damage to the samples. The blister tops were surgically excised and
the loose underlying wet cells removed by gentle swabbing. Since the discovery that
cantharidin is toxic, it is not permitted to use it for skin treatments anymore.
Ammonia vapour
In the sixties and seventies, it was common to isolate stratum corneum through exposure
to ammonia vapour. The latest protocols reported around 30 min exposure to separate the
dermis and epidermis [72,73]. Adherent wet cells are subsequently removed with a
cotton swab such that the stratum corneum sheet remains [74]. Thereafter, the stratum
corneum sheet was allowed to dry on silicone-coated paper at ambient conditions. In
addition, it was noticed that the success of this treatment is variable. Since more
consistent techniques causing less damage became available, this method is no longer
used.
26
Chapter 2
2.4.3 Enzymatic digestion
Trypsin
The working of trypsin throughout the epidermal strata has been extensively studied
[75]. It appeared that the architecture of the stratum corneum remains unaffected by
trypsinization. Corneodesmosomes and composite desmosomes shared by corneum and
granular cells are normal. Tonofilaments attached to these junctions also appear
unchanged [76]. However, concentrations of trypsin above 0.125% might damage the
stratum corneum such that its elastic properties change [5].
In order to enable the working of trypsin on the epidermal cells, the subcutaneous fat
layer and the lower dermis has to be removed. In our laboratories, the remaining skin
was immersed in a porcine 0.1% trypsin (SV30037.01, Hyclone) solution in PBS
(Phosphate Buffer Saline). For quick processing, the samples were then placed for over 2
hours in an incubator at 37°C. For this study, dermatomed skin of approximately 300 μm
thick and a surface area of 2 cm2 was placed in 3 ml trypsin. Similar results can be
obtained through an overnight culture at 4°C and 15 min at 37°C. Due to the lipids
within the stratum corneum, the thin layer floats to the surface while the remaining
epidermis sinks to the bottom. In order to prevent post trypsinization effects, stratum
corneum is rinsed with distilled water a few times to wash out trypsin and treated with
anti-trypsin. The overnight protocol can be considered as the golden standard, which is
frequently described and commonly used within several research fields.
(a)
(b)
Figure 2.7. (a) After staying overnight at 4°C, the extracellular matrix of the viable
epidermis is still attached to the stratum corneum; (b) Only stratum corneum is obtained
after leaving the skin sample for 1 hour at 37°C.
Table 2.2: Overview of effectiveness of isolation techniques for the stratum corneum.
Type
Mechanical
Method
Cutting
(cryotome)
Cantharidin
Ammonia
Ionic change
Enzymatic digestion Trypsin
Treatment Tissue
duration integrity
Tissue
viability Reproducibility
24 hrs
±
-
-
4.5 hrs
45 min
2-24 hrs
+
+
±
+
Isolation and preservation methods for the epidermis and stratum corneum
27
2.5 Preservation of the upper skin layers
This section discusses preservation techniques regarding in vitro skin research. It is
assumed that these techniques are equally suitable for all skin grafts, i.e. full-thickness,
split-thickness, and epidermal grafts. From studies on skin grafts used as burn wound
dressings, it is known that in order to provide the best clinical outcome, skin grafts
should be properly preserved. When procuring cadaver skin for banking, the cadaver
donor should be cooled as soon as possible to avoid/minimize structural tissue changes,
i.e. changes in basement membrane components [77], and to maintain viability. Within
12 to 30 hours from harvest, post-mortem skin allografts exhibit an average viability
index of 75% with little variation, which decreases to 40% within 60 hours. In addition,
Bravo et al. [54] found that human cadaver skin grafts only exhibited approximately
60% of the metabolic activity found in fresh skin samples from living surgical donors.
However, the availability of skin grafts from living donors is limited to certain body
sites.
Currently available methods used by skin banks for storing viable skin can be divided
into short-term and long-term techniques. As a large variation in protocols have been
published for storage of skin grafts and those have been extensively reviewed
[77,77,78,54,79], only methods useful for in vitro testing are discussed in this section. As
a consequence, some protocols that are recommended by guidelines and standards, are
not taken into account when scientific studies have shown evidence that both viability
and integrity are not maintained.
2.5.1 Short-term storage
Due to its simplicity, cost-effectiveness and ease of availability, refrigeration of skin
grafts remains the most widely used method today worldwide for short-term storage
[47]. Refrigerator storage reduces the metabolic rate of the cells and hence, the
nutritional demands and metabolic production. In addition, bacterial proliferation is
inhibited.
Without the use of preservation media, it has been reported that epidermis from porcine
ear skin, which is a proper model system for human epidermis, is still in normal
condition after 4 up to 6 hours at 4°C [80]. Degenerative changes started to occur at the
stratum corneum and are independent of storage temperature. In contrast, the lower parts
of the epidermis are generally compacted, but remained more or less structurally intact
for a relatively long period.
Today various isotonic media are in use for refrigerator storage (4 C) of skin grafts,
which can be divided into nutrient media (e.g. HHBSS, RPMI-1640, Eagle‟s MEM with
L-glutamine, McCoy's 5A) and saline solutions [75,77,78]. In general, nutrient media are
considered to be a better medium than saline, as they are rich anorganic salts, amino
acids, glucose and vitamins that are essential for graft viability. Mathur et al. [81] studied
the preservation of viable cadaver skin grafts in PBS at 4°C. The viability was intact
after 24 h of storage but rapidly declined afterwards; after 1 week the viability dropped
to 27% compared to fresh skin, after 2 weeks the tissue was non-viable. In addition, the
28
Chapter 2
integrity is lost because of oedema [54]. In contrary, human cadaver skin stored at 4°C,
in McCoy's 5A medium retains viability for 4 weeks [79]. Castagnoli et al. [82]
demonstrated that the viability of human skin stored in RPMI-1640 media at 4°C
decreased slowly, retaining 25% viability compared to that of fresh skin after 15 days of
storage, with no damage to skin architecture until 7 days post-procurement. Wester et al.
[63] found that the anaerobic metabolism, i.e. the conversion of glucose into lactate, of
dermatomed human cadaver skin maintained a steady-state value through 8 days of
culture in Eagle‟s MEM-BSS at 4°C.
For percutaneous absorption studies, basal nutrient medium is preferred over growth
medium containing blood serum, hormones and growth factors. The receptor fluid used
within a diffusion cell, should not interfere with the analytical endpoint measurement,
e.g. HPLC analyses. Recently, it was demonstrated in our laboratory that epidermis from
fresh skin grafts of living donors, isolated by using a dermatome, can maintain its
viability and integrity for 72 hours when maintained in HHBSS in an incubator at 37 C
and 5% CO2 (data not shown). This is inagreement with results of Bravo et al .[54].
For the stratum corneum, PBS is a sufficient medium for short-term storage. In order to
avoid the growth of bacteria and fungi and the loss of tissue integrity, it is recommended
to store at a temperature of 4°C if it is only for a few days.
2.5.2 Long-term storage
Cryopreservation
Long-term storage of skin is possible via cryopreservation. In general, the success rate of
freezing tissue depends on various factors, i.e. conducting medium, the cooling rate, the
number of cell types in the tissue, the addition of a cryoprotective agent, storage
temperature, the cooling rate and thawing rate. The viability of the epidermis (and
dermis) can be well-retained when cooling to ultralow temperature by using
cryoprotective agents (CPA‟s), without the formation of ice crystals. Cryporeservation is
likely the most routinely used method for long-term storage of skin, because the skin can
then be stored for months to years [83].
Any cell type has its optimum cooling rate producing maximum cell survival. If the
cooling rate is higher than the optimum, intracellular ice appears, causing the cell to die.
In contrast, if the cooling rate is slow, free water is removed from solution to form
extracellular ice crystals increasing the salt concentrations in the tissue. The cells also
shrink because of osmosis. It is unlikely that each cell type within a tissue will exhibit
the same optimum cooling rate. Although epidermis mainly consists of keratinocytes,
maintaining high viability for all epidermal cells would be challenging. This can be
achieved, however, if cryoprotective chemicals are added before freezing. The most
common CPA‟s are glycerol and dimethyl sulfoxide (DMSO). These cryoprotective
agents act as solvents for the salts. In addition, their presence within the cells prevents
excessive shrinkage of the cells during the cooling phase. Therefore in the presence of
CPA‟s, it is possible to use very slow cooling rates that minimize intracellular ice
Isolation and preservation methods for the epidermis and stratum corneum
29
formation while protecting the cells against solution effects. High viabilities of all cell
types can be achieved using this slow cooling rate: a cooling rate of -30°C per minute
was shown to maintain the viability of keratinocytes [77].
When skin tissue cryopreserved with 15% glycerol in PBS or nutrient medium has been
cooled by a controlled-rate process to at least -80°C, it can be transferred for long-term
storage into the vapor phase of liquid nitrogen (below -130°C). Once the skin is at a
temperature lower than -130°C, i.e. the glass transition temperature of water, no further
loss of cell viability is incurred.
The optimum thawing procedure is a rapid warming method. This can be achieved by
plunging the skin into a 37°C water bath until the tissue is just thawed. Prolonged
storage at 37°C in the presence of CPA would be detrimental. Because the cells contain
high concentrations of CPA, they are hyperosmotic compared with normal saline. To
avoid osmotic lysis of the cells, either the saline can be added gradually or an
impermeant solute such as sucrose can be added to the saline to reduce the difference in
osmolarity. It has been reported that viability declines rapidly after thawing of the skin,
even if the epidermis is stored in nutrient media [54].
It should be noted that it is prefered to use glycerol rather than DMSO, because it has a
lower toxicity to the cells and is more effective [84,81]. Nevertheless, the skin viability
might be somewhat lower after cryopreservation with glycerol [54].
Although CPA‟s are relatively non-toxic at low temperatures, the toxicity can become
significant at higher temperatures. However, structurally intact skin tissue is relatively
resistant to cryogenic damage compared to single cells. In addition, the rate at which
CPA‟s enter the cell depends on the temperature and the CPA, being faster at higher
temperatures. CPA can be best dissolved in a HEPES or TES buffer, because those
zwitterionic buffers do not lose their buffering capacity at lower temperatures.
Many different methods are in use for the packaging of frozen skin, ranging from rolls of
skin within a tube to the use of flat pack bags in metal laminated pouches. The latter are
preferred, in that the greater surface area to volume ratio ensures more even cooling
across the skin tissue, and the metal laminates are good heat conductors [77].
Snap freezing
Snap freezing in a well conducting medium, e.g. salt water, isopentane or hexane,
provides an effectice, rapid storage method without causing structural damage due to
water phase transitions. In practice, skin samples packed in a metal pouch can be
emerged in a 2-methylbuthane, which is cooled down by liquid nitrogen to -80°C. The
skin samples will immediately freeze and can then be stored at a -80°C freezer until use.
Since this is above the glass transition temperature of water, the slow progressive decline
in viability limits the maximum storage time to months.
After slowly thawing at room temperature, there is no need to thaw in a buffer before
using the tissue.
Although the tissue is not viable anymore, the tissue integrity is well maintained. As
Foutz et al. [85] showed that the mechanical properties of human skin are not affected by
freezing as well, it might be sufficient to snapfreeze samples for mechanical
30
Chapter 2
characterization. Snapfrozen tissue is used for penetration and permeation studies as
well.
Drying stratum corneum
The routinely used method of drying stratum corneum is presumably the best method to
store isolated stratum corneum. According to the protocol of Bouwstra et al. [86], drying
and storage should take place in a cool dark room under an atmosphere of argon or
krypton. Because of possible detoriation of the lipid organization, it is recommended to
adhere to a maximum storage period of approximately three months.
Drying stratum corneum facilitates handling of the specimen. Most commonly is to dry
the stratum corneum on filter paper, but damage may occur to the fragile sheet upon
removal from the filter paper. The use of a sieve instead of the filter paper solves this
problem, since the stratum corneum can be removed even dried. Before the stratum
corneum sheet can be assessed, it needs to be immersed in pure water or PBS.
SC
epidermis
Table 2.3: Overview of the ease of handle and success rate of various isolation
techniques.
Treatment Storage Tissue Tissue
Type
Method
duration
time integrity viability
short-term saline solutions
none
days
0
nutrient media
none
weeks
+
+
long-term cryopreservation
< 1 hr
years
+
+
snap freezing
10 s
3 months
+
short-term PBS
none
3-5 days
+
long-term drying
few days
years
+
+
2.6 Discussion
Isolation and preservation techniques of both epidermis and stratum corneum are of
importance for various in vitro studies to evaluate drugs, cosmetics and other household
products. Various skin isolation and preservation techniques are commonly used today,
although the effectiveness of each of these techniques has not been properly reviewed.
This study provides an overview of current techniques of which the isolation methods
can be divided into mechanical, ionic change, heating, enzymatic digestion and
irradiation techniques for skin isolation. The study describes the advantages and
disadvantages of the various methods in terms of reliability and maintaining skin
integrity and viability (Error! Reference source not found. and Table 2.2). Since the
cleavage plane is another indicator for the succes rate of a method, the cleavage location
is also specified for each of these isolation methods. In Table 2.3, the effect of various
storage conditions on the skin structure and viability are discussed. Here, the acceptable
storage time is also indicated per method.
Isolation and preservation methods for the epidermis and stratum corneum
31
The overview in Error! Reference source not found. shows that only few isolation
methods are suitable for obtaining intact viable epidermis. Although the response of the
skin to stresses such as mechanical suction and exposure to hyperosmolar salt solutions
supports the concept of the lamina lucida being the natural cleavage plane of the skin
[source], these methods are not recommended. The exact cleavage location due to
hyperosmolar salt solutions can also be between epidermal layers, because the cleavage
strongly depends on the duration of the treatment. In addition, these treatments are
detrimental to the isolated epidermis. Because their protocols are also very time
consuming, the techniques are incovenient for routine labaratory application as well. It
was decided not to include further analysis of this method in this study to asses the effect
on viability and integrity. The effectiveness of mechanical suction depends on the exact
suction blister time. As the suction blister time strongly depends on various individual
factors, it is considered to be impossible to obtain samples with consistent quality.
Compared to the methods discussed above, isolation using heat or irradiation is much
less time consuming. However, although frequently used, heat treatment does not result
in an intact viable epidermis. The cleavage disrupts the basal layer, the viability declines
and structural changes like cell separation have been observed. In our own lab, for gently
removing the epidermis from the dermis, much longer heating times were needed than
reported in the literature. This is probablyly due to specimen type or experimental
conditions such as the humidity level. Due to the longer heating time, the susceptibility
to structure changes and loss of viability are increased so this method is considered
unfavourable as well. Isolation using microwave irradiation has been explored with
satisfying results regarding tissue viability, but it is not commonly used yet. To fully
assess the usefullness of microwave irradiation, more studies are needed.
Three enzymes are known to induce the dermal-epidermal split. The obvious advantage
of enzymatic digestion is that isolation takes place because of differences between cells
meaning that the undulating structure of the epidermis and stratum corneum is followed.
Trypsin, however, can cause cleavage at various planes in the epidermis making the
treatment using trypsin unreliable. Thermolysin might be an alternative, but practicle
studies with this enzyme are rare. What is widely used and extensively studied is
enzymatic digestion using dispase. This method is very robust compared to other
isolation methods: the epidermis has a consistent quality and the viability and integrity
can be fully maintained. The cleavage plane is the lamina densa, so also the basal layer is
completely intact. The duration of the treatment might be considered as a limitation as it
is an overnight procedure. However, the number of handling steps is small and an
additional advantage is that the cleavage plane is fairly independent of the treatment
duration.
Although it is sure that enzymatic digestion is by far the best method to isolate the
epidermis, sometimes additional reasons might lead to the choice for another isolation
method for certain applications. For example, the benefit of a sample with nearly perfect
geometry can be more important in an in vitro set-up than experiments on epidermis
only. Then, cutting slices of skin using a dermatome is an attractive quick method.
Applications such as mechanical characterization benefits from the fact that the natural
32
Chapter 2
pre-stress in the skin is better retained. It was also demonstrated that the viability is
retained.
When looking at the options to isolate the stratum corneum from the viable epidermis,
fewer methods are available from which only enzymatic digestion using trypsin gives
satisfying results. It should be no surprise that both cantharadin and ammonia are
harmful to the skin. Furthermore, it is without doubt that enzymatic digestion by trypsin
is the common method to isolation the stratum corneum for any application field. The
robustness of the method and, hence, consistent quality does not give the incentive to
investigate new techniques.
As it is evident that there are means to obtain intact viable sheets of epidermis or stratum
corneum alone, the next challenge is to retain these properties over time. In literature,
however, skin storage is mostly considered in relation to skin grafts used as burn wound
dressings. As a consequence, the focus is on split-thickness skin instead of isolated skin
layers, although the requirements in terms of viability and integrity are likely to be more
strict for in vitro testing than for the use of burn grafts, .
Preservation methods can be classified either based on technique, temperature or on
storage time. The latter was chosen here because in the case of in vitro testing one can
either immediately do the testing or needs to have a large batch available over a longer
time. Short-term storage can usually be done in a refrigerator. Storage in an incubator at
37°C is also satisfactory. Saline solutions certainly induce tissue damage, while various
nutrient media can keep the tissue viable and intact for at least a whole week before
degradation slowly begins. For the stratum corneum, it makes sense to use a saline
buffer. However, it is advised to do this only when using the samples within a week. For
longer storage, stratum corneum should be dried under the right conditions, as infections
and fungi may easily grow.
In the long-term, cryopreservation is a routine laboratory technique which can produce
large batches which can be beneficial for years. There is a risk of inconsistent quality of
the samples due to the sensitivity for tissue damage during thawing. When viability is
not requirement, snap freezing is a convenient and reliable method for long-term storage.
Although the variety of topics for in vitro skin research is enormous, this review has
shown that the isolation and storage protocols can be identical. Future in vitro research
should make use of isolated epidermis, which is separated by the enzyme dispase or cut
using a dermatome because of its convenient geometry. When the epidermal samples are
subsequently stored for just a short period and tissue growth is not the goal, it is advised
to use a nutrient media such as HHBSS. For long-term storage, the only option for intact
viable tissue is cryopreservation. Regarding the stratum corneum, trypsin and drying
remain by far the best methods to isolate and preserve this skin layer.
Isolation and preservation methods for the epidermis and stratum corneum
33
Acknowledgements
First of all, we would like to thank professor Bouwstra for her contribution in the
discussions. We are also very grateful to professor Hagisawa for providing the aldehydefuchsin staining procedure. Last, we would like to thank the plastic surgery department
of the Catharina hospital in Eindhoven for providing the skin tissue.
Chapter 3
Linear shear response of the upper skin
layers
36
Chapter 3
3.1 Introduction
Knowledge about the mechanical behavior of human skin is of great importance for
various clinical and cosmetic treatments. The human skin is composed of a non-uniform
layered structure and the mechanical behavior of all the layers is highly complex: i.e.
anisotropic, inhomogenous, non-linear and viscoelastic. Therefore, the most appropriate
approach seems to be to determine the mechanical properties of each individual skin
layer all loading directions in order to understand the full skin response.
The present study focuses on the contribution of the outer skin layer, the epidermis,
when in-plane forces are applied to the skin surface. Because of the anisotropic nature of
the epidermis, the response in tensile and shear are most probably very different.
Usually, tensile properties are addressed in research studies. However, the shear
component plays a key role in applications such as the development of pressure sores,
the removal of skin adhesives and skin-device contact such as with prosthetic limbs and
shavers. The collective shortcoming for all these applications is the poor knowledge of
the mechanical response of the epidermis to shear, obstructing the further improvement
of current treatments and devices.
As the epidermis is the chemical and physical barrier between the human body and its
environment, it possesses extraordinary structural properties. Epidermis is a stratified
epithelium, consisting of four different layers, defined by position, shape, morphology
and state of differentiation of the keratinocyte, the main cell type. The epidermal tissue is
renewed constantly: cells are lost from the skin surface by desquamation and this loss is
balanced by cell division and growth in the basal layer [87]. The most superficial layer,
the stratum corneum, has a thickness of typically 10-20 μm, and is considered as a
separate layer because of its specific barrier function. The stratum corneum has a „brickand-mortar‟ structure with the corneocytes, which are differentiated non-viable
keratinocytes, as „bricks‟ in a „mortar‟ of lipid membranes and desmosomes. The
thickness of the remaining part of the epidermis, the viable epidermis, ranges from 30100 μm. To strengthen the attachment of the epidermis to the dermis, the junction has an
undulating shape resulting in large cones of epidermal tissue penetrating the dermis. The
properties of both viable epidermis and stratum corneum are influenced by
environmental conditions such as temperature (T) and relative humidity (RH).
Usually, load-bearing soft tissues are composed of a fiber network, providing the
strength and elasticity to the tissue, but this is not the case for the epidermis. Its
extensibility is mainly due to the possibility to smooth out the skin surface, while the
strength and cohesiveness are due to the rigid tonofilament cytoskeleton and the
numerous desmosomes at the periphery of the keratinocytes. Furthermore, the viable
epidermis is a very compact tissue; the intercellular spaces occupy less than 2% of the
volume [5,13]. Consequently, the viable epidermis is suspected to be more rigid than
other soft tissues. In the stratum corneum, the cellular membranes are thickened, the
Linear shear response of the upper skin layers37
water content is decreased and a larger amount of keratin is present and thus, its
mechanical stiffness and strength are suggested to be even higher.
Due to the complex skin structure, the mechanical response of the epidermis cannot be
easily distinguished from that of the dermis in an in vivo experiment. This results into
two important implications for mechanical characterization of epidermis: 1) skin layers
need to be measured individually, and 2) in vitro measurements are required. Regarding
the first implication, stratum corneum and the entire epidermis can be isolated from other
skin layers, but there are no means to isolate viable epidermis. So both isolated and
combined skin layers need to be characterized to assess the mechanical response of the
viable epidermis. Furthermore, in vitro measurements opens up a broad range of reliable
standard techniques used in mechanical engineering. Nevertheless, these methods need
to be adapted to enable the measurement of thin layers of soft materials. Moreover,
issues regarding the complex sample geometry, the heterogeneous tissue composition
and the sensitivity to environmental conditions have to be dealt with.
Currently, there is a paucity of papers describing mechanical properties of the entire
epidermis or viable epidermis only. Studies so far were either on a small-sized scale
[37,88], not reproducible [89] or included the total papillar dermis [38] and none of them
investigated the shear response. Mechanical properties of stratum corneum have been
studied and reviewed more extensively [34,90,5,21]. However, also for the stratum
corneum, very few studies are investigating shear properties. Consequently, quantitative
shear data for the upper skin layers is sparse or not existent. It is hypothesized that the
shear modulus of the epidermal layers is far below the broad range of tensile Young‟s
moduli found in literature because of the anisotropic structure of epidermis.
We measured the mechanical behavior of various human skin layers subjected to shear
over a wide frequency range and with varying environmental conditions, i.e. temperature
and relative humidity (RH). Because of the complexity, we limit ourselves in this study
to determine the small strain behavior of stratum corneum and viable epidermis. To
validate the experimental approach, also tests with silicone rubbers are performed.
3.2 Methods
3.2.1 Sample preparation
Skin
Skin was obtained from patients undergoing abdominoplastic surgery, who gave
informed consent for use of their skin for research purposes under a protocol approved
by the ethics committee of the Catharina Hospital, Eindhoven, The Netherlands. Only
abdominal skin of Caucasian women from a age group between 35 and 55 years old is
used. Abdominal skin with stria, cellulite, damage due to UV exposure or tremendously
hairy skin is excluded from the study.
38
Chapter 3
Immediately after excision, the skin is brought into the laboratory and processed within 4
hr. Skin slices are obtained using an electric dermatome (D42, Humeca, The
Netherlands) of which the set thickness was refined for this purpose by the supplier. In
order to separate the epidermis, the thickness is set to 100 μm. Subsequently, circular
tissue samples of the epidermis are obtained from the slices using an 8 mm diameter cork
borer. The epidermis is estimated to vary from 50 to 150 μm on this body site [12,5].
Depending on various factors such as skin surface roughness, tissue hydration,
smoothness of the cutting, some papillar dermis could remain attached (Figure 3.1).
To obtain stratum corneum, dermatomed skin slices of 300 μm are also punched into 8
mm diameter samples before immersion in a solution of 0.1% trypsin (SV30037.01,
Hyclone) in PBS at 37°C for 2-3 hr. Thereafter, samples are rinsed with PBS.
Also split-thickness skin of 200 and 400 μm in thickness is obtained using the
dermatome. As can be seen in Figure 3.1, the 200 μm split-thickness skin is composed of
epidermis and papillar dermis. In the 400 μm split-thickness, also reticular dermis is
included. For isolating the reticular dermis, the top layer of skin is dermatomed until the
white opaque dermis is on top. Then, a 400 μm thick layer of reticular dermis is
dermatomed.
The stratum corneum samples were stored in PBS at 4°C for maximal 7 days but dried
when longer storage is needed. All other samples were stored in a Hank‟s Hepes
Balanced Salt Solution (HHBSS) for a maximum of 72 hrs in an incubator until use. The
viability of the samples was determined by a standard colometric MTT (Thiazolyl Blue
Tetrazolium Bromide) assay. The tests proved that the tissue viability does not change
after a storage period of 72 hours (data not shown).
Silicone rubber
In order to validate the experimental approach, a highly elastic silicone rubber
(Köraform 42 A , Alpina Siliconee, Germany) was chosen. The silicone rubber was
poured under vacuum into various thicknesses: 0.05, 0.12 and 2.00 mm. Circular
samples were obtained by using an 8 mm diameter cork borer.
3.2.2 Experimental set-up
All experiments are performed on a rotational rheometer (ARES, Rheometric Scientific,
USA) with parallel plate geometry in combination with a Peltier environmental control
unit and a fluid bath. Plates are sand-blasted to prevent slippage. An eccentric
configuration is used, where the sample is placed at the edge of the plate with a radius of
33 mm (Figure 3.2), allowing for the measurement of soft tissues [91,92,93]. The shear
stress 𝜏 and shear strain 𝛾 are then calculated from the measured torque 𝑀 and the angle
𝜃 by:
𝜏
=
2𝜋𝑟12
𝑀𝑟
𝑟 − 𝑟1
2
2
+
𝑟12
8
,
𝑟
𝛾=𝜃 ,
ℎ
(3.1)
Linear shear response of the upper skin layers39
SC
VE
PD
(b)
(a)
RD
(c)
(d)
Figure 3.1: Histological cross-sections of dermatomed skin: (a) 100 μm split-skin with
stratum corneum (SC) and viable epidermis (VE), (b) 100 μm split-skin containing
epidermis and some papillar dermis (PD), (c) 200 μm split-skin consisting of epidermis
and papillar dermis, (d) 400 μm split-skin including reticular dermis (RD).
M
r1
r
Figure 3.2. Eccentric configuration for rotational shear experiments. A sample with radius
𝒓𝟏 is rotated at a radius 𝒓 with a torque 𝑴. The groove following the perimeter facilitated
the positioning of the samples.
40
Chapter 3
where 𝑟 is the radius of the plate, 𝑟1 is the sample radius and ℎ is the sample height. The
advantages of shifting the sample to the edge of the plate are that the measured torque
signal is increased and the deformation is more homogeneous than in the conventional
centered configuration.
Samples are gently placed in the correct position by using tweezers. In order to spread
out the stratum corneum sample, a droplet of PBS is placed in which the stratum
corneum sample unfolds. Subsequently, the droplet is extracted by using a tissue. The
other skin samples can be placed using tweezers only. Visible droplets on the surface of
all sample types are gently removed. Next, the upper plate is lowered until the sample
experienced the intended normal force.
Samples are measured in a controlled environment using a home-built system, see Figure
3.3. Via a pressure switch, dry compressed air enters two channels, of which one channel
conducts through a chamber filled with water to obtain fully humidified air. In the next
chamber, dry and fully hydrated air are mixed to obtain the desired RH by regulating the
flow inlets. The mixing chambers are placed in a water bath to control the temperature.
Finally, the air is brought via a temperature controlled tube (HT 20, Horst GmbH,
Germany) into the measurement chamber, in which the temperature is controlled
through the air inlet as well as via the bottom plate by the Peltier environmental control
unit. A RH/T-sensor (Hytemod-USB, Hygrosense Instruments GmbH, Germany) is
located near the sample.
G
F
E
A
D
C
B
Figure 3.3: Measurement set-up. Pressurized air goes via the pressure switch (A),
whereafter the air is split up into two tubes, passes flow regulators (B) and flow meters
(C), before entering the humid and/or mixing chambers in the waterbath (D). Then, the air
goes via a temperature-controlled tube (E) into the measurement chamber of the
rheometer (F), where a RH/T-sensor (G) is giving feedback about the actual RH and
temperature.
Linear shear response of the upper skin layers41
3.2.3 Rheological methods
Linear viscoelastic material behavior is described by a multi-mode Maxwell model:
𝜏𝑖 +
1
𝜏
𝜆𝑖 𝑖
= 𝐺𝑖 𝛾 ;
𝑖 𝜖 1, 𝑛
(3.2)
where 𝜏𝑖 is the shear stress contribution of mode 𝑖 with the relaxation time 𝜆𝑖 and
modulus 𝐺𝑖 . The applied strain rate is denoted with 𝛾 . The total stress (𝜏) is the sum of
the stress contributions of all modes:
𝑛
𝜏=
𝜏𝑖
(3.3)
𝑖=1
A frequency (𝜔) dependent input 𝛾 = 𝛾0 sin 𝜔𝑡 will lead, for linear viscoelastic
behavior, to a sinusoidal shear stress:
𝜏 = 𝐺𝑑 𝛾0 sin(𝜔𝑡 + 𝛿),
(3.4)
where 𝐺𝑑 (𝜔, 𝑇) is the dynamic modulus and 𝛿(𝜔, 𝑇) the phase shift. The response can
be written in an in-phase and out-of-phase wave:
𝜏 = 𝜏 ′ + 𝜏′′ = 𝜏0′ sin 𝜔𝑡 + 𝜏0′′ cos 𝜔 𝑡
(3.5)
From this, the moduli can be computed:
𝑛
′
𝐺 =
𝜏0′
𝛾0 =
𝑖=1
𝑛
𝜆2𝑖 𝜔2
𝐺𝑖
1 + 𝜆2𝑖 𝜔 2
𝐺 ′′ = 𝜏0′′ 𝛾0 =
𝐺𝑖
𝑖=1
𝜆𝑖 𝜔
1 + 𝜆2𝑖 𝜔 2
(3.6)
(3.7)
where 𝐺 ′ is the storage modulus, representing the elastic part of the behavior and 𝐺 ′′ is
the loss modulus, representing the viscous behavior. The two moduli, 𝐺 ′ and 𝐺 ′′ , form
the dynamic shear modulus:
𝐺𝑑 =
𝐺′2 +𝐺′′2
The phase shift 𝛿
to 𝐺′ and 𝐺" via;
(3.8)
42
Chapter 3
𝐺 ′′
tan 𝛿 =
𝐺′
(3.9)
3.2.4 Experimental procedures
The ultimate goal of this study is to determine the loss and storage moduli of stratum
corneum and viable epidermis as a function of frequency, temperature and relative
humidity (RH). If skin layers can be isolated, they are measured separately. If not,
measurements are performed on combinations of skin layers. In order to determine the
mechanical parameters, the linear viscoelastic strain regime, i.e. the strain range in which
the material properties are independent of the strain amplitude, has to be identified.
Moreover, the typical characteristics of the upper skin layers makes that other
preliminary tests are essential, i.e. the right experimental conditions needs to be defined
to ensure reliable results.
First of all, the samples, especially the stratum corneum samples, are extremely thin
(under 20 μm). Measuring such thin samples is at the limit of the possibilities of the
apparatus used. Therefore, to validate the experimental approach, a well-defined
homogeneous soft material, i.e. silicone rubber, with different thicknesses was tested.
Also, the approach of using a stack of layers to increase the sample thickness was tested.
Furthermore, the natural wrinkling shape of the thin sample (see Figure 3.1) may cause
contact problems between the sample and the parallel plates. Flattening the wrinkles may
reduce these contact problems. Therefore, the influence of the level of normal force
applied was determined for various numbers of stratum corneum samples stacked. Last
but not least, the sensitivity of the upper skin layers to its environment needs to be
translated into conditioning times, i.e. the times required for stationairy mechanical
behavior.
The experimental procedures for the topics mentioned are discussed in the order given
below. For each experimental procedure is stated which types of samples are used.
 validation of the experimental approach (silicone rubber)
 stacking (stratum corneum)
 determination of the linear viscoelastic strain regime (stratum corneum,
epidermis, epidermis + papillar dermis, epidermis + dermis, reticular dermis)
 determination of the conditioning time (stratum corneum, epidermis)
 determination of linear viscoelastic properties over a frequency range as a
function of temperature and humidity (stratum corneum, epidermis).
Validation of the experimental approach
In order to validate whether the experimental method applies for thin samples using this
measurement set-up, experiments are conducted on silicone rubber samples with varying
thickness but similar diameter (8 mm). The shear modulus is determined for various
frequencies increasing stepwise from 1 to 100 rad/s at 0.01 strain.
Linear shear response of the upper skin layers43
Stacking
A possible way to resolve the problem of thin samples with a complex wrinkled sample
geometry is to stack a few of these on top of each other. This approach is checked for 1,3
and 5 layers of dried stratum corneum, respectively. First, the dried samples are
conditioned at room temperature for 1 hr. The normal force is varied between 1-10 g,
measuring the corresponding thickness and the shear modulus at 10 rad/s and 0.01 strain
at the same time. The measurements are performed at room conditions (50% RH, 22°C).
As the other skin layers are thicker and more pliabble than stratum corneum, it is
assumed that the space between the plates is better filled up and that the skin surface
roughness is negligble. A normal force of 1 g is applied on these samples.
Linear viscoelastic strain regime
The linear viscoelastic strain regime can be determined using oscillatory shear
experiments with constant frequency and varying strain (strain sweep). The strain
sweeps are performed at 10 rad/s for strains varying from 0.001 up to 0.1 at room
conditions (50% RH, 22°C) on all skin sample types: e.g. stratum corneum, epidermis,
epidermis and papillair dermis, epidermis and dermis and reticular dermis only. As the
samples are already placed in the room for over 1 hr, it is assumed for now that 20
minutes conditioning in the closed chamber of the measurement set-up prior to the start
is sufficient. Samples consisting of only reticular dermis are measured in a humid
environment to prevent dehydration.
Conditioning times
Conditioning times are derived from oscillatory shear experiments with a strain of 0.01
at 10 rad/s for 1 hr at various RH at 22°C. These time sweep series are performed on
stratum corneum and epidermis. Data points are collected every 30 s.
Determination of linear viscoelastic properties
The previous tests should prove that the experimental approach enables the measurement
of the small strain behavior of epidermis and stratum corneum. As a result, frequency
sweeps ranging from 0.1-100 rad/s at 0.01 strain can be applied at 25%, 50%, 75% and
98% RH and 22°C and 37°C. Conditioning time varies from 20 minutes at 25% RH, 35
min at 50% RH and 75%RH, up to 45 min for 98% RH.
3.2.5 Histological examination
Histological examination of the used samples provides a control for the thickness
measurements with the rheometer, the skin layer composition and for abnormalities. All
samples are fixated in 10% phosphate-buffered formalin after mechanical testing and
processed for an aldehyde-fuchsin staining. The sections are cut at 5 μm and stained with
aldehyde (Merck) and light-green (Merck). The tissue morphology and sample thickness
is studied by light microscopy.
44
Chapter 3
3.3 Results
For all tests, the linear viscoelastic behavior is presented in terms of the shear modulus,
𝐺𝑑 , and the phase angle, 𝛿. Because it appeared that 𝛿 remains constant for all measured
conditions, these data are not always displayed.
Validation of experimental approach
In order to prove that the experimental approach is appropriate for thin samples,
frequency sweeps were applied for silicone rubbers of varying thickness. The results are
shown in Figure 3.4. No significant differences between the storage modulus 𝐺 ′ and the
loss modulus 𝐺" for the various thicknesses are measured. When assuming NeoHookean
material, such that 𝐸 = 3𝐺𝑑 , the derived Young‟s modulus is also similar to those
obtained from tensile testing. It is conluded that the experimental approach is appropriate
for measuring thin, soft materials.
Figure 3.4: Frequency sweeps performed on silicone rubber of various thickness: 0.050,
0.120 and 2.00 mm.
Stacking
In this test, stratum corneum samples were studied. As shown in Figure 3.5a, increasing
the force from 1 to 3 g results in large differences for the measured gap in the
measurement set-up, indicating that the wrinkling surface is unfolded. Increasing the
force from 3 g up to 10 g causes relatively small deformations, indicating compression.
Thus, a normal force of 3 g applied on one stratum corneum sample of 8 mm in diameter
should provide sufficient contact between the sample and the parallel plates. The
constant value of the shear modulus at this normal force in relation to the number of
stacked samples supports this conclusion (see Figure 3.5b).
Linear viscoelastic strain regime
As shown in Figure 3.6, the linear viscoelastic strain regime is similar for stratum
corneum, epidermis, dermis and split-thickness skin. For all those skin types, it is
observed that the shear response is independent of the applied shear strain until nearly
0.01. As the conditioning time for epidermis and stratum corneum could only be
estimated during this test, the measured value of 𝐺𝑑 might slightly differ from the actual
𝐺𝑑 when those skin layers are involved. Therefore, data shown are normalized.
It should be noted that the value of 𝐺𝑑 for the reticular dermis is much less than for skin
samples including epidermis. Furthermore, the measured gap could deviate more than
50% from the set thickness of the dermatome for samples containing epidermis and
Linear shear response of the upper skin layers45
dermis (not shown). However, histological examination showed that the composition of
those skin samples is in agreement with the expectations.
(a)
(b)
Figure 3.5: The effect of stacking dried stratum corneum samples: (a) total sample
thickness vs. the measured gap at various normal forces: the dotted line represents the
linear relationship between gap and number of stratum corneum (SC) samples stacked; (b)
the shear modulus at varying axial forces vs the number of SC layers at a frequency of 10
rad/s: the dashed line represents the average of the measurements using an normal force of
3 g ().
Figure 3.6: The normalized 𝑮𝒅 of the average results of strain sweeps performed on various
skin layers. For each skin layer, 3 samples from each of the 3 specimens were tested.
Conditioning
To reduce measurement time, the conditioning times were identified for epidermis. Since
the thicker epidermis needs more time to adjust to a certain temperature and humidity, it
is assumed that its conditioning time also holds for stratum corneum. The results of the
time sweeps are depicted in Figure 3.7. At low RH, the mechanical response is stabilized
within 20 minutes. Since hardly any difference is observed between the settling times for
50% and 75% RH, both conditioning times are set at 30 minutes. At 98% RH, the moduli
46
Chapter 3
slightly decrease until about 40 min. Therefore, fully hydrated skin samples are
preferably conditioned for 45 minutes. A considerably increase in the standard deviaton
for the higher RH was noted.
Figure 3.7: Average values of 3 measurements for 𝑮𝒅 (𝝎 = 𝟏𝟎𝒓𝒂𝒅/𝒔, 𝑻 = 𝟐𝟎°𝑪) and the
standard deviation over time (dotted lines) for the epidermis at various RH. The vertical
grey band indicate the necessary conditioning time.
Determination of G’ and G”
The dependency on RH and temperature were measured for both the epidermis and
stratum corneum (Figure 3.8 and 3.9). For each RH/T combination, it was aimed to
measure 3 samples per subject. However, the test sequence could not be completed for
subject 2 and 4 within 72 hr. Subject 4 was therefore excluded from the study on
epidermis.
For both stratum corneum and epidermis, the modulus is slightly frequency dependent
(see Figure 3.8). The phase angle is not significantly different for the various RH. As
similar results were obtained for epidermis and stratum corneum, only results from the
latter are shown in Figure 3.8. Because of this mild frequency dependency, a comparison
between the different environmental conditions could be done at one frequency only. In
this case, we chose 10 rad/s (see Figure 3.9 and Figure 3.10). The results for stratum
corneum show a decrease in modulus with increasing humidity, but no correlation with
temperature. For the epidermis, data were less consistent, especially at 20°C. It is
suggested that its mechanical properties are independent for small temperature changes.
Linear shear response of the upper skin layers47
(a)
(b)
Figure 3.8: Linear viscoelastic behavior of stratum corneum from one subject for various RH
at 20°C. (a) The average shear modulus 𝑮𝒅 ; (b) The average phase angle δ.
Figure 3.9: Linear viscoelastic behavior of the stratum corneum for various RH at 20°C and
37°C. The average values and standard deviations are shown for 𝑮𝒅 and δ per subject.
48
Chapter 3
Figure 3.10: Linear viscoelastic behavior of the epidermis for various RH at 20°C and 37°C.
The average values and standard deviations are shown for 𝑮𝒅 and δ per subject.
3.4 Discussion
In the past, mechanical behavior of epidermis has been described qualitatively due to the
lack of experimental data. The skin curvature and the undulating dermal-epidermal
junction cause inherent difficulties during mechanical characterization of the epidermis
in vivo. In addition, in vivo measurement methods for shear, such as elastography,
cannot be applied (yet) due to limitations in resolution. Therefore, this study presents an
in vitro measurement method to determine shear properties of the epidermis. Preliminary
testing was essential to validate the methods applied and to obtain the right experimental
conditions allowing for reliable final measurements.
In order to measure shear properties of a soft biological material, measurement methods
as developed for muscle, brain and thrombus [91,92,93] could be used. However, pretesting was needed to prove that the experimental approach is also appropriate for
samples within the order of a few micrometer, while retaining a relative large diameter to
avoid the effect of local properties. As there are inherent difficulties in determining the
actual thickness of stratum corneum from histology, the sample thickness was defined by
the measured gap between the plates. Although the thickness of the stratum corneum on
the abdomen is reported to be 14±4 μm [5], the measured gap at a normal force of 3g
varied from 15 to 60 μm due to local variations and skin surface roughness. Applying a
Linear shear response of the upper skin layers49
higher normal force causes compression of the sample, which influences the measured
modulus, and, therefore, should be avoided.
The skin surface roughness becomes less significant for the other thicker skin layers. In
addition, these layers are more pliable than stratum corneum. Whether stratum corneum,
epidermis only or epidermis and dermis together are measured, the shear response does
not differ significantly for the strain sweeps. It is hypothesized that loading in shear
causes cell deformation in the epidermis while hardly affecting the desmosomes. As inplane shear hardly effects the transversal dermal fibers as well and few cells are present
in the dermis, the dermal response will be mainly due to ground substance deformation.
It is likely that this substance has a lower shear resistance than the highly organized
epidermis. In contrast, the tissue response on in-plane tensile loading will be determined
by the strength of the desmosomes, the elasticity of the dermal fibers and the direction of
the Langer lines.
Recently, the linear viscoelastic response on oscillatory shear strains of human whole
skin and dermis-only was measured [94,95]. The increase of the moduli was more
pronounced for the dermis-only at higher frequencies, so the authors concluded that the
epidermis is only slightly frequency dependent. At lower frequencies, 𝐺𝑑,𝑑𝑒𝑟𝑚𝑖𝑠 was in
the order of kPa. In accordance to this study, we observed in our frequency sweeps that
the epidermis is indeed slightly frequency-dependent. Our strain sweeps also resulted in
a value for 𝐺𝑑 ,𝑑𝑒𝑟𝑚𝑖𝑠 of a few kPa.
For the stratum corneum, the values of 𝐺𝑑 are similar to these of the epidermis, but might
be somewhat differented when corrected for uncertainties in the sample thickness. The
results for epidermis and stratum corneum suggest that the small strain shear properties
of viable epidermis and stratum corneum hardly differ. Currently, our shear moduli can
only be compared with in-plane tensile properties of stratum corneum from literature.
When doing so, our shear moduli are one order of magnitude below the properties in dry
conditions and up to two orders of magnitude when fully hydrated when using the lowest
reported values of the Young‟s moduli [22,24,96]. This clearly supports the highly
anisotropic behavior of stratum corneum and epidermis.
A decrease in stiffness of the stratum corneum could be observed with increasing RH. In
accordance with our observations, delamination studies with stratum corneum, which
also showed the pre-failure mechanical response, showed no temperature-dependence for
this temperature range [97].
No clear relationship between the mechanical properties of the epidermis and RH could
be established. Time sweeps showed that moduli set after a certain conditioning time.
However, both time sweeps and frequency tests for epidermis showed larger variations
per RH and per subject compared to stratum corneum. This might be related to the less
well-defined tissue composition. For example, the direction of Langer lines or
irregularities such as sweat pores and hair follicles can have a more substantial role in
the mechanical behavior in fully hydrated epidermis than for stratum corneum. Future
experiments should clarify the variance in these results.
Longer conditioning times and larger variations were observed in fully hydrated
epidermal samples than for less humid samples. Examination of fully hydrated SC
50
Chapter 3
structure has revealed swollen corneocytes and water pools in the extracellular spaces
after storage in PBS [98]. Furthermore, water disrupts the lipid lamellae to varying
degrees and causes degradation of intercellular corneosomes [87,87,99]. It is likely that
the desmosomes in the viable epidermis are also highly susceptible to damage. However,
histological examination did not show any sign of degradation in our samples. The
prolonged time of conditioning the sample in the set-up at higher RH limited the number
of experiments that could be done with epidermis from one donor within 72 hours.
The present study demonstrated that reproducible results can be obtained for the shear
properties of epidermis in an in vitro set up. Viable epidermis could not be measured as
an isolated skin layer, but its properties can be derived from the other skin samples
tested. The 𝐺𝑑 for stratum corneum roughly ranges from 4 to 12 kPa, decreasing with
increasing RH. The values are far below the shear modulus value based on tensile
Young‟s moduli (i.e. 𝐸 = 𝐺𝑑 ) in literature, assuming anisotropic material behavior.
Results for the epidermis were in the same order of magnitude, but was less consistent. A
reason might be the less well-defined tissue composition. Therefore, it would be
interesting to combine mechanical testing with real-time imaging techniques to follow
changes in tissue deformation. It was already shown by histological examination after 2
days of loading that shear forces induce cell displacement in skin, and particularly in the
epidermis [100].
Furthermore, electron microscope imaging techniques could support histological
examination in assessing tissue damage due to preparation, storage or handling. Last but
not least, the shear response needs not only to be correlated with the tensile loading but
also with the effects of perpendicular loading, such as indentation or compression.
Chapter 4
A new indentation method to determine
the mechanical properties of epidermis
52 Chapter 4
4.1 Introduction
The outer skin layer possesses important characteristics that make them a favorable site
for pain-free drug delivery with minimal damage: a rich population of immunologically
sensitive cells as well as the lack of blood vessels and sensory nerve endings. Today, the
development of drug delivery using microneedles or microjets is challenging. because of
the poor understanding of the mechanical behavior of the human skin layers. In
particular, the key mechanical properties of the outer skin layer, i.e. the epidermis
composed of stratum corneum and viable epidermis, should be better understood.
The structure and function of this layer are well-known [10]. The outer layer, the stratum
corneum. is an effective physical barrier of dead cells in a „brick-and-mortar‟ structure:
the anucleate corneocytes form „bricks‟ and the intercellular lipid membranes and
corneosomes are considered as mortar. The viable epidermis mainly consists of
migrating keratinocytes towards the stratum corneum, continuously changing in
composition, shape, and function. The junction with the underlying dermis is
strengthened by its undulating pattern such that large cones of epidermal tissue penetrate
the dermis (see Figure 4.1). Furthermore, epidermal properties are influenced by
environmental factors such as temperature, humidity and UV radiation.
In order to deliver drugs transdermally, the microneedle or nozzle should penetrate the
stratum corneum to deliver the drug 100-150 μm below the skin surface, e.g. in the
viable epidermis or papillar dermis. It is therefore important, besides penetration studies,
to investigate the mechanical behavior of epidermis to understand the delivery path
through the epidermis as well as the tissue repair and remodeling mechanisms associated
with the treatment. Until today, however, studying mechanical properties of skin was
limited to dermis and stratum corneum, ignoring viable epidermis.
As sharp indentation leads to the penetration of a microneedle or nozzle, the specific
interest in this study are the mechanical properties of the epidermis during indentation at
a micron lengthscale. Recently, Kendall et al. [37] were the first publishing mechanical
properties of the (viable) epidermis during penetration, using modified standard tips on
murine skin. They observed a decrease in storage modulus when the 2 μm probe
penetrates through the stratum corneum, which is in accordance with studies on stratum
corneum only [101,102]. The authors argued that this is because of an increasing
moisture content with depth. In the viable epidermis, the storage modulus remained
nearly constant. In contrast, penetration of the 5 μm probe showed a negligble decrease
in storage modulus throughout the stratum corneum and a gradual increase in the viable
epidermis but still below the shear modulus values for the 2 μm probe.
A variety of in vivo and in vitro indentation techniques were developed to measure the
stratum corneum. In the eighties, Hendley et al. developed an indentation device to
measure force variations in vivo due to age, sex and body site [102]. A needle of 11 μm
at the tip was held perpendicular to the surface and moved rapidly into the skin. They
claimed that the speed of the indentation ensures that predominantly stratum corneum
properties are tested [33]. Measured forces were typically in the order of 3.0 N. Recently,
A new indentation method to determine the mechanical properties of epidermis 53
some nano-indentation studies have been performed on isolated stratum corneum
[103,31,104,104,101]. The tips used varied between 1-10 μm, while corneocytes have a
diameter ranging from 26-45 μm [105,3,106]. As a consequence, very local properties
are determined in experiments using those tips. Furthermore, in some of the studies, the
three-sided Berkovich tip that come to a sharp point is used. This tip easily induce
damage on the sample‟s surface, which interferes with the load-displacements results of
the indentation. Three of the nano-indentation studies were based on continuous stiffness
measurements (CSM) protocols [37,101,104]. The drawback of CSM is that the results
are influenced by the selected amplitude and frequency for viscous materials. Taken all
nano-indentation studies on stratum corneum together, the measured Young‟s moduli
varied from 10 MPa [107] for wet porcine samples up to 1 GPa for dried human samples
[101]. This broad range is likely caused by the differences in testing apparatus and
protocols, differences between species and body sites, and the heterogeneity of the
material. A reliable method to determine the mechanical properties of the stratum
corneum only on the tissue level is therefore also required.
The objective of the present study is to determine Young‟s modulus of the epidermis,
e.g. the stratum corneum and viable epidermis, by means of a micro-indentation test. The
typical complex geometry, undulating and less than 150 μm in thickness, and the
porosity of the epidermis put high demands on the mechanical characterization.
Therefore, isolated epidermis and isolated stratum corneum were tested using equipment
that is known for its accuracy and reliablity. As the used device was originally designed
for solid materials of which well-defined samples can be obtained, the protocol was
adapted to be applicable. To validate that the testing protocol holds for thin materials
with a low stiffness, tests have been performed with silicone rubber.
4.1.1 Sample preparation
Skin
Indentation tests have been carried out on ex vivo abdominal skin of Caucasian women
from a similar age group (43±4 years old) undergoing abdominoplasty surgery. All
patients gave informed consent for use of their skin for research purposes under a
protocol approved by the ethics committee of the Catharina Hospital, Eindhoven, The
Netherlands. Abdominal skin with striae markers, cellulite, damage due to UV exposure
or excessively hairy skin is excluded from the study.
Immediately after excision, the skin is brought into the laboratory and processed within 4
hours. Epidermal sheets were obtained using a dermatome (D42, Humeca) of which the
set thickness was refined for this purpose by the supplier. The dermatomed slices of 100
μm thickness were cut in pieces of approximately 1 cm2. Depending on various factors
such as skin surface roughness, tissue hydration, and the amount of cones and ridges,
samples may consist of epidermis and/or some papillar dermis [Figure 4.1].
To obtain stratum corneum samples, dermatomed skin slices of 200 μm were immersed
in a solution of 0.1% trypsin (Hyclone, SV30037.01) in an incubator at 37°C for 2-3 hr.
54 Chapter 4
Thereafter, the sheets were rinsed in PBS and also cut into pieces of approximately 1
cm2.
All samples were stored at -80°C until further use.
SC
SC
VE
VE
(b)
PD
SC
RD
VE
PD
(a)
(c)
Figure 4.1. An aldehyde-fuchsin staining is used to visualize the morphology of the
various skin layers: (a) Full-thickness skin including the stratum corneum (SC), viable
epidermis (VE), papillar dermis (PD) and reticular dermis (RD); (b) Dermatomed skin
with a set thickness of 100 μm consisting of the epidermal layer only; (c) Dermatomed
skin of 100 μm consisting of epidermis and some fragments of papillar dermis.
Silicone rubber
In order to validate that the experimental procedure is valid for thin samples, a highly
elastic silicone rubber (Köraform 42 A, Alpina Siliconee, Germany) was measured using
various sample thicknesses. The silicone rubber was poured under vacuum into various
thicknesses: 0.05, 0.12 and 2.0 mm. Then, samples of about 1 cm2 were cut out.
4.1.2 Experimental procedure
Skin
The skin sample was placed on a substrate such that in-plane tissue movement cannot
occur. The large number of pores in the epidermis hardly allowed any fixation method. It
appeared, however, that the adhesive, sticky behavior of the skin sample is strong
enough so no other fixation was required. Immediately after thawing at room
temperature, samples were spread out on an aluminum disc with the outer skin surface
facing up. Possible air or liquid below the tissue was gently squeezed out. The samples
were allowed to acclimatize for 20 min before the first indentation commenced.
On each skin sample, nine indentation locations were manually selected with use of the
built-in microscope of the NanoIndenter XP (MTS Systems, USA). Each location is at
least 500 μm away from the others to avoid that measurements influences each other.
A new indentation method to determine the mechanical properties of epidermis 55
The top center of the triangles formed by the glyphics, i.e. the primary and secondary
lines, is chosen as indentation location to optimize the contact between the indenter and
the tissue [Figure 4.2].
All experiments are performed using a sapphire sphere with a radius of 500 μm. The
system has load and displacement resolutions of respectively 1 nN and 0.0002 nm. The
maximum load depends on the depth limit of indentation, which was set to be maximal
10% of the sample thickness [108]. Preliminary testing demonstrated that this indicates a
maximum load of 0.2 mN for stratum corneum and 1 mN for epidermis. The
loading/unloading rate was 0.01 mN/s. The maximum load was held for a period of 30 s.
The low stiffness of the skin samples required a low surface approach sensitivy and
contact stiffness. For both epidermis and stratum corneum, the protocol was repeated on
three samples for each subject. Test series were completed within 2 h.
Within the laboratory, the temperature and humidity are kept constant at 22°C and 28%
RH, respectively.
Silicone rubber
The skin samples, and especially the stratum corneum samples, are extremely thin (under
20 μm). Measuring such thin samples might be at the limit of the possibilities of the
apparatus used. Therefore, to prove the usefulness of the protocol for thin materials, a
well-defined homogeneous soft material, the silicone rubber, with different thicknesses
was tested with the indentation protocol similar to that for the epidermis. The samples
were placed on the substrate without fixation. Indentation locations were pointed
automatically, using a 3x3 grid with a distance of 500 μm between the various locations.
1.8 mm
Figure 4.2. The top center of the triangles, highlighted by the large red points, formed by
the glyphics was chosen as indentation location on the skin samples.
4.1.3 Determination of the Young’s modulus
Analytical approach
In order to derive a first estimate of the Young‟s modulus, the experimental data of the
skin and silicone rubber samples are analysed by the method proposed by Oliver and
56 Chapter 4
Pharr [108], assuming a fully elastic response upon unloading. From the initial unloading
slope of the load-displacement (𝑃, ℎ) curve, the reduced modulus Er is obtained:
𝐸𝑟 =
𝜋 𝑑𝑃/ℎ
2
𝐴
(4.1)
where 𝐴 is the contact surface. The measured displacement of the tip is in practice hardly
ever equal to the contact depth, because at the vicinity of the tip, the surface can sink-in
or pile-up (see Figure 4.3). For the special case of frictionless contact of a spherical
indenter with a flat linearly elastic half space, the projected contact area 𝐴𝑝 can be
calculated for small deformations according to:
𝐴𝑝 = 𝜋𝑎2 = 𝜋(2𝑅 − ℎ𝑐 )ℎ𝑐
(4.2)
Subsequently, the Young‟s modulus is calculated following:
1
1 − 𝜈 2 1 − 𝜈𝑖2
=
+
𝐸𝑟
𝐸
𝐸𝑖
(4.3)
where E and ν are the Young‟s modulus and the Poisson‟s ratio for the specimen and Ei
and νi are the same parameters for the indenter. The epidermis, stratum corneum and
silicone rubber are considered to be close to incompressible materials, using a Poisson‟s
ratio of 0.495.
Sink-in
Pile-up
a
hp
h
a0
hc
hp
a
h
hc
Figure 4.3: Contact profile developed during indentation where 𝒉 is the indentation depth,
𝒉𝒄 is the contact depth, and 𝒂 is the radius. Obtained from Pelletier et al. [109].
Numerical model
To be able to compare the estimated Young‟s moduli via the analytical method, a finite
element calculation using MSC.Marc was introduced. An axisymmetric mesh was used
to fit the experiments using a Neo-Hookean model, assuming incompressible material
behavior. The mesh consisted of 4329 linear quad4 elements, using full integration. The
size of the mesh was chosen such that the edges do not influence the stress distribution
and contact between the indenter and the sample was assumed to be frictionless.
A new indentation method to determine the mechanical properties of epidermis 57
For the silicone rubbers, the Young‟s modulus, 𝐸𝑆𝑅 , was estimated by fitting the average
load-displacement curve of the 50 μm thick samples. The value for 𝐸𝑆𝑅 was then used to
calculate the unloading curves of the 120 and 2000 μm thick samples. These unloading
curves are compared with the experimental data.
Since the deformations were small, linear elastic behavior was also assumed for the skin
samples. Furthermore, the thickness of the stratum corneum was considered to vary from
10 to 20 μm. The thickness of the viable epidermis was kept constant at 80 μm. First, the
Young‟s modulus for the stratum corneum, 𝐸𝑆𝐶 , was derived by fitting the average loaddisplacement curve of the stratum corneum samples. The obtained modulus for stratum
corneum, 𝐸𝑆𝐶 , was used to fit the experimental data of the epidermis, such that the
modulus for the viable epidermis, 𝐸𝑉𝐸 , could be derived. In order to assess the sensitivity
of the fitting approach, the effect of increasing or decreasing 𝐸𝑆𝐶 with a factor 2 on the
maximum indentation depth was studied for the epidermis.
4.2 Results
Silicone rubber
The load-displacement curves obtained from the silicone rubber samples are shown in
Figure 4.4. The results were highly reproducible for each thickness. The maximum
indentation depth slightly decreases with decreasing sample thickness. Consequently, the
slope of the initial unloading curve decreases, which is reflected in the average values for
the Young‟s moduli using Oliver & Pharr: i.e. 3.67±0.20, 2.22±0.10 and 1.69±0.04 MPa
for a sample thickness of 50, 120 and 2000 m, respectively.
From the FE model, the Young‟s modulus was estimated to be 2.16 MPa. When using
this values to obtain the unloading curves for the 120 and 2000 m thick sample, it can be
shown that the unloading curves and maximum indentation depth for all thicknesses are
comparable to the experimental data.
Figure 4.4. All force-indentation (𝑷, 𝒉) curves for silicone rubbers with different
thicknesses.
58 Chapter 4
Figure 4.5: Fitting curves based on applying a NeoHookean model on the experimental
data of the silicone rubbers.
Epidermis and stratum corneum
An example of the results from one subject is shown in Figure 4.6. Data that
significantly displayed measurement errors or deviated from the general response, were
ignored. In practice, usually 2 or 3 tests out of a series of 9 meausurements were left out
when calculating the average indentation curve (see Table 4.1). Figure 4.7 clearly shows
that the average curves are well-overlapping for all subjects. It appears that indentical
slopes were obtained for stratum corneum and epidermis. The Young‟s moduli derived
via the analytical approach can also be found in Table 4.1. Here, 𝐸𝑆𝐶 is about twice the
value of 𝐸𝑆𝐶+𝑉𝐸 .
The results of the FE-model are shown in Figure 4.8. For a 20 μm thick stratum corneum
and 80 μm thick viable epidermis, 𝐸𝑉𝐸 is identical for 𝐸𝑆𝐶 . Decreasing the thickness of
the stratum corneum to 10 μm hardly affects 𝐸𝑉𝐸 . Also increasing in the stiffness of the
stratum corneum did not have an effect. As expected, lowering the stiffness of the much
thicker viable epidermis causes an increasing indentation depth, from approximately 8 to
12 μm.
(a)
(b)
Figure 4.6. All indentation curves of 1 subject for stratum corneum (a) and epidermis (b).
Note that the scales are different.
A new indentation method to determine the mechanical properties of epidermis 59
(a)
(b)
Figure 4.7 Average indentation curves per subject for stratum corneum (a) and
epidermis (b).
Table 4.1: The number of tests excluded from 9 tests in total and the analytically derived
Young’s modulus of all subjects.
Stratum corneum
Subject # excluded
Eanalytical [MPa]
1
3
2.00±0.72
2
2
3.10±2.10
3
3
2.31±0.94
Epidermis
# excluded
3
3
1
Eanalytical [MPa]
0.88±0.01
1.07±0.10
1.21±0.38
Figure 4.8. Results of NeoHookean fit on the unloading curves of the epidermis. The
thickness of the stratum corneum is varied from 10 (dashed lines) to 20 μm (solid lines).
Also the effect of increasing or decreasing the stiffness of the stratum corneum is shown.
4.3 Discussion
The major problem in performing indentation experiments on skin is probably the skin‟s
surface roughness. In order to have a smooth as possible surface, we used a large
spherical indenter (ø=500 μm) such that the contact area was much greater than the
diameter of individual cells and also more homogeneous. During preliminary tests that
were performed close to the glyphics, it was observed that the poor contact definition in
60 Chapter 4
those areas resulted in an unacceptably high variability per subject. When positioning the
indenter at the highest point between a triangle formed by the glyphics, establishing the
initial contact between indenter and the tissue was not a problem. In addition, the use of
a spherical tip minimizes plastic deformations and stress concentrations and avoid
damaging the sample [110]. Using the introduced measurement protocol, highly
reproducible data could be obtained for all subjects and the variance between the
subjects was negligbly small.
In order to obtain reproducible data from an in vitro experiments that are meaningfull, a
correct sample preparation is essential. In this study, the epidermal samples were isolated
using a dermatome. Although this method does not allow for separating the epidermis at
the basal membrane only, its benefit is that the bottom side of the sample with this
obtained geometry is in full contact with the substrate. As only small deformations were
applied, the results are not influenced by the possible fragments of papillair dermis in the
sample. Our tests were performed with epidermis that was thawed and immediately used
in a dry environment. As an increasing moisture content in the epidermis decreases the
stiffness, it becomes more difficult to define the initial contact surface at higher various
humidities in the future.
The analytical method of Oliver and Pharr provides an easy method to asses the order of
magnitude of the Young‟s modulus from the experimental data. However, the theory
holds for materials responding fully elastically upon unloading. In the case of soft
tissues, this assumption is far from correct because material responses like piling-up and
sinking-in cannot be captured correctly. Due to piling up of the tissue, the projected
contact area is bigger then used in the calculations (see Figure 4.3). In our study, the
deviation is relatively small, because the use of a large spherical indenter causes a more
homogeneous contact surface.
The introduction of a numerical model should result into a better approach. The used
Neo-Hookean model, however, is also far from correct, but provides a first comparison
with the analytical method. The results show that the stiffness of the viable epidermis is
comparable to that of the stratum corneum instead of a factor two lower as calculated
with Oliver and Pharr. For both epidermal layers, the stiffness of the two layers is
approximately 1 MPa, which proves that the viable epidermis considerably contributes to
the mechanical response of skin at this lengthscale. In comparison with literature, our
values for stratum corneum are on the low side of the published range [101,107,34]. This
can be explained by the fact that the local properties studied in literature were mainly
determined by the stiffness of individual corneocytes, while our studies focused on the
tissue level. In comparison with values for full-thickness skin stiffness from in vivo
indentation tests, our values are two orders of magnitude higher [101,111,112].
The mechanical behavior of many soft tissues is described with a multimode Maxwell
model. Extending the NeoHookean model into such model would be a logical step
forward. However, the relaxationspectrum and corresponding low shear moduli that can
A new indentation method to determine the mechanical properties of epidermis 61
be derived from rheological experiments (see Chapter 3) does not influence the fitting on
the load-displacement curve. The short relaxtion times that are ranging from 0.002 up to
2 s, are only relevant during high impact loading and are in accordance with the observed
small viscoelastic plateau at the maximum applied force in the indentation experiments
(see Figure 4.6 and Figure 4.7). Moreover, also a multimode Maxwell model assumes an
isotropic material and cannot capture variations in mechanical properties with changes in
morphology, composition and moisture content through the epidermis. A better
approximation should therefore be an anisotropic model. Then, experimental data from
indentation, tensile and shear can be captured using a layered structure.
To conclude, the small deformation behavior of epidermis was studied in this study. We
have introduced a reliable experimental approach to evaluate the mechanical behavior of
epidermal tissue. The results demonstrated that the stiffness of the viable epidermis is
comparable to that of the stratum corneum for perpendicular direction at a lengthscale
relevant for clinical and cosmetic treatments. The applied load in this study covers the
physiologically relevant range. For clinical applications such as transdermal drug
delivery, the large deformations and, the ultimate goal, the failure behavior of the
epidermal layer needs to be understood. The methods presented in this study are
considered to be a suitable tool that can be extended for these purposes.
Acknowledgments
We would like to thank the plastic surgery department of the Catharina hospital in
Eindhoven for providing the skin tissue. Furthermore, we are gratefully to dr. Hagisawa
providing the protocol for the histological examination.
Chapter 5
Linear viscoelastic behavior of
subcutaneous adipose tissue
The content of this chapter is based on M. Geerligs, G.W.M. Peters, P.A.J. Ackermans,
C.W.J. Oomens, and F.P.T. Baaijens (2008), Linear viscoelastic behavior of
subcutaneous adipose tissue, Biorheology; 45(6): pp 677-688.
64 Chapter 5
5.1 Introduction
The mechanical behavior of subcutaneous adipose tissue, also called hypodermis, is a
widely ignored topic in the biomechanics literature. A plethora of papers can be found on
properties of skin and skeletal muscle, but only few papers have addressed the properties
of the layer in between [38,43,39,113,114]. This is noteworthy, because adipose tissue
plays an important role in the load transfer between different structures in the body
during breathing, body movements or exercise, or when exposed to therapeutic
stretching during physiotherapy and massage. It is well recognized that the subcutaneous
fat experiences larger strains than the dermis during suction and that its stiffness is likely
to be a few orders less than that of the dermis [1,115]. However, it is still not common
practice to take the adjacent adipose layer into account when the combined mechanical
behavior of skin, fat and muscle tissue is modeled. Currently it would be difficult to do
so, because values for mechanical parameters of adipose tissue are limited and
inconsistent in the literature. Thus, there is a need to develop a parametric and
constitutive model of subcutaneous adipose tissue, which can be implemented in
numerical models of the whole skin as well as in multilayer models including skin, fat
and muscle. Numerical models including the subcutaneous fat layer are needed in a wide
field of applications, e.g. studying skin device contact, needle insertion procedures and
the removal of skin adhesives. Rheological experiments are accepted to be a good
starting point to develop such a constitutive model.
For a meaningful interpretation of the mechanical behavior of the adipose tissue, it is
essential to know the tissue composition. The present paper is focused on subcutaneous
adipose tissue, which is a type of connective tissue throughout the body found between
the dermis and the aponeurosis and fasciae of the muscles. However, the fat pads on the
palm of the hand and foot are considered to be different, since they contain a much
higher ratio of unsaturated versus saturated fatty acids and are therefore morphologically
different. Relatively small differences in tissue composition exist at the other body sites.
Subcutaneous adipose tissue is a loose association of lipid-filled cells called white
adipocytes, of which 90-99% is triglyceride, 5-30% water and 2-3% protein. Lipids
within the white adipocytes are organized in one droplet. The diameter of the white
adipocytes ranges from 30 to 70 μm, depending on the site of deposition [17].
Collections of white adipocytes comprise fat lobules, each of which is supplied by an
arteriole and surrounded by connective tissue septae. Each adipocyte is in contact with at
least one capillary. In healthy adults, only one third of the subcutaneous adipose tissue
contains mature adipocytes [17]. The remaining two thirds consists of blood vessels,
nerves, fibroblasts, and adipocyte precursor cells.
The subcutaneous adipose tissue of the lower trunk and the gluteal-thigh region is further
divided into two distinct layers: the superficial and deep subcutaneous adipose tissue
[116,19]. Both morphological and metabolic differences were found between those two
layers [116,117,118], but it is not clear if these layers differ in terms of the mechanical
properties.
Linear viscoelastic behavior of subcutaneous adipose tissue 65
To our knowledge, only a few authors studied the mechanical properties of the
subcutaneous adipose tissue. Of those, focus has been associated with breast tissue,
particularly in the early detection of cancerous tissues [95,44,119,119,120,41,43]. These
studies have generally utilized indirect and non-invasive measurements. The largest
study involving 70 samples of breast fat tissue using ex vivo indentation experiments
yielded a mean Young‟s modulus of 3.21 kPa [19]. Linear viscoelastic behavior was
shown up to 50% strain during uniaxial tension for abdominal subcutaneous tissue of rats
when applying incremental displacement steps of 1 mm followed by a 1 second
relaxation period [113]. Patel et al. [39] measured the storage and loss moduli of
subcutaneous fat tissue, also from the abdomen, for strains up to 20%. The results
showed a frequency-dependent shear moduli decreasing, which decreased with
increasing strain. These data, however, involved measurements outside the linear
viscoelastic strain range. Recently, the mechanical behavior of subcutaneous adipose
tissue of the buttock was measured in relation to pressure ulcers by performing confined
compression tests, but no mechanical parameters for modeling could be derived from the
results [121,114].
All the above-mentioned studies only give limited descriptions of the mechanical
behavior, either because the focus was only on the differences between breast tissue
types, or on long term quasi-static behavior [113,121], or because the authors were only
interested in a comparison of properties between human fat and a mimicking material
[39].
Our ultimate goal is to develop a skin model that includes the mechanical properties of
all skin layers separately, and can be used in a numerical model. Since it may be
predicted that the mechanical behavior of adipose tissue contributes considerably to the
overall skin behavior, there is a need to develop a thoroughly tested constitutive model
describing the mechanical behavior for large strains. The formulation of such a model
will be based on rheological experiments in vitro. The first step is to investigate the
material bulk properties within the linear viscoelastic strain region, which is defined as
the range of strain amplitudes where the material properties are independent of the
applied strain. The types of experiments are relatively simple to perform and hence, it is
appropriate to design experimental procedures as well as to identify experimental
problems. The linear viscoelastic parameters obtained will form the basis for a non-linear
viscoelastic model in future work. The concept will be developed for porcine
subcutaneous adipose tissue because of the availability and minimal biological
variability among specimens. The objective of the current study is to use dynamic
mechanical thermal analysis (DMTA) in combination with Time Temperature
Superposition (TTS) to determine the small oscillatory strain behavior of subcutaneous
adipose tissue in vitro. DMTA is performed through oscillatory shear experiments up to
100 rad/s at various temperatures. Next, the linear viscoelastic power-law memory
function, commonly used for soft-solids, will be introduced to describe the small strain
viscoelastic behavior of this tissue.
66 Chapter 5
5.2 Methods and Materials
5.2.1 Sample preparation
Porcine subcutaneous fat tissues were obtained from a local slaughterhouse (Ballering,
Son, The Netherlands), where they were cut into transverse slices of 1.5-2.0 mm thick
and stored at 4°C. In porcine species, the back fat is divided in an outer, middle and
inner layer of subcutaneous tissue because the adipocyte features of these layers differ
with respect to size, number and metabolic activity. The porcine middle layer, which is
used in the present study, is comparable to the deep subcutaneous layer in the abdominal
region of humans [122]. All pigs were Landrace, having a dressed carcass weight of
approximately 83 kilograms, and were 14-18 weeks old at necropsy.
Within 48 hr of collection, circular tissue samples were obtained from the slices with an
8 mm diameter cork borer. Next the samples were stored ice-cooled in a PBS solution
and tested within the subsequent 4 hours. An overview of the number of specimens and
the number of samples from each specimen per test is given in Table 5.1.
Methods of tissue preservation may change the mechanical properties of tissue due to
changes in tissue quality [85]. Rapid freezing, which has not been demonstrated to
change the fatty acid composition compared to fresh tissues [123], is an attractive
solution for storing tissue for prolonged periods. Thus, in order to assess whether snap
freezing preserves mechanical properties, adipose tissue was snap-frozen by immersion
in 2-methylbutane cooled by liquid nitrogen and stored at –80°C until use for mechanical
testing. Thawing of the samples was done slowly within an ice-cooled box. In order to
assess these storage conditions, histological sections were examined by light
microscopy. For that, the specimens were fixed in 10% phosphate-buffered formalin and
processed for conventional paraffin embedding. The specimens were cut into 5-μm thick
sections and stained with hematoxylin and eosin (H&E). Since all lipids were extracted
out of the adipocytes by using the conventional paraffin embedding technique, other
specimens were embedded in O.C.T. compound (TISSUE-TEC) and frozen for lipid
staining. These specimens were cut into 8-μm thick sections at –20°C, stained with oil
Red O (Sigma) and counterstained with hematoxylin.
5.2.2 Rheological methods
To determine the linear viscoelastic properties, oscillatory shear experiments were
performed using a rotational rheometer (Advanced Rheometric Expansion System
(ARES), Rheometrics Scientifc, USA) with a controlled strain mode, and parallel plate
geometry in combination with a Peltier environmental control unit. Sand-blasted plates
were used to prevent slippage. An oscilloscope was used to ascertain that the shape of
the torque signal was indeed sinusoidal. Samples were compressed between the plates by
lowering the upper plate until an axial force of 0.1 g was reached.
Linear viscoelastic behavior of subcutaneous adipose tissue 67
In the experiments a sinusoidal strain γ(t)
insteady
state and within the range of linear viscoelastic behavior, resulted in a sinusoidal shear
rate, g(t), and shear stress,τ(t) with a phase shift δ:
 (t )   0 sin(t ) ,
(5.1)
 (t )  Gd  0 sin(t   ) .
(5.2)
The dynamic shear modulus Gd(ω,T) and the phase shift δ(ω,T) are both a function of the
angular frequency ω and temperature T. It is common to separate the dynamic shear
modulus into a storage modulus, G', representing the elastic behavior since this describes
the stress in phase with the strain, and a loss modulus, G'', representing the viscous
behavior, 12  out of phase with the strain, i.e. in phase with the strain rate:
Gd  G' 2 G"2 .
(5.3)
The phase shift d
to (5.3:
tan( ) 
G"
G' .
(5.4)
The Time-Temperature Superposition (TTS) principle is applicable when data can be
shifted to and from a reference temperature T0 to form a master curve [124]. The
advantage of this principle is that the frequency domain can be extended beyond the
measurement limits as well as that data can be shifted to other working temperatures. A
smooth master curve is obtained by shifting frequency sweep curves obtained at different
temperatures horizontally and vertically on the curve obtained at the reference
temperature, until all the curves overlap. Normally the horizontal shift factor aT is
applied to the phase angle δ. Subsequently, the dynamic shear modulus Gd, and also G'
and G'', can be shifted along the horizontal and vertical axis to a reference temperature
with the horizontal shift factor aT and a vertical shift factor bT:
tan  (, T )  tan  (aT , T0 ,
Gd ( , T ) 
1
Gd (aT  , T0 ).
bT
(5.5)
(5.6)
68 Chapter 5
5.2.3 Testing procedure
Test protocols were based on measuring the linear viscoelastic properties of other soft
biological tissues, such as brain [125], muscle [126] and thrombus [93]. The linear
viscoelastic regime was determined using oscillatory shear experiments with constant
frequency and varying strain. Strain sweeps were performed from 0.04% to 10% at
frequencies of 1, 10 and 100 rad/s and 20°C. A constant strain within the determined
linear regime of 0.1% was chosen for the subsequent frequency sweep tests.
The frequency sweep was repeated three times to avoid tissue conditioning phenomena,
observed during preliminary testing. We did not carry out traditional preconditioning.
Instead we performed always three frequency sweeps, increasing the frequency stepwise
logarithmically from 1 to 100 rad/s and then performing the data analysis on the third
frequency sweep. This protocol was also used to examine the influence of snap freezing
and thawing on the mechanical properties of subcutaneous fat tissue. For this purpose,
samples from 3 pigs were tested, both fresh and after freezing and thawing. All tests
were performed at 20°C.
To investigate whether the TTS principle is applicable to subcutaneous adipose tissue,
frequency/temperature sweeps were successively performed at temperatures of 5, 20, 35
and 40˚C, at 0.1% strain and frequencies ranging from 1-100 rad/s. Again, two
successive frequency sweeps from 1-100 rad/s were performed prior to these
frequency/temperature sweep tests. The temperature range is bounded at the low end by
the phase transition temperature of water and above by temperatures at which protein
degradation is likely to occur. To check the possible influence of the order of heating or
cooling, 3 samples were also subjected to a frequency/temperature sweep with
decreasing temperatures.
As a control for the applied power-law model, a stress relaxation experiment additional
to the frequency sweep tests was performed. In these experiments a step strain of 0.1%
was applied during 100 s.
Table 5.1: Overview of number of specimens and number of samples per specimen used
for the experiments.
Test
Strain Sweep
Frequency Sweep
5x repeated
Model fit
Effect snapfreezing
Frequency/Temperature Sweep
Increasing T
Decreasing T
Stress relaxation
* sample number per condition
# Specimens (# samples per specimen)
3 (4,5,3)
3 (3,3,3)
3 (3,3,3)
3 (3,5,3)*
2 (3,3)
2 (2,1)
1 (5)
Linear viscoelastic behavior of subcutaneous adipose tissue 69
5.2.4 Statistics
For the strain sweep, frequency sweep and stress relaxation tests with fresh tissue, the
average values and standard deviations were calculated for the mechanical parameters at
different testing strains or frequencies. In order to determine whether snap-freezing has a
significant effect on the mechanical parameters, data were analyzed with the linear
mixed model [127] by using the software Splus. For this purpose, the log of the
frequency sweep data was used. The linear mixed model was chosen because it accounts
for biological variability among samples and among specimens while analyzing freezing
effects.
5.3 Results
5.3.1 Small oscillatory strain behavior
Fig. 1 shows the results for the strain sweep tests at 10 rad/s for both G' and G''. Both
moduli and phase shift, which is not shown here, were found to be nearly independent of
strain for amplitudes up to 0.1%. Tests at other frequencies revealed similar results and
are therefore also not shown.
Preliminary testing showed that tissue conditioning phenomena are minimised by
performing two frequency sweeps before the actual measurement (Figure 5.1). Results
for the storage and loss moduli and the phase angle, as functions of the applied
frequency, are shown in Figure 5.2. The biological variation appeared to be small.
Taking all samples from fresh specimens together, the shear modulus Gd is found to be
14.9 kPa ± 4.8 kPa at 10 rad/s. The average phase angle is approximately 21.0° over all
frequencies, indicating that the complex modulus is dominated by elastic behavior.
Results of stress relaxation are depicted in Figure 5.3. The shear modulus decreases over
a decade over 100 s.
Figure 5.1: Results from strain sweep tests. Average G’ and G” demonstrate a linear
viscoelastic regime up to 0.1% strain at a frequency of 10 rad/s.
70 Chapter 5
Figure 5.2. Frequency sweep results: (a) mean G’, and G”, the standard deviations and the
fitted model; (b) mean δ , standard deviation and the estimated fit.
Figure 5.3. Stress relaxation behavior.
5.3.2 Model application
The shear stress response for linear viscoelastic behavior is usually described in terms of
the Boltzmann integral:
t
   G (t  t ' ) (t ' )dt '

,
(5.7)
where G(t) is the relaxation function and  is the shear rate. The results of the frequency
sweeps indicate that a power-law relation can adequately describe the storage and loss
moduli:
Linear viscoelastic behavior of subcutaneous adipose tissue 71
G' ()  G' (1) p ,
(5.8)
with G'(1) and p as constants [128]. The same relation is used for G''.
The phase angle can be expressed in terms of the exponent p [128,129]:
tan  
G"
p
 tan
G'
2 .
(5.9)
So the small oscillatory strain behavior is captured by an approximation with only two
constants (p,G(1)). It is known [128] that the relaxation function G(t) in Eq. (5.7 can be
written as
G(t )  G(1)t  p .
(5.10)
The constants G(1) is related to G'(1) by
G (1) 
2G ' (1)( p!)
p
sin
,
p
2
(5.11)
where p! is the factorial function.
The expressions for G' and G'' were fitted simultaneously, resulting in one value for p
per sample. Next, the exponent p was used to calculate the phase angle corresponding to
the frequency sweeps (Figure 5.2) and the relaxation modulus for the stress relaxation
experiments (Figure 5.3). In all cases, the exponent p was in the range from 0.18-0.25,
with a mean value of 0.21.
5.3.3 Time-Temperature Superposition
Results of the frequency/temperature sweeps show that the phase angle is not dependent
on temperature for increasing temperature (data not shown). However, the shear modulus
Gd can be shifted along the horizontal frequency axis to obtain a smooth master curve at
a reference temperature of 20°C (Figure 5.4), in such a way that Gd (, T )  Gd (aT , T0 ) .
Results of the frequency/temperature sweeps with decreasing temperature were similar to
those with increasing temperature and are therefore not shown here. The curves of Gd for
different temperatures show curves that overlap extensively such that the frequency
domain could be extended to almost 3 decades (Figure 5.4). The horizontal shift factors,
as a function of the temperature at which each dataset was acquired, can be captured
reasonably well with an exponential function with a quadratic power:
aT  e
aT0 2 bT0 c
,
(5.12)
72 Chapter 5
with a = -0.0046 ± 0.0021, b = 2.54 ± 1.25 and c = -351.39 ± 183.39 (Figure 5.5). From
this, it can be calculated that Gd at body temperature is approximately 7.5 kPa at 10 rad/s.
Figure 5.4. (a) Example of frequency sweeps performed at different temperatures, which
can be shifted horizontally; (b) master curve of Gd obtained for two specimens each within
3 samples.
Figure 5.5. Shift factor aT versus temperature T. Experimental data from three sepcimesn
(,○,□) from two specimens are shown together with the mean fit.
5.3.4 Freezing effects
From the histological sections, severe damage could be observed in 2 out of 12 samples.
Either cells were less packed or cell membranes were ruptured (Figure 5.6). However,
less or no damage occurred when tissue was embedded in the O.C.T. compound. So it
remains unclear, whether the damage was only due to the snap freezing method and/or
preparation artefacts.
The frequency sweeps showed that the differences of the intercepts of regression lines
were not statistically different, whereas the differences in the slopes of the lines for G’
were statistically different (Figure 5.7). However, the biological variance among all
samples is larger than the difference between the fresh and snap frozen samples. This can
Linear viscoelastic behavior of subcutaneous adipose tissue 73
be seen in Figure 5.7, where the regression line of the frozen samples lies within the
biological variation of the fresh samples. So from a practical viewpoint, the observed
difference of slopes for the two conditions is negligible for G'. In the case of the G''
slopes, there was no statistical difference. Taken this all together means that snap
freezing does not show any effects on the mechanical properties compared to fresh
tissue.
Figure 5.6. (a) Fresh adipose tissue, (b) adipose tissue after snap freezing without damage,
(c) tissue damage after snap freezing.
Figure 5.7. The biological variation on the slope of the normalized regression lines of G’ is
shown. The dotted lines represent the limits of two times the standard deviations on both
sides of the belonging regression line.
5.4 Discussion
The results indicate that the shear moduli can be shifted to measurement conditions
described in the literature when using the Time-Temperature Superposition. From the
literature it is known that the linear region for other soft-solids consisting of loosely
bounded soft particles is below 1%, which is consistent with the present observations. In
fact, the linear region is considered to be only up to 0.2% strain. This small strain was
the maximum strain that could still represent linear behavior within an acceptable signalto-noise ratio. Too large strain amplitudes are outside the linear strain regime and reduce
the “apparent” modulus, which might explain the difference with Patel‟s data [39]. In
74 Chapter 5
comparison with Samani et al. [43], who applied a quasi-static loading with a frequency
of 0.1 rad/s resulting in a Young‟s modulus of 3.2 kPa., our shear modulus Gd(ω = 0.1
rad/s, T = 20°) is 5.6 kPa, which results into a higher Young‟s modulus. In addition, the
present results show an obvious temperature dependency and a specific start-up
behavior. The reasons for these differences are unknown. The reproducible long term
variations in the beginning of a frequency sweep, a change in the slope of G', are not yet
understood. Snap freezing may cause tissue damage resulting in less packed cells or
ruptured membranes, but it is more likely that the observed artifacts are caused by the
chosen histological technique. Snap freezing did not appear to have an effect on the
mechanical behavior. Although the slopes of the regression lines for G' demonstrated
significant differences, the observed difference is smaller than the biological variation
between samples. Many of the environmental conditions, other than temperature, are
difficult to control. Since the snap frozen samples were measured on separate days to the
fresh samples, the environmental conditions might have influenced the measurement
outcomes per specimen.
In the present study porcine tissue from the slaughterhouse was used. The nature of the
source of biological material at the present study was such that biological variation
between specimens was relatively small. The adipocytes of the pigs had a diameter of 70
μm or greater whereas that of human adipocytes varies from 30 to 70 μm. The question
arises whether other tissue composites contribute more to the mechanical behavior of the
bulk tissue than the adipocytes. Besides blood vessels and the collagen fiber network, no
other significant composites are present in the adipose tissue. Tissue with visible blood
vessels was excluded from testing. Therefore, it is conceivable that the stiff collagen
fiber network surrounding the fat lobules plays an important role in the overall
mechanical behavior.
To our knowledge, it is the first time that this common rheological model has been
applied to biological soft tissue. The power-law model fits the experimental data well.
The p-values obtained are comparable to those of other soft materials in the literature. It
should be noted, however, that the fit on the slope of the stress relaxation behavior could
be improved although an optimization process would not yield any further benefit. More
interesting is the fact that we have introduced a model that can be extended to a threedimensional non-linear model capturing large deformations with the possibility to
include the build up and breakdown behavior of initial structures. Nevertheless,
experiments in the non-linear strain regime are necessary to prove whether or not this
promising model can fit those predictions.
Also, Time-Temperature Superposition is applicable to this type of biological tissue.
Mechanical properties measured at any temperature can be shifted to body temperature
by applying the Time-Temperature Superposition. However, the applicable temperature
range for experiments is physically bound by phase transitions at low temperatures and
the solidifying of proteins above 41°C. The measurements already showed a much
larger variation at the upper limit of the temperature range, i.e. at 40°C, than at any other
Linear viscoelastic behavior of subcutaneous adipose tissue 75
temperature. This indicates that it is recommended to avoid this boundary of the
temperature range.
Chapter 6
Does subcutaneous adipose tissue behave
as an (anti-)thyxotropic material?
The contents of this chapter are based on M. Geerligs, G.W.M. Peters, P.A.J.
Ackermans, C.W.J. Oomens, and F.P.T. Baaijens (2008), Does subcutaneous adipose
tissue behave as a thyxotropic material?, Journal of Biomechanics, accepted.
78 Chapter 6
6.1 Introduction
The mechanical load transfer from a skin contact area to deeper tissues involves several
tissue layers. On most body sites, the subcutaneous adipose tissue considerably
contributes to this load transfer. However, when numerical models are used to predict the
stress response due to external loading, the focus is either on the skin-device contact or
on the deeper tissue layers while the subcutaneous fat layer is often ignored. This
omission might be related to the lack of defined parameters, which describe the
mechanical behavior of adipose tissues. This is particularly surprising given the critical
roles for adipose tissues in the medical and cosmetic fields, involving, for example,
implantable drugs delivery, skin adhesive removal, deep tissue injury and needle
insertion procedures.
Recently, our previous work on the linear behavior of subcutaneous adipose tissue has
shown that the linear strain regime is valid for very small strains only, i.e. 0.001 [130]. In
most applications, however, much higher deformations occur in the adipose tissue for
prolonged periods. Indeed, for wheelchair or bedridden patients, for example, this might
lead to the development of deep tissue injury under bony prominences within a time
frame of minutes to hours, during which stress relaxation in the compressed tissue might
occur [40]. Numerical models based on experimental data are of indispensable value to
predict the onset and progression of such mechanical-induced damage.
Currently, there is a paucity of papers on the mechanical properties of the subcutaneous
adipose tissue found beneath hairy skin. Viscoelastic properties of single human
adipocytes have been recently characterized using AFM resulting in a relaxed modulus
and relaxation time for either load or deformation [131]. Few related in vitro studies on
tissue behavior exist. Of these, rheological measurements demonstrated a decrease in
viscosity with increasing shear rate [39]. In addition, the authors suggested that adipose
tissue loses firmness with increasing strain and frequency, a state which is not
recoverable. In a separate study, ovine subcutaneous tissue was subjected to ramp-andhold cycles during confined compression tests at various ramp rates [39,40]. The results
were given in the form of a transient aggregate modulus and short-term elastic moduli.
They also found a strong deformation rate dependency. Short-term moduli were in the
order of 20 kPa. In an alternative in vivo approach, a suction device yielded experimental
parameters which, when combined with numerical modeling, led to a first estimation of
non-linear material parameters for human skin [115]. To our knowledge, there are no in
vivo studies considering subcutaneous adipose tissue as a single layer. By contrast, some
in vivo studies have examined the mechanical properties for a compliant system
consisting of skin and subcutaneous adipose tissue [121,132].
The work mentioned above describes a range of loading conditions, often combining
techniques involving indentation, confined compression, stress relaxation and constant
shear responses. Clearly, this makes comparison of data from the studies problematic in
Does subcutaneous adipose tissue behave as an (anti-)thyxotropic material? 79
nature. For a general constitutive model for adipose tissue a more systematic approach is
required.
The material structure of subcutaneous adipose tissue does not relate conveniently to
other biological tissues. Its main component is the white adipocyte. The remaining
components are water (5-30% weight) and protein (2-3% weight). The white adipocytes
are filled with a large fat droplet imposing forces on both the nucleus and the small
cytoplasmic volume at the cell periphery. The composition of the white adipocytes
depends on the specific function and body site. As an example, differences throughout
the human body are known for the proportions of saturated fatty acids, monosaturated
versus polysaturated fat and the lipolysis rate [17]. White adipocytes are collected in a
surrounding fiber network. The adipose tissue is well-vascularized throughout with each
adipocyte in contact with at least one capillary. Hence, adipose tissue is susceptible to
ischemia and hypoxia, which influence its mechanical response.
Our previous work on the small strain behavior of adipose tissue has shown that
reproducible results are obtained in an in-vitro set-up using a rheometer with parallel
plate geometry and that the behavior can be described with a power-law model [130].
However, sometimes tissue samples were found to be much stiffer than the mean value
and early work at higher strains has suggested that (reversible) structural changes start to
play a role. In addition, earlier large strain studies formed the incentive for a more
systematic approach at higher strains to elucidate the phenomena that havealready been
described. Therefore, the present study aims to provide systematic data for long-term
small strain behavior as well as the effect of strain history, with the purpose of
contributing to the development of a constitutive model.
Accordingly, the work is divided in two parts. The first part contains long term
oscillatory tests at small strains to investigate temporal effects of the adipose tissue
samples. Subsequently, strain-dependency tests, comprising constant shear, stress
relaxation and constant strain rate, are applied. From these tests, non-linear parameters
can be obtained useful for constitutive modeling. Such an experimental approach is
designed to gain insight on the mechanical response of adipose tissue under shear where
the effect of strain history, strain level and duration is taken into account.
6.2 Materials & Methods
6.2.1 Sample preparation
In porcine species, the subcutaneous fat layer on the back is divided in an outer, middle
and inner layer. The porcine middle layer was selected for use, as it is considered to be
the most comparable with the deep subcutaneous layer in the abdominal region of
humans [122]. The tissue was obtained from a local slaughterhouse, where they were cut
into transverse slices of approximately 1.5 mm thick. In our laboratories, circular
samples were obtained from the slices with an 8 mm diameter cork borer. The samples
80 Chapter 6
were stored in a Phosphate Buffered Saline solution (PBS) in ice-cooled boxes and tested
within 48 hr of collection. If measurements were repeated after a certain period of
recovery, each sample was stored in PBS between measurements. All pigs were
Landrace, having a dressed carcass weight of approximately 83 kilograms, and were 1418 weeks old at necropsy.
6.2.2 Rheological methods
All experiments were performed on a rotational rheometer (ARES, Rheometric
Scientific, USA) with parallel plate geometry in combination with a Peltier
Environmental control unit and a fluid bath. Plates were sand-blasted to prevent
slippage. The upper plate was lowered to compress the sample until the sample
experienced an axial force of 1 g. All loading protocols, which were based on previous
experiments on soft biological tissues (Van Dam, 2008; Hrapko, 2006), are summarized
in Figure 6.1.
Long-term dynamic behavior within the linear viscoelastic regime was studied with time
sweep tests (Figure 6.1a). Tests were performed at a frequency of 10 rad/s with a strain
amplitude of 0.001 at body temperature (37°C), lasting at least 45 minutes. The chosen
strain amplitude was previously determined to be the maximum strain within the linear
viscoelastic regime [130]. Time sweeps were repeated after various time periods of
recovery, namely 0, 0.5, 1 and 3 hours.
Shear experiments in the non-linear regime were preceded by two successive frequency
sweeps with a frequency of 1-100 rad/s and a strain amplitude of 0.001. This procedure
was adopted to minimize the effects of pre-conditioning [130]. Subsequently, the sample
was tested in either a series of constant shear rate experiments, constant shear
experiments or stress relaxation experiments (Figure6.1b-e). The measurement protocols
were based on previous experiments on soft biological tissues.
Constant shear rate experiments with various strain amplitude were designed to
investigate any potential damaging effect in the mechanical behavior due to the previous
strain history on the immediate mechanical response. The first series of sequences were
loading-unloading tests conducted with a constant shear rate of 1 s−1 and strains
incrementally increasing from 0.01 up to 0.5 (Figure 6.1b). The sample was left to
recover at zero strain for at least 10 times the loading time after each loading-unloading
cycle. In total, 20 cycles were applied. In another series of sequences with the same
constant shear rate, strains were applied in decreasing order (Figure 6.1c). Again the
sample was left to recover at zero strain for at least 10 times the loading time after each
loading-unloading cycle. In order to investigate possible reversible changes, this
sequence was repeated after 0, 1 and 3 hours of rest.
The next set of experiments was designed to apply constant shear at increasing shear rate
(Figure 6.1d). Loading-unloading cycles were conducted with constant shear rate
increasing from 0.01 s−1 to 1 s−1 per cycle with maximum strain amplitude of 0.15.
Does subcutaneous adipose tissue behave as an (anti-)thyxotropic material? 81
Between two cycles, the sample was again left to recover for at least 10 times the loading
time.
Finally, stress relaxation experiments were composed of a series of ramp-and-hold tests
at different strain levels (Figure 6.1e). During the loading and unloading phase, a
constant strain rate of 1 s-1 was imposed. The maximum strain was held for 10 s during
which the relaxation of the material was recorded. The sample was left to recover for a
period of at least 100 s during which time the tissue response was monitored. The test
was repeated for four different strain levels, namely 0.01, 0.05, 0.1 and 0.15.
An overview of the number of specimens and the number of samples from each
specimen per test is given in Error! Reference source not found..
Figure 6.1: Schematic illustration of test sequences. (a) Time sweep tests; (b) Constant
shear rate experiments with increasing shear strains; (c) constant shear rate experiments
with decreasing shear strains; (d) constant shear experiments with increasing shear rate;
(e) stress relaxation experiments.
82 Chapter 6
Table 6.1: Overview of number of samples used for the experiments.
Test
Time sweep
Constant shear rate
increasing shear
decreasing shear
Constant shear
Stress relaxation
# specimens (# samples per specimen)
4 (1,6,4,2)
1(3)
1(2)
2(3,3)
2(3,3)
6.3 Results
6.3.1 Long term small strain behavior
An interesting qualitative trend was observed during the time sweep experiments (Figure
6.2a). The samples showed a gradual increase of both initial storage modulus and initial
loss moduli over time from the start of the experiment. However, after a period, a rapid
increase in stiffness, G‟, occurred in all samples indicating a change in tissue structure.
The moduli showed a further slight increase until a steady state was reached. During the
steep increases the moduli increased by a range of roughly 1.5-15 kPa. The rapid
stiffening occurred at some time between 250 s and
1200 s. An overview of the stiffness increase and start time for all 13 samples is given in
Figure 6.2b.
Figure 6.2: (a) Typical result of a time sweep: the arrow indicates the measured increase in
the storage modulus G’ during quick stiffening phase (ΔG’). (b) ΔG’ against the start time
of the stiffening for samples from all specimens.
Experiments with repeated time sweeps show that the material behavior is reversible,
although recovery takes several hours to complete (Figure 6.3). To enable comparison
between specimens, the shear moduli of each specimen were normalized to a scale r
from 0 to 1, e.g. from the initial modulus up to the final steady state level of the initial
test. When the second time sweep is immediately performed after the first time sweep,
Does subcutaneous adipose tissue behave as an (anti-)thyxotropic material? 83
the initial moduli remains constant at the plateau value, see Fig. 3a. After a recovery
period of 1 hour, the initial value for the moduli is reduced, although not reaching the
level corresponding to that during the first time sweep. After 3 hours the material
appeared to be totally recovered and a qualitatively comparable curve could be obtained.
A third test on the same sample after a further 3 hours of recovery (trest =6 hr in Figure
6.3b) demonstrated a qualitatively similar curve.
Figure 6.3: Repetition of time sweeps. (a) The shear moduli are scaled from 0 to 1, from
the start value of the initial test on the specific sample up to the stationary state at the
higher plateau. The initial response from one sample is shown here by the thick line; the
other lines represent the response after various periods of rest time for the same sample;
(b) A sample is loaded again after 3 and 6 hours of rest to demonstrate the reversible
behavior.
6.3.2 Large strain experiments
In the constant shear rate experiment with increasing strains (Figure 6.4a), three phases
can be distinguished as delineated by strain values of 0.15 and 0.30 in Fig. 4b. If the
stress strain curve (Figure 6.4b) is enlarged to highlight the first phase, it is evident that
the responses at strains up to 0.15, within reasonable limits, overlap (Figure 6.4c) and
can be considered to be reproducible. For strains above 0.15, however, the loading
curves are changing. For increasing strain, the stress is decreasing for subsequent loading
cycles indicating strain induced changes in the tissue. By contrast, above 0.3 strain, the
curves appear to overlap for repeated load cycles suggesting that tissue structure is not
changing further. Although the stress response greatly differs for the three phases for the
large strain range, the stress response within the linear strain region did not change.
The results of the constant shear rate experiments with decreasing strain are depicted in
Figure 6.5. Notice that the tissue structure immediately changed in the first cycle, and
that the subsequent loading cycles followed the first curve. In addition, despite applying
strains of approximately 0.3, the specimens were able to recover after a sufficient
recovery period.
84 Chapter 6
Figure 6.4: Average results from constant shear rate experiment with increasing strain
amplitude. (a) Applied shear strain with reproducible strain rate; (b) the three different
phases of the stress-strain response; (c) stress-strain response up to 0.1% strain.
Figure 6.5: Average results from constant shear rate experiment with decreasing strain
amplitude. (a) Applied shear strain with reproducible strain rate; (b) Stress-strain curves
from constant shear rate experiments with decreasing strain. The applied sequences have
been repeated after various rest periods (dotted lines).
Constant shear rate experiments with increasing strain rate were applied up to a
maximum strain of 0.15. From the results it can be observed that the stress as a function
of strain is strain rate dependent and that the response stiffens with increasing strain rate
for both the linear and non-linear range (Figure 6.6).
Results of the stress relaxation experiments are illustrated in Figure 6.7. The results show
practically overlapping curves for the loading phase in the linear strain regime (Figure
6.7). The stress response in the non-linear strain region followed a nearly identical curve
for each sample (Figure 6.7c). During stress relaxation, the relaxation modulus did not
reach yet a plateau value within the relaxation time allowed (Figure 6.7d). The averaged
relaxation modulus decreases as a function of applied strain, where the difference
becomes smaller for larger strains.
Does subcutaneous adipose tissue behave as an (anti-)thyxotropic material? 85
Figure 6.6: Constant shear experiments with increasing strain rate.
Figure 6.7: Results of stress relaxation experiments in shear (test sequence C). (a) stress vs.
time for one sample; (b) stress-strain response for one sample; (c) peak stress variations
(n=6); (d) average relaxation modulus vs. time.
6.4 Discussion
For this study, both long term behavior at small strains and strain history effects at large
strains were investigated. Samples from porcine subcutaneous adipose tissue
86 Chapter 6
demonstrated noteworthy behavior for both types of loading. The long term behavior
obtained at small strains is qualitatively reproducible. However, in quantitative terms,
both the time of onset and the amount of increase in moduli values varied considerably
(Figure 6.2b). The cause for those variations is not yet understood. Nevertheless, the
observed sudden stiffening of the material up to a decade is crucial for understanding
and measuring the material behavior of adipose tissue. The rapid increase in tissue
stiffness implies structural changes, which are reversible, and might influence
mechanical testing over longer time periods.
Responses in the large strain regime were examined initially by performing constant
shear rate experiments (Figure 6.1b). The stress-strain response changed for increasing
strains and can be divided in three phases (Figure 6.4b). Material behavior changed
dramatically. Additional experiments were therefore performed to ratify the tissue
structure changes due to mechanical loading, as well as to investigate tissue recovery.
These experiments with decreasing shear confirmed that the stress-strain response is
dependent on the strain history. The applied large strains here are in accordance with
physiologically relevant strains, for example equivalent to that estimated during sitting
[121].
From the constant shear rate experiments it can be concluded that up to 0.15 strain, the
adipose tissue might behave mechanically similar to other biological tissues such as
brain tissue and thrombus [125,133,92]. Because tissue structure changes might occur
above 0.15 strain, the subsequent large strain experiments were performed up to this
limit. The constant shear experiments and stress relaxation tests indicate both reliability
and reproducibility of the test method and show similar trends as those reported for
samples from brain and thrombus tissues. These findings therefore support the
appropriateness of a Mooney-Rivlin like model for the simulation of the first phase of
large strains.
Structural changes due to mechanical loading are an indication of thixotropic behavior.
Thixotropic behavior is defined as a time-dependent decrease of viscosity or modulus
induced by deformation which is a reversible effect when the deformation is removed
[134]. When the deformation causes a reversible, time-dependent increase, it is called
antithixotropy. (Anti-)thixotropic materials may or may not be viscoelastic in nature.
Both the long term behavior at small strains and the constant shear rate experiments
indicate reversible structural changes. However, the small strain results indicate an antithixotropic behavior, while the large strain results show a thixotropic behavior that is
observed at the large strain only. The stress relaxation response evidently indicates
viscoelastic behavior. In the human body, blood and synovial fluid are known to behave
thixotropically [135,136,134]. For adipose tissues, it would be interesting to visualize
using a confocal microscope to see whether adipocytes and/or the surrounding collagen
network behavior rearrange with mechanical loading. In addition, to examine the
mechanical behavior for strains above 0.15 specific test methods are needed, as
summarized in a recent overview [134]. When establishing such experiments, the large
Does subcutaneous adipose tissue behave as an (anti-)thyxotropic material? 87
strain behavior of adipose tissues should be studied preferably before stiffening occurs at
small strains to be independent of time effects.
The outcome of our large strain studies were not influenced by time effects. From the
large deformation studies, the experiment with an increasing strain up to 50%
represented the most prolonged lasting approximately 2300 s, including the preceding
frequency sweeps. The loading-unloading cycle was maintained at a maximum strain for
only 1 s, which amounted to only 20 s in total. The duration of the other experiment with
increasing strains was less than 500 s. The increasing shear rate experiments and stress
relaxation experiments lasted approximately 900 and 750 s with short term loadingunloading cycle as well. So the long-term time effects did not influence the outcome of
the strain-dependency studies.
The observed reversible behavior is in contradiction with a previous study [39]. These
authors argue that even at small deformations human adipose tissue is not able to recover
during creep tests. Since the linear strain regime is only applicable to very small strains,
it might be that those measurements are performed outside this region or that the
recovery time was insufficient.
The described phenomena may have major consequences for the interpretation of results
of biomechanical studies. A field of interest of the authors is the development of pressure
ulcers, tissue degeneration after prolonged loading, usually occurring in bedridden or
wheelchair bound patients. Recent studies have shown that these ulcers can start at the
skin, but also in deeper tissue layers close to bony prominences [137,138]. This pressure
induced “deep tissue injury” is a major issue for wheelchair bound paraplegic patients
because they are insensate to pressure-induced effects and injury is very difficult to
diagnose in the absence of visible damage at the skin surface. In the studies on etiology
and development of methods for prevention, biomechanical modeling is a valuable tool.
The fat layer plays a very important role in these analyses and the stiffness changes
described in the current paper will have a major impact on the stress and strain
distributions within the different tissue layers overlying the bony prominences. This
highlights the need for further research on this subject and to derive a theoretical model
for the description of fat behavior.
In conclusion, the time sweeps tests and the large strain experiments demonstrate that
time effects and strain effects result in different material behavior. This indicates (anti-)
thixotropic material behavior meaning that a constitutive model should contain
parameters to describe the build-up and breakdown of material structure. When only
large strains up to 0.15 are considered, a Mooney-Rivlin model should be able to capture
the experimental data. The application of the Mooney-Rivlin model would demand extra
parameters to include the effect of prolonged mechanical loading as well as the
physiologically relevant high strains. Additionally, a power law model describing the
linear viscoelastic behavior has been introduced in our previous work. This model would
88 Chapter 6
also be suitable for implementing a build-up and breakdown structure properties. We
believe, however, it is better to set-up more experiments to fully understand the material
behavior before continuing the building of a constitutive model.
This paper shows the high complexity of the material behavior and particularly
demonstrates more work is needed on this topic. The described effects should be taken
into account when setting up new experiments. The follow-up experiments should clarify
the effects of time and strain and the reversibility of the material.
Acknowledgements
We would like to thank Prof. Dan Bader for his valuable contribution to our discussions
during preparation of this article.
Chapter 7
General discussion
90 Chapter 7
7.1 Introductory remarks
The mechanical behavior of skin is of utmost importance for clinical and cosmetic
treatments. However, there is a paucity of information regarding the role of tissue
mechanics in disease progression, skin-device interaction, tissue repair, and remodeling
mechanisms associated with those treatments. As the skin is a challenging material
composed of a layered hierarchical structure, a wide range of measurement methods for
mechanical characterization of skin have been developed. Most researchers tended
toward in vivo testing for obvious reasons. Non-invasive studies can then be applied on
skin in its natural environment at different body sites and it is reasonably cost-effective.
Although in vivo testing requires ingenious procedures and a lot of assumptions to
simplify the models describing the experiment or else, numerical-experimental
procedures including inverse parameters estimations, the in vivo methods are quite
succesful for mechanical chararcterization of the dermis. The overall mechanical
behavior of skin is often considered to be equivalent to the dermal properties.
Most clinical and cosmetical applications require more detailed knowledge about
individual layers at the skin surface, viable epidermis and stratum corneum, and about
the deeper hypodermis. It is required to accurately measure displacements in all the
layers with non-invasive methods like ultrasound, MRI, Raman spectroscopy, and
Optical Coherence Tomograph. The different lengthscales, ranging from 10 μm of the
stratum corneum to the cm scale for the hypodermis, and the inverse relation between
penetration depth and resolution of all above mentioned techniques form a major
problem. The length scales as well as the variety of stiffnesses found in the different
layers also form a major difficulty for the numerical simulation tools as well as for the
parameter algorithms [21].
That is why instead of investigating the mechanical behavior of human skin layers in
vivo, this thesis aimed to prove that individual human skin layers can be mechanically
characterized in a reliable and reproducible manner using an in vitro set-up. The layers
of interest were the stratum corneum, epidermis and hypodermis, because their
mechanical behavior is unknown or results in literature are inconsistent. As it is
important to measure samples of consistent quality, isolation and preservation techniques
for the various skin layers were analyzed. Subsequently, testing apparatus were adapted
to be applicable. Because only epidermis is already a layered structure by itself, the small
strain behavior was determined in an in-plane and perpendicular direction under various
environmental conditions. For the hypodermis, rheological experiments were used to
study the linear and non-linear behavior.
In the following sections the in vitro model (Section 7.2) and the mechanical testing
methods (Section 7.3) are discussed. Thereafter, the implications for clinical and
cosmetic treatments (Section 7.4) and recommendations for further research (7.5) are
provided.
General discussion
91
7.2 In vitro model
An in vitro model enables improved control of the experimental conditions and offers the
potential of performing well-controlled mechanical experiments on a specific skin layer.
Skin obtained from plastic surgery is preferred above cadavers, because the skin has a
higher viability and skin is available from more age groups. However, the number of
body sites is limited. In this thesis, skin obtained from abdominoplastic surgery was used
to study the epidermis and stratum corneum. In obese people, the structure of adipose
tissue has undergone changes in comparison with healthy subjects [17,139]. Therefore, a
porcine model was introduced for this layer. Besides its comparable structure and
function to human adipose tissue [122], the frequent availability and reproducibility of
the samples was considered as very attractive.
After harvesting the skin tissue, the necessary skin layer must be isolated. It has been
observed that five hours after harvesting no viability could no longer observed in ex vivo
mice [140]. As the time between harvesting and mechanical testing is usually too long to
maintain the tissue viability and intregrity, means of preservation were needed as well.
It is essential to ensure that skin preparation treatments does not have an effect on the
mechanical properties. The use of ex vivo human skin in percutaneous and absorption
studies is well established. Current standardized isolation and preparation protocols for
skin [141] are mainly guided by cost and time effectiveness and ease of use. However, it
is widely known and demonstrated that the tissue preparation influences mechanical
properties [142]. In particular, the epidermal layers are known to be highly sensitive for
chemical and physical changes in the environment. Therefore, available and new
techniques to isolate and preserve epidermis and stratum corneum were assessed on their
succesfulness regarding the maintenance of tissue integrity and viability. Furthermore,
the ease of handling and the reproducibility of the protocol was considered. Much
knowledge is already available from skin grafting techniques for burn wounds. However,
the definition of a proper tissue condition is different from in vitro testing [77,54].
From the numerous techniques in use to isolate the epidermis, our studies showed that
the number of available methods dramatically decreases when taking into account the
maintenance of tissue integrity and viability. Cutting using a dermatome and enzymatic
digestion with dispase fulfills both requirements and also score high on ease of handling
and reproducibility. Although dispase causes the cleavage in the basement mebrane and
using a dermatome not, using a dermatome is the only option for mechanical testing. In
that case, the split results into a better well-defined sample geometry. As shown in
Chapter 3 and 4, fragments of papillar dermis present in the epidermal samples did not
influence the results for small strain behavior. As it is proven in these chapters that the
mechanical behavior of stratum corneum and viable epidermis are comparable and both
have a higher stiffness than de dermal layer, it can be assumed that the influence of
fragments of papillar dermis in the samples can be ignored in large deformation studies
as well. Regarding the isolation of the stratum corneum, the golden standard is
enzymatic digestion with 0.1% trypsin. Some other techniques were analyzed as well,
none showed a better performance.
92 Chapter 7
For our studies, it was preferred to store samples in HHBSS at an incubator at 37°C and
5% CO2 after separation. It should be noted that proper storage is much better
achievable than during transport and the mechanical tests. Although some equipment are
now being developed for controlling the environment of biological tissues, this does not
hold for most apparatus yet. Usually, temperature control is built in a device but
implementation of a humidity control sytem remains difficult because it might influence,
for example, the sensitive load cells.
The practical problems that have to be dealt with, emphasizes the importance of careful
handling according to strict protocols for all skin layers. Although the dermatome was
refined by the supplier, the extent of stretching the skin to use the dermatome and its
intrinsic properties cause that the thickness of the epidermis was still variable. Generally,
handling of the sample might induce damage, which influences the outcome of the
mechanical test. The thin fragile stratum corneum easily tears during transport and
cannot be placed in a set-up without using a droplet of water. Skin samples including
reticular dermis curl up and twist, which makes a gently treatment challenging.
Regarding the adipose tissue, every touch causes geometric deformations, which hinders
the correct placement of the the sample in the set-up.
7.3 Mechanical methods
In this thesis, new protocols for available reliable, accurate equipment were developed
for the mechanical characterization of separate skin layers. The traditonal techniques for
the in vitro mechanical characterization of skin layers are uniaxial and biaxial testing.
Uniaxial tensile tests are easy to perform, cost-effective and testing equipment is a
commodity in most biomechanical laboratories. Although uniaxial tensile tests do not
provide sufficient information for a full characterization of the in-plane mechanical
properties, it provides a means for direct comparisons between specimens, body sites,
and the influence of environmental conditions for the various treatments. Biaxial testing
and its interpretation are more difficult and time-consuming to perform. In addition, the
equipment is more expensive and not widely available. Disadvantages of both uniaxial
and biaxial testing are that it is difficult to clamp the samples without influencing the
measurement, to determine the cross area due to the presence of the skin lines and to
define the unloaded configuration because of the natural pre-stress in the skin.
Other techniques, such as indentation and rotational shear, are able to deal with this
issues and are therefore an attractive alternative for axial testing. In addition, smaller
samples can be used. In order to perform these tests on our skin samples, measurement
methods known for their accuracy and reliability from mechanical engineering were
used: the ARES rheometer and MTS NanoIndenter XP. The major measurement
problems were due to the highly non-linear viscoelastic material behavior, the low
stiffness, and the sample thickness and rough surface of the epidermis and stratum
corneum only. The newly developed protocols that were validated with silicone rubbers
General discussion
93
resulted into a set of repoducible data for all measured skin layers. Although some
problems in measuring the linear shear properties of the epidermis were encountered .
In general, rheological experiments aim to characterize the viscoelastic response of soft
materials, requiring relatively large homogeneous samples. To be able to obtain a
homogeneous strain field as well as to increase the accuracy, an eccentric configuration
was used for the upper skin layers. Temperature and humidity could be well regulated by
a home-built system. The measurement chamber with controlled environment could not
be closed completely, because then it would interfere with the applied shear. For that
reason, the temperature and humidity sensors were placed closed to the sample to ensure
a stabile environment in that area. In addition, for the upper skin layers, the required
settings were close to the limitations of the apparatus. The axial resolution is 1 μm,
which is close to the thickness of stratum corneum (10-20 μm). As a result, the
rheometer could not be used to perform compression tests. In addition, there is some
uncertaintity about the shear data for the stratum corneum, because of the thin,
undulating geometry of the sample (Chapter 3). Nonetheless, the obtained resultes were
reproducible, which indicates that the measurement itself is reliable and that a possible
deviation in the measured response is a constant factor.
Many phenomena such as the frequency-dependency and the large deformation behavior
in adipose tissue could not have been measured in vivo and are also difficult to measure
with other in vitro testing techniques. Since the applied protocols did not give a
definitive answer on the non-linear behavior of adipose, another set of experiments
aimed to describe thixotropic behavior needs to be designed. Although thixotropic
studies have been extensively discussed, appropriate protocols for biological tissues are
not yet available.
The MTS NanoIndenter XP is more and more used to probe the mechanical response of
biological materials [s]. Because of the variable probe size, indentation can be used to
measure the mechanical properties from a biological material ranging from cell
membranes up to the global tissue level. In addition, the system is appropriate for thin,
small and heterogeneous samples. This allows testing of tissue specimens that are
unsuitable for traditional mechanical testing techniques. Compared to the rheological
tests on epidermis and stratum corneum, a small region of the sample is loaded with a
relatively large spherical indenter to insist a more homogeneous surface and good
contact during indentation. Because of the sensitivity of the load sensors, it will be
challenging to regulate humidity in the future. Another related problem could be the
definition of the initial sample height, because the role of adhesive forces increases. In
addition, visualization of the experiment is not yet possible. Therefore, other indentation
set ups as developed by Cox et al. [143] might be useful .
From a mechanical point of view, the Nanoindenter XP is a very interesting technique
for further research like the non-linear behavior of the upper skin layers and finally,
failure behavior studies. Wu et al. [34] already developed methods to determine
properties like fracture behavior from load-displacements curves of stratum corneum.
94 Chapter 7
When a direct coupling between structure and loading is essential, other methods needs
to be considered as well.
7.4 Main findings
7.4.1 Small strain behavior of the epidermal layers
In the present study, the stratum corneum and viable epidermis were measured in various
loading directions. The variations within the studies were very small, emphasizing the
reproducibility and reliability of the experimental approach. The Young‟s moduli
derived from shear (in-plane) and indentation (perpendicular) studies are compared with
the tensile Young‟s moduli literature in Table 7.1. For the shear experiments, the
Young‟s modulus was derived from the obtained shear modulus assuming NeoHookean
material behavior, such that 𝐸 = 3𝐺𝑑 . Although some authors have assessed the stiffness
of the (viable) epidermis either in combination with the papillar dermis [38,144,37], this
study provides the first data that are indeed obtained from epidermis only within a small
strain regime.
According to the highly anisotropic structure of the epidermis with the keratinocytes and
corneocytes change in shape over depth, enormous differences in value exist between
loading directions. The differences can be further explained by the fact that different
structural components are the dominant factor during various loading types. The
resistance of the keratinocytes mainly determine the mechanical response during shear,
while the tensile stiffness is determined by the connections bewteen the keratinocytes,
i.e. the desmosomes. As indentation is a mixture of compression, tensile and shear
forces, it is less obvious which structural component is the most dominant factor. The
variability in stiffness for the various loading directions emphasizes the need for an
anisotropic model based on a set of experimental data in all loading directions.
Another important finding is that the stiffness of the viable epidermis has a same order of
magnitude in shear and indentation as the stratum corneum. This implies that the
mechanical behavior of the viable epidermis cannot be ignored in the measured
lengthscales. In addition, it was observed that the shear moduli are decreasing with
increasing humidity, but was hardly influenced by temperature and frequency.
Table 7.1: Overview of measured Young’s moduli in MPa for the stratum corneum and
viable epidermis at room temperature.
SHEAR
INDENTATION
TENSILE
Eshear [MPa] Eanalytical [MPa] EFE-model [MPa] Euniaxial [MPa]
Stratum corneum 25% RH
98% RH
(Viable) epidermis 25% RH
0.03
0.01
0.03
2.5
n.a.*
1.1
1.2
n.a.*
1.2
40-10000
6-10
n.a *
98% RH
0.01
n.a.*
n.a.*
n.a.*
*n.a =not available
General discussion
95
7.4.2 Mechanical behavior of the subcutaneous adipose tissue
For the adipose tissue, the shear modulus is about 8 kPa at 10 rad/s and 20°C and
changes with temperature and frequency. The obtained modulus is in good agreement
with literature [39,42]. Prolonged loading results into a dramatic stiffening of the
material. This behavior is reversible with a recovery time of about 3 hours.
The studies on its non-linear behavior suggest tissue structure changes with increasing
strains. Up to 0.15 strain, the adipose tissue looks like to behave as a Mooney-Rivlin
material. Thereafter, the stress response decreases with increasing strains and becomes
stationary after 0.3 strain. Also this appeared to be reversible behavior.
Although it is generally assumed in literature that adipose tissue behaves non-linear
viscoelastic, these experimental data suggest thixotropic material behavior. As a
consequence, this layer between the skin and muscle tissue cannot be ignored. Before
numerical models can be developed, more experiments are required to fully describe the
non-linear behavior of adipose tissue.
7.5 Implications for clinical and cosmetic applications
The research presented in this thesis is part of larger research programmes being pursued
within Philips Research and Eindhoven University of Technology (TU/e). The relevance
of the work in this thesis was already shown in Chapter 1. In this section, the
implications for some of the applications are discussed.
In Philips Research, part of the innovation is related to consumer products that are in
contact with skin, like shavers. During shaving, the skin penetrates the slits of a shaver in
which the hairs are cut. To enhance shaving performance, the hair must be cut as close to
the skin surface without causing irritation or other damage to the skin. The small length
scale of the doming, which is the penetration of the skin in the slits, requires that the top
layers are included in numerical simulations. To date, the influence of the top layer on
doming has been difficult to incorporate. The underlying tissue during shaving might
vary from soft tissues such as adipose tissue to bone. The material parameters of the
different skin layers obtained in this study are useful to improve numerical models
predicting shaver performance. Moreover, the use of hydrativing additives might affect
the mechanical behavior of the top layers and thereby affect doming.
At the TU/e, an ongoing research programme on the early detection and evaluation of
(deep) presssure ulcers is running. Pressure ulcers are defined as areas of soft tissue
breakdown that result from sustained mechanical loading of skin in shear and
compression and underlying tissues. Until today, this work was mainly focused on early
markers in skin [145,146] and the mechanisms associated with muscle injury
[147,148,149,150]. The poor understanding of the mechanical behavior of adipose has
hampered to involve this layer in the research. The reversible mechanism behavior
during prolonged or severe loading of the subcutaneous fat affect the mechanisms that
are initiated. In this study, the relation between mechanical behavior and tissue damage
96 Chapter 7
could not yet be made. Currently, other research groups gained interest in the role of
subcutaneous adipose tissue [40,121]. Furthermore, in an in vitro study by Ohura et al
[100], the purpose was to estimate the impact of external shear force and pressure on
superficial skin and subcutaneous fat similtaneously. Shear force combined with a small
amount of pressure is accepted as a major factor in the pathogenic mechanism of a
superficial, shallow ulcer or blister. In this thesis, a method is presented to accurately
measure shear forces on epidermis separately. For small strains, the epidermis is much
more stiff than the dermis, which will affect the initiation of a superficial pressure ulcer.
7.6 Recommendations
Some important questions remained unanswered in this thesis. The content of the thesis
was focused on a reliable in vitro mechanical characterization of separate skin layers.
However, to fully understand the mechanical behavior of a heteregeneous sample, it is
necessary to understand how the tissue changes due to mechanical damage. The specific
role of keratinocytes and desmosomes in the epidermis and the role of collagen fibers in
the adipose tissue needs to be fully unraveled. In addition, real-time imaging techniques
can help to solve measurement problems such as for the epidermis at high humidities.
Although a variety of imaging techniques are available, factors such as the depth of
imaging, resolution, field of view and the sample rate frequency are limiting the use for
visualization during mechanical testing of the epidermis. For the epidermal layers, it
would be interesting to track cell shape deformations by multiphoton laser scanning
microscopy, allowing visualization of cellular and subcellular structures of the epidermis
and upper dermis [source]. In addition, confocal imaging techniques are able to track the
cell nuclei with more than 10 images per second [source]. Both techniques have the
advantage that images can be obtained from intrinsic tissue properties only. This makes
those techniques also appropriate for in vivo imaging and in particular, . Another
imaging technique combining second-harmonic signal and 2-photon imaging is
developed by Palero et al. [thesis], who demonstrated on both in vivo and ex vivo
epidermal tissue from mouse that the viability of cells and the cell membranes could be
measured simultaneously. In the far future, this technique is very attractive for failure
studies.
Before visualization of the mechanical tests on adipose tissue can be performed, it is
recommended to study the fiber network surrounding the adipocytes. The relative large
structures, i.e. adipocytes have a diameter up to 70 μm and are collected in lobs, limit the
number of possible techniques. For instance, histological examination and confocal
microscopy cannot visualize the three-dimensional structure of the collagen fiber
network. Another problem is that current staining probes cannot enter thick native tissue
[bron Anita]. If this can be solved, then three-dimensional techniques such as optical
projection tomography can be useful. In the meantime, it does make sense to study the
geometric deformations of the adipose tissue samples during mechanical behavior. In
General discussion
97
particular, the response on stiffening during prolonged loading and the different phases
with increasing strains have to be studied.
In this thesis, only the small deformation behavior of the upper skin layers was studied.
For clinical and cosmetic applications, it is essential to study the non-linear behavior of
those layers as well. In principle, the experimental approaches presented in this thesis
can be used to develop testing series for the non-linear region. Ultimately, experimental
studies on the failure behavior is necessary. When doing so, also transport models and
structural damage should be incorporated. Therefore, it is desirable to perform those
studies also on in vitro human skin.
Our tests proved that the non-linear behavior of adipose tissue is rather complex and
cannot be captured in a constitutive model yet. Therefore, a new set of experimental data
have to be collected to be able to built a constitutive model in the future. An overview of
these type of tests is described by Dullaert et al. [151,152]. Mechanical tests in other
loading directions should be performed as well. Compression tests are most relevant to
clinical and cosmetic applications and can also be performed on a rheometer.
In addition, this thesis demonstrated mechanical parameters for abdominal skin from
Caucasian women in the age group of 35-55 years. Skin with striae, cellulite, UV
damage or excessively hairy skin was excluded from the study. Other studies should
include other skin types, other body sites with a high density of hairs or UV exposure,
ageing effects, and so on.
Ultimately, a full-thickness constitutive model consisting of individual skin layers may
be developed not only to study damage development, but also to serve as a model for
investigating new prevention and treatment strategies. For applications such as
decubitus, it is advised to incorporate transport models, which offers the potential to
asses the correlation between tissue damage and biological markers.
7.7 General conclusion
This thesis presents methods to determine mechanical properties of individual skin layers
in vitro. The two main findings are: 1) the stratum corneum and viable epidermis behave
highly anisotropic in the small strain behavior and the stiffness of the viable epidermis is
equivalent to that of the stratum corneum in each loading direction, and 2) the
hypodermis initially shows typical small strain behavior for soft tissues, but seems to
behave thixotropically during prolonged deformation and for larger strains. These two
main findings highlight the importance of mechanical characterization of individual skin
layers as well as the need for anisotropic models involving separate skin layers in
numerical simulations. The used experimental methods represent valuable tools for
studying the mechanical properties in relation to disease and treatments in future.
98 Chapter 7
Samenvatting
De mechanische eigenschappen van de menselijke huid zijn van belang voor vele
klinische en cosmetische toepassingen. Vaak wordt de huid beschouwd als één geheel,
maar inmiddels is gebleken dat het voor diverse toepassingen van belang is het
mechanische gedrag van de afzonderlijke huidlagen te begrijpen. Voorbeelden hiervan
zijn: het toedienen van medicijnen door de huid, de interactie tussen de huid en een
(scheer)apparaat en de preventie en behandeling van doorligwonden. Totnutoe is veel
onderzoek naar de mechanische eigenschappen van de huid uitgevoerd in an in vivo
situatie waarbij aangenomen dat de middelste huidlaag met zijn vezelstructuur
representatief is voor de huid in zijn geheel. Het doel van dit promotieonderzoek was om
de mechanische eigenschappen van de afzonderlijke huidlagen te karakteriseren. Hierbij
is specifiek gericht op die huidlagen waarvan nog nauwelijks literatuur beschikbaar is of
de resultaten in de literatuur inconsistent zijn.
Allereerst is er onderzocht wat de beste methoden zijn om de verschillende huidlagen
van elkaar te scheiden en levensvatbaar te houden in an in vitro omgeving.
Aandachtspunten hierbij waren het effect van een methode op de weefselstructuur en de
levensvatbaarheid alsook de betrouwbaarheid, duur en de mate van uitvoerbaarheid van
een methode. Hieruit kon geconcludeerd worden dat voor dit onderzoek de epidermis het
beste geisoleerd kan worden met een dermatoom. Vervolgens is de epidermis bewaard in
een medium, HHBSS, of door in te vriezen volgens een specifieke protocol. De stratum
corneum kan van de epidermis geisoleerd te worden door gebruik te maken van het
enzym trypsin om vervolgens bewaard te worden in een saline buffer, PBS, of in
gedroogde vorm.
Vervolgens zijn er verschillende methoden ontwikkeld om de mechanische reactie van
de afzonder huidlagen te kunnen meten. Voor de bovenste huidlagen, de opperhuid en
hoornlaag, zijn in vitro meetopstellingen gebouwd om de respons op kleine rekken te
kunnen meten. In het dagelijks leven, worden grote rekken in principe opgevangen door
het ontvouwen van het huidoppervlak en dus alleen bij niet-fysiologische
omstandigheden, zoals een naald door de huid prikken, zal de epidermis grote rekken
ondergaan. Omdat schuif en druk sterk aan elkaar gerelateerd zijn en het bekend is dat de
opperhuid een inhomogene gelaagde structuur heeft, is er gekozen voor het opleggen van
zowel een schuif- als indentatiebelasting. Voor beide soorten belasting is aangetoond dat
100 Samenvatting
er geen significant verschil is tussen de mechanische eigenschappen van de epidermis en
stratum corneum. Verder bleken deze huidlagen bij een schuifrek wel gevoelig voor
vochtigheid maar niet voor temperatuuur. Als de kracht loodrecht op de huid staat,
gedraagt de opperhuid zich veel stijver dan bij het opleggen van een schuifrek. De
uitkomsten van deze experimenten tonen aan dat het essentieel is om het anisotrope
gedrag van deze afzonderlijke huidlagen mee te nemen in numerieke huidmodellen.
De onderhuidse vetlaag is belast met kleine en grote schuifrekken gedurende korte en
lange termijn. De frequentie en temperatuur afhankelijkheid van de mechanische
parameters zijn gemeten bij kleine rekken. Het is gebleken dat al bij zeer kleine rekken
de onderhuidse vetlaag ernstig gaat vervormen na langdurige belasting, maar dat na een
rustperiode het gedrag reversibel is. Dit duidt erop dat er veranderingen in de
weefselstructuur optreden door mechanische belasting maar zonder blijvende schade.
Ook het opleggen van grote schuifrekken resulteerde in veranderingen in de
weefselstructuur die reversibel bleken. Tot zekere schuifrekken is het gedrag van
onderhuidsvet vergelijkbaar met andere zachte lichaamsweefsels. Bij zeer hoge
schuifrekken wordt het materiaalgedrag complexer. Om dit goed te kunnen begrijpen,
zijn er eerst meer experimenten nodig voordat er numerieke modellen gebouwd kunnen
worden die ook deze grote schuifrekken kunnen beschrijven. Een goede basis voor een
numeriek model zou een Mooney-Rivlin of powerlaw model kunnen zijn.
In dit proefschrift zijn mechanische eigenschappen van individuele huidlagen bepaald in
een in vitro omgeving met behulp van nauwkeurige apparatuur, resulterend in
reproduceerbare resultaten. Het wordt aanbevolen om in de toekomst de relatie tussen de
deformatie in het weefsel en het mechanisch gedrag te bestuderen met behulp van
visualizatietechnieken. Daarnaast zal het onderzoek uitgebreid moeten worden met
studies naar het faalgedrag van de individuele huidlagen in relatie tot klinische en
cosmetische applications.
Dankwoord
Graag wil ik iedereen bedanken die direct of indirect een bijdrage heeft geleverd aan de
totstandkoming van dit proefschrift. Een aantal mensen wil ik specifiek bedanken.
Allereerst wil ik Frank en Paco bedanken voor het mogelijk maken van mijn project
binnen deze bijzondere constructie tussen Philips en TU/e. Door de samenwerking heb ik
kunnen profiteren van de faciliteiten van beide zijdes alsook van de kennis over de huid
als van de (bio)mechanica. Cees, bedankt voor het vertrouwen en je positieve
relativerende kijk op zaken. Zonder jou en Sigi had ik de stap om te gaan promoveren
nooit genomen. Gerrit, bedankt voor het kijkje in de wereld van de polymeren en
rheologie. Hoewel ik je kunstzinnige hierogliefen tegenwoordig lees alsof het
geschreven is in Times New Roman is, zal ik ze toch gaan missen! Paul, ik vind je
enthousiasme, vertrouwen, en kritische blik altijd erg bijzonder. Bedankt dat je altijd
voor me klaar stond! Dear Dan, I really appreciate your contribution to my thesis.
Daarnaast zijn er nog een aantal mensen die me op praktisch vlak vooruit hebben
geholpen. Hoewel al een poosje terug, wil ik Matej en Evelyne bedanken voor de
kennismaking met het meten aan zachte weefsels aan de rheometer. Ik wist toen nog niet
dat het rheohok mijn huiskamer zou gaan worden. Henk en ook de mannen van de TU
werkplaats, bedankt voor de mooie verzameling rheometer hulpstukken! Lambert, we
hadden samen een voorbeeldig MaTe-project met jouw W en mijn BMT achtergrond, en
dan ook nog experimenteel en numeriek. Jan, ik ben zeer blij dat mijn
statistiekproblemen voor jouw een wetenschappelijke uitdaging waren. Sjoerd, zullen we
nog een keer een speklapje opeten, terwijl je de kurkboor scherp maakt? Sarita, bedankt
voor al het snij- en kleurwerk dat je voor me gedaan hebt. Henny, jouw tekeningen
hebben dit boekje aanzienlijk opgefleurd. Ik wil de stagaires Francois, Roman Ditmar en
Suzanne Stolk en verscheidene derdejaars projectgroepjes bedanken voor hun bijdrage in
het onderzoek. Lisette, Debbie, Roel en Susanne, fijn dat er ook andere mensen met ex
vivo huid bezig waren. Anke, jij bent ook zeker een bedankje waard.
In een samenwerkingsverband tussen Philips en TU/e heb ik veel dubbel mogen beleven.
Het is erg bijzonder om te werken in twee groepen met zoveel collegae. Ik zou mijn
kamergenootjes bij Philips alsook op de TU/e specifiek willen bedanken voor hun
gezelschap. Rachel, goed bezig! I‟m glad that someone invented Facebook to keep
sharing our daily complaints and gossips! Anke en Maria, ik blijf het leuk vinden om af
102 Dankwoord
en toe het vijfde wiel aan de wagen te zijn en hoop dan ook dat er nog veel etentjes
komen!
Ik wil het personeel van de afdelingen plastische chirurgie en de operatiekamers in het
Catharina Ziekenhuis in Eindhoven bedanken voor alle emmertjes met huid. In het
bijzonder de plastisch chirurgen Van Rappard en Hoogbergen die deze samenwerking
mogelijk hebben gemaakt alsook Marjolein (en je directe collega‟s) en de OK-receptie
voor alle telefoongesprekken.
Lieve OLT en andere scoutingvriendjes, het is erg relativerend om een potje te koken en
een biertje te drinken in het bos, bij een kampvuur, in de disco of in de kroeg. Na al die
jaren en kampen blijft het gezellig en voor mij erg waardevol! Vrouwenweekendjes (en
de autorit heen en terug, Margo!) ben ik ook gaan waarderen. Daarnaast is het erg leuk
om in de wachttijd van een experiment over de scoutingorganisatie na te denken:
regiegroep, grote kampen, Georgie, enz., enz. Peter, mutsen en onderbroeken staan
garant voor leuke herinneringen. Ik ben benieuwd welke kledingstukken we de komende
jaren er nog bij weten te verzamelen.
Frank, Pe, Xander en Elizabeth, Nicole, en alle anderen bedankt voor jullie interesse in
mijn onderzoek. Lieve Iksiks, zonder Betty Boo en mijn roze kledingset was mijn
promotietijd toch een stuk minder vrolijk geweest! Nicole en Jannet, ik heb weer tijd
voor onze etentjes en bezoekjes aan ons wereldwijde vriendennetwerk (sorry!). Lieve
Rianne, ik heb weer zeeen van tijd voor onzinnige projectjes. Ook mijn wandelstokken
en bergschoenen staan te popelen (wordt het een graad 4?). Lieve papa en mama,
dankjullie wel voor jullie geduld. Het komt wel goed met me. Gerrie, Dick en Sebas, het
is erg ontspannend om met zo‟n gezellige schoonfamilie op stap te zijn!
Lieve Martijn, altijd komt toch alles goed? Maar eerlijk is eerlijk, zonder jouw luisterend
oor (ergens in een auto), je releativerende woorden en onvoorwaardelijke steun had ik
het nooit gered. Ga je mee naar Nice?
Marion Geerligs,
Eindhoven, november 2009.
Curriculum Vitae
Marion Geerligs is geboren op 21 juni 1979 in Hoogezand-Sappemeer. In 1998 behaalde
zij haar Gymnasium diploma aan het CSG Vincent van Gogh in Assen. Aansluitend
studeerde zij een jaar Bewegingswetenschappen aan de Vrije Universiteit Amsterdam.
Na een jaar besloot zij over te stappen naar de studie Biomedische Technologie aan de
Technische Universiteit Eindhoven. Als onderdeel van deze studie liep zij stage in het St.
Mary Hospital in Mumias (Kenia), waar zij onderzoek deed naar de preventie van
doorligwonden bij aan bedgebonden patienten. Haar afstudeerwerk richtte zich op het
ontwerpen van een testobject voor ge-automatiseerd bloed prikken waarin de
mechanische en ultrasoundeigenschappen van de huid, vet, vaatwand en bloed werden
nagebootst. Dit onderzoek werd uitgevoerd binnen de groep Care & Health Applications
van Philips Research. Vanwege haar interesse in het onderzoek naar de biomechanica
van zachte weefsels besloot zij in 2005 verder te gaan met een promotieonderzoek bij
dezelfde groep in een samenwerkingsverband met de Technische Universiteit
Eindhoven.
References
[1] Berson, M., Vabre, V., Karlsson, B., Gregoire, J. M., Yvon, C., Patat, F., Auriol, F.,
Vaillant, L., Diridollou, S., Black, D., and Gall, Y., (1998). An in vivo method for
measuring the mechanical properties of the skin using ultrasound. Ultrasound in
Medicine and Biology, 215-224, 24(2), 215-224.
[2] Plewig, G. and Marples, R. R., (1970). Regional differences of cell sizes in the
human stratum corneum. I. J.Invest Dermatol., 13-18, 54(1), 13-18.
[3] Holbrook, K. A. and Odland, G. F., (1974). Regional differences in the thickness
(cell layers) of the human stratum corneum: an ultrastructural analysis. J.Invest
Dermatol., 415-422, 62(4), 415-422.
[4] Brody, I., (1977). Ultrastructure of the stratum corneum. Int.J.Dermatol., 245-256,
16(4), 245-256.
[5] Agache, P. and Humbert, P. (2004). Measuring the skin. Springer,
[6] Plewig, G., (1970). Regional differences of cell sizes in the human stratum
corneum. II. Effects of sex and age. J.Invest Dermatol., 19-23, 54(1), 19-23.
[7] Rawlings, A. V., Scott, L. R., Harding, C. R., and Bowser, P. A., (1994). Stratum
corneum moisturization at the molecular level. Journal of Investigative
Dermatology, 731-740, 103(5), 731-740.
[8] Agache, P., Boyer, J. P., and Laurent, R., (1973). Biomechanical properties and
microscopic morphology of human stratum corneum incubated on a wet pad in
vitro. Archives of Dermatological Research, 271-283, 246(3), 271-283.
[9] Marks, R., (2004). The stratum corneum barrier: The final frontier. Journal of
Nutrition, 134(8 SUPPL.),
[10] Elias, P. M. and Feingold, K. R. (2005). Skin Barrier. Informa Healthcare,
References 105
[11] Blank, I. H., Moloney, J., III, Emslie, A. G., Simon, I., and Apt, C., (1984). The
diffusion of water across the stratum corneum as a function of its water content.
J.Invest Dermatol., 188-194, 82(2), 188-194.
[12] Sandby-M°ller, J., Wulf, H. C., and Poulsen, T., (2003). Epidermal Thickness at
Different Body Sites: Relationship to Age, Gender, Pigmentation, Blood Content,
Skin Type and Smoking Habits. Acta Dermato-Venereologica, 410-413, 83(6), 410413.
[13] Silver, F. H., Siperko, L. M., and Seehra, G. P., (2003). Mechanobiology of force
transduction in dermal tissue. Skin Research and Technology, 3-23, 9(1), 3-23.
[14] Montagna, W., Kligman, A. M., and Carlisle, K. S. (1992). Atlas of Normal Human
Skin. Springer Verlag, New York.
[15] Odland, G., (1991). Structure of skin. In: Goldsmith, L. A., Physiology,
biochemistry, and molecular biology of the skin. Oxford University Press, Oxford.
[16] Briggaman, R. A. and Wheeler, C. E., Jr., (1975). The epidermal-dermal junction.
J.Invest Dermatol., 71-84, 65(1), 71-84.
[17] Avram, A. S., Avram, M. M., and James, W. D., (2005). Subcutaneous fat in
normal and diseased states: 2. Anatomy and physiology of white and brown adipose
tissue. Journal of the American Academy of Dermatology, 671-683, 53(4), 671683.
[18] Ann L.Albright & Judith S.Stern, Dept of Nutrition and Internal Medicine
Unicersity of California at Davis,(2005). adipose tissue.
[19] Shen, W., Wang, Z, Punyanita, M, Lei, J., Sinav, A, Kral, J. G., Imielinska, C.,
Ross, R., and Heymsfield, S. B., (2005). Adipose tissue quantification by imaging
methods. Obesity research, 5-11, 11(1), 5-11.
[20] Elsner, P., Berardesca, E., Wilhelm, K. P., and Maibach, H. I. (2002).
Bioengineering of the skin; Skin Biomechanics. CRC Press,
[21] Hendriks, F. M. (2005). Mechanical Behaviour of Human Epidermal and Dermal
Layers. Ph.D. thesis, Technische Universiteit Eindhoven.
[22] Wildnauer, R. H., Bothwell, J. W., and Douglass, A. B., (1971). Stratum corneum
biomechanical properties. I. Influence of relative humidity on normal and extracted
human stratum corneum. Journal of Investigative Dermatology, 72-78, 56(1), 7278.
106 References
[23] Park, A. C. and Baddiel, C. B., (1972). Rheology of stratum corneum-I: A
molecular interpretation of the stress-strain curve. Jourbal of Society Cosmetical
Chemistry, 3-12, 233-12.
[24] Papir, Y. S., Hsu, K. H., and Wildnauer, R. H., (1975). The mechanical properties
of stratum corneum. I. The effect of water and ambient temperature on the tensile
properties of newborn rat stratum corneum. Biochimica et Biophysica Acta, 170180, 399(1), 170-180.
[25] Wolfram, M. A., Wolejsza, N. F., and Laden, K., (1972). Biomechanical properties
of delipidized stratum corneum. J.Invest Dermatol., 421-426, 59(6), 421-426.
[26] Van Duzee, B. F., (1978). The influence of water content, chemical treatment and
temperature on the rheological properties of stratum corneum. Journal of
Investigative Dermatology, 140-144, 71(2), 140-144.
[27] de Rigal, J. and Leveque, J. L., (1985). In vivo measurement of the stratum
corneum elasticity. Bioengineering and the Skin, 13-23, 1(1), 13-23.
[28] Agache, P. G., Monneur, C., Leveque, J. L., and de Rigal, J., (1980). Mechanical
properties and Young's modulus of human skin in vivo. Archives of Dermatological
Research, 221-232, 269(3), 221-232.
[29] Leveque, J. L., de Rigal, J., Agache, P. G., and Monneur, C., (1980). Influence of
ageing on the in vivo extensibility of human skin at a low stress. Archives of
Dermatological Research, 127-135, 269(2), 127-135.
[30] Pierard, G. E., Nikkels-Tassoudji, N., and Pierard-Franchimont, C., (1995).
Influence of the test area on the mechanical properties of skin. Dermatology, 9-15,
191(1), 9-15.
[31] Yonghui, Y. and Verma, R., (2003). Mechanical properties of stratum corneum
studied by nano-indentation. Spatially Resolved Characterization of Local
Phenomena in Materials and Nanostructures.Symposium
(Mater.Res.Soc.Symposium Proceedings Vol.738), 265-272, 265-272.
[32] Gardner, T. N. and Briggs, G. A. D., (2001). Biomechanical measurements in
microscopically thin stratum corneum using acoustics. Skin Research and
Technology, 254-261, 7(4), 254-261.
[33] Graves, C. J. and Edwards, C., (2002). Hardware and Measuring Principles: The
Microindentometer. In: Elsner, P., Berardesca, E., Wilhelm, K. P., and Maibach, H.
I., Bioengineering of the skin: skin biomechanics. CRC Press LLC, Boca Raton.
References 107
[34] Wu, K. S., van Osdol, W. W., and Dauskardt, R. H., (8-8-2005). Mechanical
properties of human stratum corneum: Effects of temperature, hydration, and
chemical treatment. Biomaterials,
[35] Koutroupi, K. S. and Barbenel, J. C., (1990). Mechanical and failure behaviour of
the stratum corneum. J.Biomech., 281-287, 23(3), 281-287.
[36] Middleton, J. D., (1969). The effect of temperature on extensibility of isolated
corneum and its relation to skin chapping. Br.J.Dermatol., 717-721, 81(10), 717721.
[37] Kendall, M. A., Chong, Y. F., and Cock, A., (2007). The mechanical properties of
the skin epidermis in relation to targeted gene and drug delivery. Biomaterials,
4968-4977, 28(33), 4968-4977.
[38] Hendriks, F. M., Oomens, C. W. J., Bader, D. L., Baaijens, F. P. T., and Brokken,
D., (2006). The relative contributions of different skin layers to the mechanical
behavior of human skin in vivo using suction experiments. Medical Engineering
and Physics, 259-266, 28(3), 259-266.
[39] Patel, P. N., Smith, C. K., and Patrick, C. W., (2005). Rheological and recovery
properties of poly(ethylene glycol) diacrylate hydrogels and human adipose tissue.
Journal of Biomedical Materials Research Part A, 313-319, 73A(3), 313-319.
[40] Gefen, A. and Haberman, E., (2007). Viscoelastic properties of ovine adipose tissue
covering the gluteus muscles. Journal of Biomechanical Engineering-Transactions
of the Asme, 924-930, 129924-930.
[41] Samani, A. and Plewes, D., (21-9-2004). A method to measure the hyperelastic
parameters of ex vivo breast tissue samples. Phys.Med.Biol., 4395-4405, 49(18),
4395-4405.
[42] Samani, A., Zubovits, J., and Plewes, D., (21-3-2007). Elastic moduli of normal and
pathological human breast tissues: an inversion-technique-based investigation of
169 samples. Phys.Med.Biol., 1565-1576, 52(6), 1565-1576.
[43] Samani, A., Zubovits, J., and Plewes, D., (2007). Elastic moduli of normal and
pathological human breast tissues: an inversion-technique-based investigation of
169 samples. Physics in Medicine and Biology, 1565-1576, 52(6), 1565-1576.
[44] Van Houten, E. E., Doyley, M. M., Kennedy, F. E., Weaver, J. B., and Paulsen, K.
D., (2003). Initial in vivo experience with steady-state subzone-based MR
elastography of the human breast. J.Magn Reson.Imaging, 72-85, 17(1), 72-85.
108 References
[45] Skerrow, D. and Skerrow, S. J., (1985). A Survey of Methods for the Isolation and
Fractionation of Epidermal Tissue and Cells. In: D.Skerrow and C.J.Skerrow,
Methods in Skin Research. John Wiley & Sons Ltd., New York.
[46] Messager, S., Hann, A. C., Goddard, P. A., Dettmar, P. W., and Maillard, J. Y.,
(2003). Assessment of skin viability: is it necessary to use different methodologies?
Skin Research and Technology, 321-330, 9(4), 321-330.
[47] Basaran, O., Ozdemir, H., Kut, A., Sahin, F. I., Deniz, M., Sakallioglu, E. A., and
Haberal, M. A., (2006). Effects of different preservation solutions on skin graft
epidermal cell viability and graft performance in a rat model. Burns, 423-429,
32(4), 423-429.
[48] Collier, S. W., Sheikh, N. M., Sakr, A., Lichtin, J. L., Stewart, R. F., and Bronaugh,
R. L., (1989). Maintenance of skin viability during in vitro percutaneous
absorption/metabolism studies. Toxicol.Appl.Pharmacol., 522-533, 99(3), 522-533.
[49] D.Skerrow and C.J.Skerrow, (1985). A Survey of Methods for the Isolation and
Fractionation of Epidermal Tissue and Cells. In: D.Skerrow and C.J.Skerrow,
Methods in Skin Research. John Wiley & Sons Ltd., New York.
[50] Willsteed, E. M., Bhogal, B. S., Das, A., Bekir, S. S., Wojnarowska, F., Black, M.
M., and McKee, P. H., (1991). An ultrastructural comparison of dermo-epidermal
separation techniques. J.Cutan.Pathol., 8-12, 18(1), 8-12.
[51] Steinstrasser, I. and Merkle, H. P., (1995). Dermal metabolism of topically applied
drugs: pathways and models reconsidered. Pharm.Acta Helv., 3-24, 70(1), 3-24.
[52] Ogura, R., Knox, J. M., and GRIFFIN, A. C., (1960). Separation of epidermis for
the study of epidermal sulfhydryl. J.Invest Dermatol., 239-243, 35239-243.
[53] VAN SCOTT, E. J., (1952). Mechanical separation of the epidermis from the
corium. J.Invest Dermatol., 377-379, 18(5), 377-379.
[54] Bravo, D., Rigley, T. H., Gibran, N., Strong, D. M., and Newman-Gage, H., (2000).
Effect of storage and preservation methods on viability in transplantable human
skin allografts. Burns, 367-378, 26(4), 367-378.
[55] Kiistala, U., (1968). Suction Blister Device for Separation of Viable Epidermis
from Dermis. Journal of Investigative Dermatology, 129-&, 50(2), 129-&.
[56] Beerens, E. G., Slot, J. W., and van der Leun, J. C., (1975). Rapid regeneration of
the dermal-epidermal junction after partial separation by vacuum: an electron
microscopic study. J.Invest Dermatol., 513-521, 65(6), 513-521.
References 109
[57] COWDRY, E. V., CARRUTHERS, C., and SUNTZEFF, V., (1948). Influence of
age on the copper and zinc contest in the epidermis of mice undergoing
carcinogenesis with methylcholanthrene and a note on the role of calcium.
J.Natl.Cancer Inst., 209-213, 8(5-6), 209-213.
[58] Baumberger, J. P., Suntzeff, V., and Cowdry, E. V., (1965). Methods for the
Separation of epidermis from Dermis and Some Physiologic and Chemical
Properties of Isolated Epidermis. Journal of the National Cancer Institute, 413-423,
413-423.
[59] Dimond, R. L., Erickson, K. L., and Wuepper, K. D., (1976). The role of divalent
cations in epidermolysis. Br.J.Dermatol., 25-34, 95(1), 25-34.
[60] Scaletta, L. J., Occhino, J. C., MacCallum, D. K., and Lillie, J. H., (1978). Isolation
and immunologic identification of basement membrane zone antigens from human
skin. Lab Invest, 1-9, 39(1), 1-9.
[61] Woodley, D., Sauder, D., Talley, M. J., Silver, M., Grotendorst, G., and
Qwarnstrom, E., (1983). Localization of basement membrane components after
dermal-epidermal junction separation. J.Invest Dermatol., 149-153, 81(2), 149-153.
[62] Baden, H. P. and Gifford, A. M., (1970). Isometric contraction of epidermis and
stratum corneum with heating. J.Invest Dermatol., 298-303, 54(4), 298-303.
[63] Wester, R. C., Christoffel, J., Hartway, T., Poblete, N., Maibach, H. I., and Forsell,
J., (1998). Human cadaver skin viability for in vitro percutaneous absorption:
storage and detrimental effects of heat-separation and freezing. Pharm.Res., 82-84,
15(1), 82-84.
[64] Briggaman, R. A., (1982). Biochemical composition of the epidermal-dermal
junction and other basement membrane. J.Invest Dermatol., 1-6, 78(1), 1-6.
[65] Walzer, C., Benathan, M., and Frenk, E., (1989). Thermolysin treatment: a new
method for dermo-epidermal separation. J.Invest Dermatol., 78-81, 92(1), 78-81.
[66] Liu, S. C. and Karasek, M., (1978). Isolation and growth of adult human epidermal
keratinocytes in cell culture. J.Invest Dermatol., 157-162, 71(2), 157-162.
[67] Stenn, K. S., Link, R., Moellmann, G., Madri, J., and Kuklinska, E., (1989).
Dispase, a neutral protease from Bacillus polymyxa, is a powerful fibronectinase
and type IV collagenase. J.Invest Dermatol., 287-290, 93(2), 287-290.
[68] Kitano, Y. and Okada, N., (1983). Separation of the epidermal sheet by dispase.
Br.J.Dermatol., 555-560, 108(5), 555-560.
110 References
[69] Geggel, H. S. and Gipson, I. K., (1985). Removal of viable sheets of conjunctival
epithelium with dispase II. Invest Ophthalmol.Vis.Sci., 15-22, 26(1), 15-22.
[70] Sanchez, M. A., Diaz, N. L., and Tapia, F. J., (2002). Microwave irradiation for
rapid epidermis-dermis separation and improved epidermal cell immunodetection.
Biotech.Histochem., 183-187, 77(4), 183-187.
[71] Kholodnykh, A. I., Petrova, I. Y., Larin, K. V., Motamedi, M., and Esenaliev, R.
O., (1-6-2003). Precision of measurement of tissue optical properties with optical
coherence tomography. Appl.Opt., 3027-3037, 42(16), 3027-3037.
[72] Ferguson, J., (1977). A method to facilitate the isolation and handling of stratum
corneum. British Journal of Dermatology, 21-23, 96(1), 21-23.
[73] Humphries, W. T. and Wildnauer, R. H., (1971). Thermomechanical analysis of
stratum corneum. I. Technique. J.Invest Dermatol., 32-37, 57(1), 32-37.
[74] Humphries, W. T. and Wildnauer, R. H., (1971). Thermomechanical analysis of
stratum corneum. I. Technique. J.Invest Dermatol., 32-37, 57(1), 32-37.
[75] Nicollier, M., Agache, P., Kienzler, J. L., Laurent, R., Gibey, R., Cardot, N., and
Henry, J. C., (1980). Action of trypsin on human plantar stratum corneum. An
ultrastructural study. Arch.Dermatol.Res., 53-64, 268(1), 53-64.
[76] Morejohn, L. C. and Pratley, J. N., (18-5-1979). Differential effects of trypsin on
the epidermis of Rana catesbeiana. Observations on differentiating junctions and
cytoskeletons. Cell Tissue Res., 349-362, 198(2), 349-362.
[77] Kearney, J. N., (2005). Guidelines on processing and clinical use of skin allografts.
Clin.Dermatol., 357-364, 23(4), 357-364.
[78] Zieger, M. A. J., Tredget, E. E., and McGann, L. E., (1996). Mechanisms of
cryoinjury and cryoprotection in split-thickness skin. Cryobiology, 376-389, 33(3),
376-389.
[79] Ben Bassat, H., (2005). Performance and safety of skin allografts. Clin.Dermatol.,
365-375, 23(4), 365-375.
[80] Meyer, W., Zschemisch, N. H., and Godynicki, S., (2003). The porcine ear skin as a
model system for the human integument: influence of storage conditions on basic
features of epidermis structure and function--a histological and histochemical study.
Pol.J.Vet.Sci., 17-28, 6(1), 17-28.
References 111
[81] De, A., Mathur, M., and Gore, M. A., (2008). Viability of cadaver skin grafts stored
in skin bank at two different temperatures. Indian J.Med.Res., 769-771, 128(6),
769-771.
[82] Castagnoli, C., Alotto, D., Cambieri, I., Casimiri, R., Aluffi, M., Stella, M., Alasia,
S. T., and Magliacani, G., (2003). Evaluation of donor skin viability: fresh and
cryopreserved skin using tetrazolioum salt assay. Burns, 759-767, 29(8), 759-767.
[83] Han, B. and Bischof, J. C., (2004). Engineering Challenges in Tissue Preservation.
Cell Preservation Technology, 91-112, 2(2), 91-112.
[84] Lawrence, J. C., (1972). Storage and skin metabolism. Br.J.Plast.Surg., 440-453,
25(4), 440-453.
[85] Foutz, T. L., Stone, E. A., and Abrams, Jr, (1992). Effects of freezing on
mechanical properties of rat skin. American journal of veterinary research, 788-792,
53(5), 788-792.
[86] Bouwstra, J. A.,(2009).
[87] Bouwstra, J. A., de Graaff, A., Gooris, G. S., Nijsse, J., Wiechers, J. W., and van
Aelst, A. C., (2003). Water distribution and related morphology in human stratum
corneum at different hydration levels. Journal of Investigative Dermatology, 750758, 120(5), 750-758.
[88] Chistolini, P., De Angelis, G., De Luca, M., Pellegrini, G., and Ruspantini, I.,
(1999). Analysis of the mechanical properties of in vitro reconstructed epidermis:
preliminary results. Medical & Biological Engineering & Computing, 670-672,
37(5), 670-672.
[89] Rochefort, A., Druot, P., Leduc, M., Vassalet, R., and Agache, P., (1986). A New
Technique for the Evaluation of Cosmetics Effect on Mechanical-Properties of
Stratum-Corneum and Epidermis Invitro. International Journal of Cosmetic
Science, 27-36, 8(1), 27-36.
[90] Geerligs, M,(2007). Mechanical behaviour of human skin layers: stratum corneum,
living epidermis and hypodermis. A literature review. PR-TN 2006/00450(
[91] van Turnhout, M., Peters, G., Stekelenburg, A., and Oomens, C., (2005). Passive
transverse mechanical properties as a function of temperature of rat skeletal muscle
in vitro. Biorheology, 193-207, 42(3), 193-207.
[92] Hrapko, M., van Dommelen, J. A., Peters, G. W., and Wismans, J. S., (2006). The
mechanical behaviour of brain tissue: large strain response and constitutive
modelling. Biorheology, 623-636, 43(5), 623-636.
112 References
[93] van Dam, E. A., Dams, S. D., Peters, G. W., Rutten, M. C., Schurink, G. W., Buth,
J., and van de Vosse, F. N., (2006). Determination of linear viscoelastic behavior of
abdominal aortic aneurysm thrombus. Biorheology, 695-707, 43(6), 695-707.
[94] Holt, B., Tripathi, A., and Morgan, J., (28-8-2008). Viscoelastic response of human
skin to low magnitude physiologically relevant shear. J.Biomech., 2689-2695,
41(12), 2689-2695.
[95] Gennisson, J. L., Baldeweck, T., Tanter, M., Catheline, S., Fink, M., Sandrin, L.,
Cornillon, C., and Querleux, B., (2004). Assessment of elastic parameters of human
skin using dynamic elastography. Ieee Transactions on Ultrasonics Ferroelectrics
and Frequency Control, 980-989, 51(8), 980-989.
[96] Park, A. C. and Baddiel, C. B., (1972). Rheology of Stratum Corneum .2.
Physicochemical Investigation of Factors Influencing Water-Content of Corneum.
Journal of the Society of Cosmetic Chemists, 13-&, 23(1), 13-&.
[97] Wu, K. S., Li, J., Ananthapadmanabhan, K. P., and Dauskardt, R. H., (2007). Timedependant intercellular delamination of human stratum corneum. Journal of
Materials Science, 8986-8994, 42(21), 8986-8994.
[98] vanHal, D. A., Jeremiasse, E., Junginger, H. E., Spies, F., and Bouwstra, J. A.,
(1996). Structure of fully hydrated human stratum corneum: A freeze fracture
electron microscopy study. Journal of Investigative Dermatology, 89-95, 106(1),
89-95.
[99] Warner, R. R., Stone, K. J., and Boissy, Y. L., (2003). Hydration disrupts human
stratum corneum ultrastructure. Journal of Investigative Dermatology, 275-284,
120(2), 275-284.
[100] Ohura, T., Takahashi, M., and Ohura, N., Jr., (2008). Influence of external forces
(pressure and shear force) on superficial layer and subcutis of porcine skin and
effects of dressing materials: are dressing materials beneficial for reducing pressure
and shear force in tissues? Wound.Repair Regen., 102-107, 16(1), 102-107.
[101] Pailler-Mattei, C., Pavan, S., Vargiolu, R., Pirot, F., Falson, F., and Zahouani, H.,
(2007). Contribution of stratum corneum in determining bio-tribological properties
of the human skin. Wear, 1038-1043, 263(7-12 SPEC. ISS.), 1038-1043.
[102] Hendley, A., Marks, R., and Payne, P. A., (1982). Measurement of forces for point
indentation of the stratum corneum in vivo: The influences of age, sex, site
delipidisation and hydration. Bioeng Skin, 234-240, 3234-240.
References 113
[103] Pailler-Mattei, C. and Zahouani, H., (2004). Study of adhesion forces and
mechanical properties of human skin in vivo. Journal of Adhesion Science and
Technology, 1739-1758, 18(15-16), 1739-1758.
[104] Potter, A., Luengo, G., Baltenneck, C., Pavan, S., and Loubet, JL,(1-7-2007).
Measuring mechanical properties of stratum corneum and isolated corneocytes at
sub-micron length scale.
[105] Plewig, G. and Marples, R. R., (1970). Regional differences of cell sizes in the
human stratum corneum. I. J.Invest Dermatol., 13-18, 54(1), 13-18.
[106] Plewig, G., (1970). Regional differences of cell sizes in the human stratum
corneum. II. Effects of sex and age. J.Invest Dermatol., 19-23, 54(1), 19-23.
[107] Yuan, Y. and Verma, R., (2006). Measuring microelastic properties of stratum
corneum. Colloids and Surfaces B: Biointerfaces, 6-12, 48(1), 6-12.
[108] Oliver, W. C. and Pharr, G. M., (1992). An Improved Technique for Determining
Hardness and Elastic-Modulus Using Load and Displacement Sensing Indentation
Experiments. Journal of Materials Research, 1564-1583, 7(6), 1564-1583.
[109] Pelletier, C. G. N., Den Toonder, J. M. J., Govaert, L. E., Hakiri, N., and Sakai,
M., (2008). Quantitative assessment and prediction of contact area development
during spherical tip indentation of glassy polymers. Philosophical Magazine, 12911306, 88(9), 1291-1306.
[110] Ebenstein, D. M. and Pruitt, L. A., (2006). Nanoindentation of biological
materials. Nano Today, 26-33, 1(3), 26-33.
[111] Jachowicz, J., McMullen, R., and Prettypaul, D., (2007). Indentometric analysis of
in vivo skin and comparison with artificial skin models. Skin Research and
Technology, 299-309, 13(3), 299-309.
[112] Boyer, G., LaquiFze, L., Le Bot, A., LaquiFze, S., and Zahouani, H., (2009).
Dynamic indentation on human skin in vivo: ageing effects. Skin research and
technology : official journal of International Society for Bioengineering and the
Skin (ISBS) [and] International Society for Digital Imaging of Skin (ISDIS) [and]
International Society for Skin Imaging (ISSI), 55-67, 15(1), 55-67.
[113] Iatridis, J. C., Wu, J., Yandow, J. A., and Langevin, H. M., (2003). Subcutaneous
tissue mechanical behavior is linear and viscoelastic under uniaxial tension.
Connect.Tissue Res., 208-217, 44(5), 208-217.
[114] Haberman, E. and Gefen, A.,(2007). Viscoelastic properties of ovine adipose
tissue covering the gluteus muscles as related to pressure ulcer modeling.
114 References
[115] Hendriks, F. M., Brokken, D., van Eemeren, J. T. W. M., Oomens, C. W. J.,
Baaijens, F. P. T., and Horsten, J. B. A. M., (2003). A numerical-experimental
method to characterize the non-linear mechanical behaviour of human skin. Skin
Research and Technology, 274-283, 9(3), 274-283.
[116] Smith, S. R., Lovejoy, J. C., Greenway, F., Ryan, D., deJonge, L., de la Bretonne,
J., Volafova, J., and Bray, G. A., (2001). Contributions of total body fat, abdominal
subcutaneous adipose tissue compartments, and visceral adipose tissue to the
metabolic complications of obesity. Metabolism-Clinical and Experimental, 425435, 50(4), 425-435.
[117] Bray, G. A., Lovejoy, J. C., Most-Windhauser, M., Smith, S. R., Volaufova, J.,
Denkins, Y., de Jonge, L., Rood, J., Lefevre, M., Eldridge, A. L., and Peters, J. C.,
(2002). A 9-mo randomized clinical trial comparing fat-substituted and fat-reduced
diets in healthy obese men: the Ole Study. Am.J.Clin.Nutr., 928-934, 76(5), 928934.
[118] Lovejoy, J. C., Smith, S. R., and Rood, J. C., (2001). Comparison of regional fat
distribution and health risk factors in middle-aged white and African American
women: The Healthy Transitions Study. Obes.Res., 10-16, 9(1), 10-16.
[119] Krouskop, T. A., Wheeler, T. M., Kallel, F., Garra, B. S., and Hall, T., (1998).
Elastic moduli of breast and prostate tissues under compression. Ultrasonic
Imaging, 260-274, 20(4), 260-274.
[120] sarvazyan, A. P., Skovoroda, A., and Emelianov, S, (1995). Biophysical bases of
elasticity imaging. Acoustic Imaging, 223-240, 21223-240.
[121] Linder-Ganz, E., Shabshin, Noga, Itzchak, Yacov, and Gefen, Amit, (2007).
Assessment of mechanical conditions in sub-dermal tissues during sitting: A
combined experimental-MRI and finite element approach. Journal of
Biomechanics, 1443-1454, 40(7), 1443-1454.
[122] Klein, J., Permana, P. A., Owecki, M., Chaldakov, G. N., Bohm, M., Hausman,
G., Lapiere, C. M., Atanassova, P., Sowinski, J., Fasshauer, M., Hausman, D. B.,
Maquoi, E., Tonchev, A. B., Peneva, V. N., Vlachanov, K. P., Fiore, M., Aloe, L.,
Slominski, A., Reardon, C. L., Ryan, T. J., and Pond, C. M., (2007). What are
subcutaneous adipocytes really good for ...? Experimental Dermatology, 45-70,
16(1), 45-70.
[123] Beynen, A. C. and Katan, M. B., (1985). Rapid Sampling and Long-Term Storage
of Subcutaneous Adipose-Tissue Biopsies for Determination of Fatty-Acid
Composition. American Journal of Clinical Nutrition, 317-322, 42(2), 317-322.
References 115
[124] Tirrel, M., (1994). Rheology of Polymeric Liquids. In: Macosko, C. W.,
Rheology: principles, measurements, and applications. Wiley-VCH, Inc, New york.
[125] Hrapko, M., Dommelen, J. A. W. v., Peters, G. W. M., and Wismans, J. S. H.
M.,(2005). The mechanical behaviour of brain tissue: large strain response and
constitutive modeling. 59-69. 59-69.
[126] van Turnhout, M., Peters, G., Stekelenburg, A., and Oomens, C., (2005). Passive
transverse mechanical properties as a function of temperature of rat skeletal muscle
in vitro. Biorheology, 193-207, 42(3), 193-207.
[127] Verbeke, G. and Molenberghs, G. (2000). Linear Mixed Models for Longitudinal
data. Springer, New York.
[128] Pipkin, A. C. (1986). Lectures on Viscoelasticity. Springer-Verlag, New York.
[129] Tanner, R. I. (2000). Engineering Rheology. Oxford University Press,
[130] Geerligs, M., Peters, G. W., Ackermans, P. A., Oomens, C. W., and Baaijens, F.
P., (2008). Linear viscoelastic behavior of subcutaneous adipose tissue.
Biorheology, 677-688, 45(6), 677-688.
[131] Darling, E. M., Topel, M., Zauscher, S., Vail, T. P., and Guilak, F., (2008).
Viscoelastic properties of human mesenchymally-derived stem cells and primary
osteoblasts, chondrocytes, and adipocytes. J.Biomech., 454-464, 41(2), 454-464.
[132] Then, C., Menger, J., Benderoth, G., Alizadeh, M., Vogl, T. J., Hubner, F., and
Silber, G., (2007). A method for a mechanical characterisation of human gluteal
tissue. Technol.Health Care, 385-398, 15(6), 385-398.
[133] van Dam, E. A., Dams, S. D., Peters, G. W., Rutten, M. C., Schurink, G. W., Buth,
J., and van de Vosse, F. N., (2008). Non-linear viscoelastic behavior of abdominal
aortic aneurysm thrombus. Biomech.Model.Mechanobiol., 127-137, 7(2), 127-137.
[134] Mewis, J. and Wagner, N. J., (2009). Thixotropy. Advances in Colloid and
Interface Science, 214-227, 147-48214-227.
[135] Mewis, J. and Wagner, Norman J.Thixotropy. Advances in Colloid and Interface
Science, In Press, Corrected Proof
[136] O'Neill, P. L. and Stachowiak, G. W., (1996). The inverse thixotropic behaviour
of synovial fluid and its relation to the lubrication of synovial joints. Journal of
Orthopaedic Rheumatology, 222-228, 9(4), 222-228.
116 References
[137] Linder-Ganz, E. and Gefen, A., (2009). Stress analyses coupled with damage laws
to determine biomechanical risk factors for deep tissue injury during sitting.
J.Biomech.Eng, 011003, 131(1), 011003.
[138] Stekelenburg, A., Gawlitta, D., Bader, D. L., and Oomens, C. W., (2008). Deep
tissue injury: how deep is our understanding? Arch.Phys.Med.Rehabil., 1410-1413,
89(7), 1410-1413.
[139] Avram, M. M., Avram, A. S., and James, W. D., (2005). Subcutaneous fat in
normal and diseased states: 1. Introduction. Journal of the American Academy of
Dermatology, 663-670, 53(4), 663-670.
[140] Palero, J. A., de Bruijn, H. S., van der Ploeg van den Heuvel, Sterenborg, H. J.,
and Gerritsen, H. C., (1-8-2007). Spectrally resolved multiphoton imaging of in
vivo and excised mouse skin tissues. Biophys.J., 992-1007, 93(3), 992-1007.
[141] OECD,(13-4-2009). OECD Guideline for the testing of chemicals. Skin
absorption: in vitro Method. Test Guideline 428(
[142] Garo, A., Hrapko, M., van Dommelen, J. A., and Peters, G. W., (2007). Towards a
reliable characterisation of the mechanical behaviour of brain tissue: The effects of
post-mortem time and sample preparation. Biorheology, 51-58, 44(1), 51-58.
[143] Cox, M. A. J., Driessen, N. J. B., Boerboorn, R. A., Bouten, C. V. C., and
Baaijens, F. P. T., (2008). Mechanical characterization of anisotropic planar
biological soft tissues using finite indentation: Experimental feasibility. Journal of
Biomechanics, 422-429, 41(2), 422-429.
[144] Hendriks, F. M., Oomens, C. W. J., Baaijens, F. P. T., and Brokken, D., (2004).
Influence of hydration and experimental length scale on the mechanical response of
human skin in vivo, using optical coherence tomography. Skin Research and
Technology, 231-241, 10(4), 231-241.
[145] Bronneberg, D. (2007). Biochemical markers for early detection of superficial
pressure ulcers. Ph.D. thesis, Technische Universiteit Eindhoven.
[146] cornelissen, l. h. (2008). Modeling the transport of biochemical markers in skin:
towards pressure ulcer assessment. Ph.D. thesis, Technische Universiteit
Eindhoven.
[147] Gawlitta, D. (2006). Compression-induced factors influencing the damage of
engineered skeletal muscle. Ph.D. thesis,
PrFont34Bin0BinSub0Frac0Def1Margin0Margin0Jc1Indent1440Lim0Lim1Width1
WidthB3WidthA3Width3Width1409tensileWidth1WidthB3WidthA3Width3Width
1409Technische Universiteit Eindhoven.
References 117
[148] Bosboom, E. M. H. (2001). Deformation as a trigger for pressure sore related
muscle damage. Ph.D. thesis, Technische Unievrsiteit Eindhoven.
[149] Stekelenburg, A. (2005). Mechanisms associated with deep tissue injury induced
by sustained compressive loading. Ph.D. thesis, Technische Universiteit Eindhoven.
[150] Ceelen, K. K. (2008). Modeling the development of in vitro and in vivo pressureinduced muscle damage. Ph.D. thesis,
[151] Dullaert, K. and Mewis, J., (2005). A model system for thixotropy studies.
Rheologica Acta, 23-32, 45(1), 23-32.
[152] Dullaert, K. and Mewis, J., (2005). Thixotropy: Build-up and breakdown curves
during flow. Journal of Rheology, 1213-1230, 49(6), 1213-1230.