skin layer mechanics - materials technology - tu eindhoven · skin layer mechanics the human skin...

Skin layer mechanics

MARION GEERLIGS

ISBN: 978-90-74445-92-4

Cover design: Marion Geerligs & Henny Herps

Printed by Universiteitsdrukkerij TU Eindhoven, Eindhoven, The Netherlands.

©Koninklijke Philips Electronics N.V. 2009

All rights reserved. Reproduction in whole or in part is prohibited without the

written consent of the copyright owner.


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 2010 om 16.00 uur

door

Marion Geerligs

geboren te Hoogezand-Sappemeer

Dit proefschrift is goedgekeurd door de promotor:

prof.dr.ir. F.P.T. Baaijens

Copromotoren:

dr.ir. C.W.J. Oomens

en

dr.ir. G.W.M. Peters

Contents

Summary ................................................................................................ ix

Skin layer mechanics ............................................................................ ix

Chapter 1 General introduction ........................................................... 1

1.1 Introduction ............................................................................................................... 2

1.2 A mechanical view of skin anatomy and physiology ............................................... 4

1.2.1 Skin topography ................................................................................................. 4

1.2.2 Stratum corneum ................................................................................................ 5

1.2.3 Viable epidermis ................................................................................................ 6

1.2.4 Dermal-epidermal junction ................................................................................ 7

1.2.5 Dermis................................................................................................................ 8

1.2.6 Hypodermis........................................................................................................ 9

1.3 Review of 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

Chapter 2 Isolation and preservation methods for the epidermis

and stratum corneum ........................................................................... 15

2.1 Introduction ............................................................................................................. 16

2.2 Skin preparation and analyses ................................................................................ 17

2.2.1 Skin preparation ............................................................................................... 17

2.2.2 Histological examination ................................................................................. 18

2.2.3 Analyses of skin viability ................................................................................ 19

2.3 Epidermal isolation techniques ............................................................................... 19

Summary vi

2.3.1 Mechanical separation ..................................................................................... 19

2.3.2 Ionic change ..................................................................................................... 20

2.3.3 Heat .................................................................................................................. 21

2.3.4 Enzymatic digestion ........................................................................................ 21

2.3.5 Microwave irradiation ..................................................................................... 23

2.4 Isolation techniques for the stratum corneum......................................................... 23

2.4.1 Mechanical separation ..................................................................................... 24

2.4.2 Chemical separation ........................................................................................ 25

2.4.3 Enzymatic digestion ........................................................................................ 25

2.5 Preservation of the upper skin layers ...................................................................... 26

2.5.1 Short-term storage ........................................................................................... 27

2.5.2 Long-term storage ............................................................................................ 28

2.6 Discussion ............................................................................................................... 30

Chapter 3 Linear shear response of the upper skin layers .............. 33

3.1 Introduction ............................................................................................................. 34

3.2 Methods .................................................................................................................. 35

3.2.1 Sample preparation .......................................................................................... 35

3.2.2 Experimental set-up ......................................................................................... 36

3.2.3 Rheological methods ....................................................................................... 39

3.2.4 Experimental procedures ................................................................................. 40

3.3 Results ..................................................................................................................... 41

3.4 Discussion ............................................................................................................... 46

Chapter 4 A new indentation method to determine mechanical

properties of the epidermis ................................................................. 49

4.1 Introduction ............................................................................................................. 50


4.1.2 Experimental procedure ................................................................................... 53

4.1.3 Determination of the Young‟s modulus .......................................................... 54

4.2 Results ..................................................................................................................... 55

4.3 Discussion ............................................................................................................... 56

Chapter 5 Linear viscoelastic behavior of subcutaneous adipose

tissue ...................................................................................................... 61

5.1 Introduction ............................................................................................................. 62

5.2 Methods and Materials ........................................................................................... 64



5.2.3 Testing procedure ............................................................................................ 65

5.2.4 Statistics ........................................................................................................... 66

5.3 Results ..................................................................................................................... 67

vii

5.3.1 Small oscillatory strain behavior ..................................................................... 67

5.3.2 Model application ............................................................................................ 68

5.3.3 Time-Temperature Superposition .................................................................... 69

5.3.4 Freezing effects ................................................................................................ 70

5.4 Discussion ............................................................................................................... 71

Chapter 6 Does subcutaneous adipose tissue behave as an

(anti-)thyxotropic material? ................................................................ 73

6.1 Introduction ............................................................................................................. 74

6.2 Materials & Methods .............................................................................................. 75



6.3 Results ..................................................................................................................... 78

6.3.1 Long term small strain behavior ...................................................................... 78

6.3.2 Large strain experiments ................................................................................. 79

6.4 Discussion ............................................................................................................... 81

Chapter 7 General discussion ............................................................. 85

7.1 Introductory remarks .............................................................................................. 86

7.2 In vitro model ......................................................................................................... 87

7.3 Mechanical methods ............................................................................................... 88

7.4 Main findings .......................................................................................................... 90

7.4.1 Small strain behavior of the epidermal layers ................................................. 90

7.4.2 Mechanical behavior of the subcutaneous adipose tissue ............................... 91

7.5 Implications for clinical and cosmetic applications ............................................... 91

7.6 Recommendations................................................................................................... 92

7.7 General conclusion ................................................................................................. 94

Samenvatting ........................................................................................ 95

Dankwoord ............................................................................................ 97

Curriculum Vitae ................................................................................. 99

References ........................................................................................... 100

Summary


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 was 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, the viable

epidermis and stratum corneum, and the deeper lying hypodermis. Studying these layers

in an in vivo set up is very 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 non-invasive 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 aim 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.

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

the right experimental conditions to ensure reliable results. Subsequently, micro-

indentation experiments were applied using a spherical tip with a relatively large

Summary x

diameter. The Young‟s moduli were 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. If the skin is compressed with an indenter, 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 take

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 at small strains 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 becomes stationary. These

results showing time 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 the 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 mechanical properties and

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 stiffness to resist 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, trauma, medical or cosmetic treatments. In order to understand

the skin behavior following the onset of these conditions, 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 into the underlying dermal layer or for

diseases such as pressure ulcers, the combined mechanical response of all individual skin

layers is important. Although often not recognized, this is also the case during the

removal of skin adhesives or the use of consumer products such as shavers. For all these

applications, the subcutaneous fat layer 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. These include the development of artificial outer skin to substitute

animal and clinical testing in evaluating drugs, cosmetics and other consumer products,

and engineered fatty tissue facilitates large volume soft tissue augmentation in plastic

surgery. Furthermore, the mechanical behavior of subcutaneous fat is critical for many

other clinical treatments beyond the scope of this thesis, such as liposuction surgery and

cellulite treatments.

To date, research on skin mechanics has mainly focused on full-thickness skin, the mid-

layer (dermis) and the top layer of the epidermis, the stratum corneum. 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. Indeed

very limited experimental data is available for those layers. In addition, there is no

consistency in data for the stratum corneum. Accordingly, the mechanical behavior of

individual skin layers could not have been yet incorporated in numerical models. This

thesis therefore focuses on the mechanical characterization of stratum corneum,

epidermis and the subcutaneous adipose tissue. Before the scope and outline of the thesis

is given, the anatomy of the skin and skin layer mechanics is shortly discussed.

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

1.2 A mechanical view of skin anatomy and physiology

Mechanical properties of skin vary considerably and depend on body site, age, race and

gender. Individual factors like exposure to UV irradiation, the use of creams and

individual health and nutritional status can also affect 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 considered to be very stiff 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 can be considered to be minimal in comparison to the bulk mechanical

behavior caused by the main components of the skin layer.

As this thesis focuses on the mechanical behavior of the other layers, i.e. stratum

corneum, viable epidermis and hypodermis, the anatomy and physiology of these skin

layers are of particular interest.

Figure 1.2 Schematic representation of the different skin layers.

1.2.1 Skin topography

The topography 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, and follow at least two

directions. The follicular orifices are located at the junction of the furrows, whereas the

sweat pores are mainly found in 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.

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. The 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 range 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, and viable epidermis, is continuously renewed within 6 to 30 days

[8]. 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. 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.

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 [9].

6 Chapter 1

(a) (b)

Figure 1.3: Morphology of the stratum corneum. (a) schematic drawing (b) cryostat

section of normal human skin treated with Sorensen’s alkaline buffer and methylene blue.

Obtained from Marks [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 applications of either pure water, moisturizers or emollients alters the

hydration state of the stratum corneum, significantly modifying some of its mechanical

properties. Under normal conditions, the hydration in the stratum corneum conditions

varies from 5-10% near the surface up to 30% near to the transition with the viable

epidermis. Bound water associated with proteins and lipids accounts for 20-30% of the

total water volume. The total water content varies little between 30% and 60% RH,

although it increases considerably at higher values [11]. When fully hydrated, the

stratum corneum swells to twice its normal thickness. In an in vitro situation, however,

the stratum corneum can increase up to 400% of its original thickness [12]. This

highlights the constraints imposed on the stratum corneum in vivo.

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 upwards to the skin surface

where they become non-viable. Other cell types within the viable epidermis include

melanocytes, Langerhans cells and Merkel cells.

Keratinocytes change their shape, size and physical properties when migrating to the

skin surface. Indeed the morhology 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. In the next layer, the stratum spinosum, the

cells are larger and polyhedral in nature and are connected by desmosomes, which are

symmetrical laminated structures. The keratinocytes adopt a more flattened morphology

at higher layers of the stratum spinosum. In this layer, they are associated with lamellar

granules, which 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 is apparent and the cytoplasm of the flattened cells become

increasingly filled with keratohyalin masses and filaments. Furthermore, the cell

membrane becomes gradually thicker.

The thickness of the viable epidermis varies roughly between 30-100 μm [13],

accomodating between 5 to 10 cell layers. The cells are communicating by very strong

desmosomes in the very compact tissue; the intercellular spaces occupy less than 2% of

the volume [5,14]. Therefore, the mechanical integrity of the viable epidermis is

considered to be stronger than other soft tissues.

Because of its non-vascular structure, the epidermal cells are nourished from plasma that

originates in the dermal blood vessels such that the nutrients transport across the

epidermal-dermal junction.

Figure 1.4: Morphology 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. [15].

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

Figure 1.5). The degree of attachment is enhanced by parts of the epidermis penetrating

the papillary dermis resulting in large cones, called rete ridges or papillae [16]. The

major point of weakness is considered to be the lamina lucida [17]. The dermal-

epidermal junction length over a straight line ranges from 1.1 to 1.3 units [5].

stratum

corneum

basal

layer

granulous

layer

spinous

layer

N D

KF

LG

8 Chapter 1

Figure 1.5: Ultrastructure of the dermal-epidermal junction.

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 relatively small and loose

distribution of elastic and collagen fibrils within a significant amount of ground

substance. 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. In the reticular dermis, fibers are horizontally oriented.

The dermis has a mainly mechanical function. The reticular dermis is able to extend up

to about 25% by stretching 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 amorphous ground

substance acts as a viscous gel-like material, which does not leak out of the dermis, even

under high pressure. The permanent tension in the reticular dermis generates the folding

of the overlying structures and hence, the skin surface. The fiber network in the papillary

dermis contributes to the protection of vessels and cells against mechanical insults.

In the papillary dermis, 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 appears relatively 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

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, the

white adipocytes, which are held in a framework of collagen fibers. However, only one

third of adipose tissue contains mature adipocytes [18], with the remainder being

stromal-vascular cells including fibroblasts, leukocytes, macrophages, and pre-

adipocytes [19]. Adipose tissue has little extracellular matrix compared to other

connective tissues.

Stored fat is the predominant component of the adipocytes; where the lipid droplet can

exceed 50 μm. The cytoplasm and nucleus appears as a thin rim at the periphery of the

cell (Figure 1.6). The diameter of the entire white adipocyte is variable, ranging between

30 and 70 μm [18]. 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, which provides the exchange of metabolites and

allows the adipocytes to function effectively. It is interesting to note that 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 [20].

(A) (B)

Figure 1.6: Schematic drawing (a) and histological section (b) of hypodermis, or

subcutaneous adipose tissue, showing white adipocytes (WA) with the nucleus (N) at the

periphery. The adipocytes are in contact with the blood circulation via arterioles which

branches the larger arteries (A) and veins (V).

A

V

N

WA

10 Chapter 1

The mechanical functions of the subcutaneous adipose tissue include allowing the

overlying skin to move as a whole, both horizontally and vertically, and the attenuation

and dispersion of externally applied pressure.

1.3 Review of skin layer mechanics

Measurement methods and mechanical properties of skin have been extensively

reviewed in the literature [5,21,22]. Therefore, given the focus of the present work, focus

will be limited to studies on the behavior of stratum corneum, viable epidermis and

hypodermis. More specifically, they include force-elongation data, either in vivo or in

vitro, and currently available constitutive models.

1.3.1 In vivo vs in vitro experiments

When measurements on skin mechanics are performed in vivo, the human skin exists in

its natural pre-stress and skin relief. The number of in vivo measurement methods is,

however, limited [22] and a numerical-experimental approach is usually adopted. 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 of the

latter is that all forms 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 skin types, and the variety of

body sites can be problematic.

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. Full extension of the stratum corneum occurs in critical, extra-

physiological situations due to disease, trauma, clinical or cosmetic applications.

1.3.2 Mechanical behavior of the stratum corneum

Force-elongation curves at constant elongation rate demonstrate one, two or three phases

depending on the hydration level in the in vitro experiment (Figure 1.7) [23]. The first

phase, up to a 10% extension, is considered to represent purely elastic behavior. The next

phase, absent at low RH, is an irreversible elongation with a low slope, with strains

ranging from 20-125%. In addition, fully hydrated stratum corneum exhibiting a final

phase, where strain hardening is observed before rupture, at approximately 200%

extension. The slope becomes steeper at increasing elongation rates, as would be

predicted of a viscoelastic response. 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 have reported tensile testing [8,23-27]. Subsequently,

torsional techniques were developed to measure the stratum corneum behavior in vivo

[28-31]. More recently, indentation techniques were introduced to determine the

Young‟s modulus in vitro [32,33], and also in vivo indentation tests have been


performed [34]. Furthermore, imaging techniques such as ultrasound and magnetic

resonance elastography have been used to estimate mechanical properties [22,33].

Reported Young‟s moduli vary considerably encompassing values from a few MPa to

GPa [24,25,35,36]. For example, the estimated tensile moduli for various RH is shown in

Figure 1.8. As indicated in this figure, the stiffness of the stratum corneum varies from

rubber-like at high RH to nylon-like at low RH values. The differences may be due to a

combination of reasons, such as regional differences, anisotropy, differences between

species, but also test conditions, such as sample preparation and difficulties in

controlling sample dimensions and 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. The influence of temperature decreases to a minimum beyond 90% RH.

More common trends due to an increase in RH or temperature include an increase of the

maximum extension and work of rupture, and a reduction of the force at rupture

[23,25,37]. Furthermore, stratum corneum behaves isotropically in transversal plane only

[36].

Preconditioning effects have not been reported for stratum corneum, which represents an

important difference with the whole skin. This finding indicates the absence of mobile

components in the stratum corneum [36].

Current constitutive models of the stratum corneum are based on traction, relaxation and

creep tests [5]. From the experimental tests, it is important that the model accomodate

elasticity, non-linear viscosity and strain hardening parameters. However, the association

Figure 1.7 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 [23].

10

20

30

40

I II III

98% RH

76% RH

30 120

32% RH

Elongation [%]

Loa

d [g

]

in v

ivo

ra

ng

e

12 Chapter 1

between the defined parameters and the anatomical components has yet to be

determined.

Figure 1.8 An overview of Young’s moduli of stratum corneum as function of the RH

derived from in vitro tensile tests.

1.3.3 Mechanical behavior of the viable epidermis

Only recently, a few studies have focused on the viable epidermis. From an indentation

approach, a local Young‟s modulus of a few MPa has been reported for the viable

epidermis of murine ear skin [38,39]. However, it is recognised that murine skin exhibits

a higher density of hair follicles and a very thin epidermis compared with human skin.

Indeed a combined experimental-numerical approach on in vivo human skin yielded an

estimated Young‟s modulus of about 0.5 kPa for the upper human skin layers including

the papillar dermis [1,39]. The authors hypothesized that this low value was due to the

negligble influence of the stratum corneum on the overall mechanical response of the

skin, when suction was performed with small aperture sizes. Due to the dearth of

experimental data, 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 subjected to applying shear [40], compression [40,41],

indentation [42,43] or suction [1,39-41,44]. Young‟s moduli varied from a few kPa to

values in excess of 100 kPa.

All studies provide limited descriptions of the overall mechanical behavior as they were

developed for very specific applications. Consequently, an appropriate constitutive

model based on experimental data is not available yet. Indeed current models are either

limited to small strain behavior [39,45] or based on other soft connective tissues.


1.4 Aim and Outline

The objective of this thesis is to develop appropriate experimental techniques and

procedures, which will enable the characterization of the mechanical behavior of

individual skin layers in vitro. The focus is on those skin layers for which available data

is relatively scarce, i.e. the viable epidermis and hypodermis, and/or inconsistent as in

the case for the stratum corneum. The results should provide insight into the relationship

between the mechanical responses to the structure of the various skin layers and, hence,

provide better understanding of the way a treatment or disease affects the skin behavior.

Furthermore, the experimental data should provide suitable 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

are not solely caused by variations in humidity and temperature, but are influenced test

conditions, anisotropy, sample preparation, etc. It is therefore essential to perform

experiments with samples of consistent quality in an accurate measurement system in a

well-controlled environment. This will be initially achieved in relatively simple small

strain experiments in various directions under different environmental conditions. If this

small strain behavior is reproducible and well-understood, then it is appropriate to extend

the work to examine 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 system has been designed to measure the shear

response of thin, soft tissues in a controlled environment (Chapter 3). A micro-

indentation 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 predict its behavior from

the experiments on stratum corneum and whole epidermis.

Subsequently, rheological methods are developed to study the linear shear response of

subcutaneous adipose tissue (Chapter 5). From those results, a constitutive model

describing the linear viscoelastic behavior of subcutaneous adipose tissue at small strains

has been developed. Then, a set of experiments were designed to study both the large

deformation and time-dependent behavior (Chapter 6).

Finally (Chapter 7), a general discussion evaluates the selected measurement methods

for the skin layers and these outcomes, as well as the significance of the findings of this

work for various applications.

Chapter 2

Isolation and preservation methods for

the epidermis and stratum corneum

The contents of this chapter are based on M. Geerligs, D. Bronneberg, P.A.J.

Ackermans, C.W.J. Oomens, and D.L. Bader, Isolation and preservation methods for the

epidermis, submitted.

16 Chapter 2

2.1 Introduction

Ex vivo human skin grafts provide a cost-effective alternative to animal and clinical

testing. Various industries, such as the cosmetic, household product and pharmaceutical,

could benefit from in vitro studies to evaluate drugs and a range of 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 the fundamental knowledge on both structural and 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 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 well-controlled in

vitro experiments. In order to obtain meaningful 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 excess 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, its complete isolation while maintaining its

structural integrity remains a challenge. The keratinocytes are surrounded by a poor

extracellular matrix and lack the support of a fiber structure, which provides the strength

and stiffness of most biological tissues. 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 periphery of the keratinocytes. At the epidermal-

dermal junction hemidesmosomes anchor the epidermis to the dermis (see Figure 1.5).

These hemidesmosomes or the adjacent anchoring filaments need to be disrupted to fully

separate the epidermis from the dermis.

In order to maintain the complex structure of the stratum corneum during isolation, it is

important to preserve the curvature. The architecture of the stratum corneum is widely

established 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, there is an increased availability of ex vivo human

skin. 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 donor body site. Furthermore, within

one skin graft, its structure might change as a result of disease or prior treatment. 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 epidermal components

Isolation and preservation methods for the epidermis and stratum corneum 17

[46]. 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 with optimal efficiency, factors such as

cleaning, preservation, and storage should be adequately addressed. In various studies,

such as transdermal drug delivery, percutaneous absorption studies, irritancy and

toxicology studies, an intact skin barrier is essential. Furthermore, adequate 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 have been extensively

described [47-49]. 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.

This paper 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. Existing reviews are considered to be out of date

and do not include recent work from the host laboratory [46,50-52]. No standards exist,

thus inter-study comparisons are problematic. In addition, much of the existing data may

have been influenced by the specific preparation technique, which have been employed.

Accordingly, the present paper describes mechanical, ionic change, heat, enzymatic

digestion and irradiation 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 in the present experiments are

described below, as well as the analysis techniques used to study the skin structure and

viability.

2.2.1 Skin preparation

Human skin was 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 (Figure 2.1a). Subsequently, split-thickness skin samples, varying in

thickness from 100-400 µm, are produced using a commercial dermatome (D42,

Humeca, The Netherlands) (Figure 2.1b).

18 Chapter 2

(a) (b)

Figure 2.1: Skin is stretched using forceps (a) and dermatomed (b).

(b)

(a) (c)

Figure 2.2. (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.

2.2.2 Histological examination

In order to examine tissue structure, 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, namely the stratum corneum, viable epidermis, papillar dermis and reticular

dermis (Figure 2.2a). The structural integrity is examined by using the H&E staining.

SC

VE

PD

RD

SC

VE

SC

VE

PD


2.2.3 Analyses of skin viability

Skin viability was studied by using the 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 of 1 mg/ml MTT solution in PBS in a well (Phosphate Buffered

Saline). The plates were incubated at 37C 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 for 2 hours 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. The

effectiveness of each is summarized in Table 2.1 at the end of the section in terms of

actual cleavage plane, maintaining of both cell viability and tissue integrity.

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. The method involves manually stretching the skin to its limit over a

slightly convex wooden surface, and anchoring it 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 can be easily

induced in the epidermis using this relatively crude 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 the present study, a cordless, battery operated dermatome was used. As previously

mentioned, 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 was still attached to the

epidermal specimens (Figure 2.2). Due to the presence of rete ridges, it was highly

unlikely that the cutting plane went through the dermal-epidermal junction only.

However, the number of skin layers present in the separated tissue can be assessed

visually; with the yellowish translucent epidermis being easily distinguishable from the

white opaque dermis. A MTT-test demonstrated that the dermatomed skin retained its

viability for 100%, which is in agreement with Wester et al. [54].

The defined geometric shape of the specimen is very convenient for assessing its

mechanical properties. It is assumed that the mechanical properties of the present

papillary dermis are similar to the surrounding epidermal tissue, because no differences

20 Chapter 2

in shear properties were found between 100 and 200 μm thick split-skin samples (see

Chapter 3).

Suction device

Suction blisters can be produced by applying suction cups on the skin, in both in vivo

and in vitro experiments. In vivo separation of the human epidermis was first reported in

1964 [55]. Kiistala et al.(1968) found that a blister could be induced within 130 minutes

with a suction gap of 25 mm. 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 had to be

maintained at 200 mm Hg or above. The cleavage occurs in the plane through the lamina

lucida, leaving the lamina densa 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 [50,56].

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

An earlier method to isolate the epidermis involved its maceration in dilute acetic acid.

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 in an alkaline environment. These

methods, however, are toxic to 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 changes according to the duration of the treatment.

For example, after 30 min incubation in 0.01 M EDTA at pH 7.4 the split occurred in the

lower granular layer, whereas after 45 min it was in a spinous-suprabasilar location and

after 60 min or more it occurred at the dermal–epidermal junction. In adition,

intracellular oedema increases with time. Accordingly, this is not considered to be a

favourable method for epidermal separation.

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 within the keratinocytes was noted [50]. Although

no other degenerative features have been reported, epidermal components may have been

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.

Indeed after 72-96 hours 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 any of the techniques

based on ionic change, they are not considered 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]. It was reported that the skin is heated up to 50 to 60C for 30 s. To maintain

enzyme activity, mild heat treatment at 52C 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]. Although heating can easily cause tissue

dehydration, this can be minimized 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 the present studies, human skin samples were heated on either a hot plate and in a

sealed bag. The former process appeared to flatten the undulating epidermal structure,

while the papillae remained intact after heating in a sealed bag in hot water. Much longer

heating times were needed than mentioned in literature. The epidermis could be peeled

from the dermis after more than 5 minutes.

For both heat separation techniques, structural tissue damage occured as evidenced by

the presence of vacuoles and a disrupted basal layer (Figure 2.3). It has been previously

reported that heat treated skin (60°C for 1 minute) and heat-separated epidermis and

dermis significantly lose viability [63]. 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.

(a) (b)

Figure 2.3. Histological 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. For example, 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%

22 Chapter 2

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 [46]. 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 noted

following trypsin treatment such that recovery may take up to a few days [46]. All these

factors lead to inconsistent epidermal separation following treatment.

Thermolysin

The epidermis can easily be separated from the dermis following incubation at 4C for 1

h in a solution containing 250-500 g/ml thermolysin, a proteolytic enzyme more

generally used for protein analysis [65]. Thermolysin can be dissolved in sterile

magnesium free PBS containing 1 mM CaCl2 at pH 7.8. However, to ensure complete

penetration of the enzyme, it is advisable to remove the subcutaneous fat and the lower

dermis from the specimen. Light and electron microscopy revealed that the separation

occurred at the lamina lucida and that the hemidesmosomes were selectively disrupted

[65]. By contrast, Willsteed et al.[50] noticed an intraepidermal split, without any lamina

lucida separation.

Dispase

Dispase II (Roche Diagnostics) has proven to be a rapid, effective, but gentle agent for

separating intact epidermis from the dermis [67,68]. This proteolytic enzyme is able to

cleave the basement membrane zone region while preserving the viability of the

epithelial cells.

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 the present studies to separate the

epidermis from the dermis. Fresh skin samples of various sizes were placed on top of

sterile gauzes in 6 cm diameter petri dishes 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 minimize loss of the skin barrier integrity. After overnight

incubation at 4C and thereafter 10 min at 37C, the epidermis was gently peeled from

the dermis using tweezers. In agreement with literature, the present study 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 2.4). The cleavage occurred in the lamina densa.

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, which were

then enough to facilitate enzyme diffusion. Therefore, it is recommended to dermatome

skin grafts prior to performing the enzyme treatment.


Figure 2.4. H&E staining of epidermis separated with Dispase.

2.3.5 Microwave irradiation

Sanchez et al. [69] 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 reported to be essential

to use the prescribed buffer and specifically adhere to the recommended microwave

exposure times. Nevertheless, microwave irradiation seems to be a rapid method for

separation of the epidermis from the dermis.

Table 2.1: Critical of isolation techniques used for epidermal tissues. Techniques that are

highlighted, are investigated in our laboratories.

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. The effectiveness

Type Method

Treatment

duration Cleavage plane

Tissue

integrity

Tissue

viability Reproducibility

Mechanical Dermatome < 1 hr variable + + +

Suction < 2hrs lamina lucida 0 0 -

Heat 5 min basal layer - - 0

Ionic NaCl 24-96 hrs lamina lucida 0 n.a.* 0

change EDTA > 1 hr n.a. - n.a.* -

PBS 72-96 hrs hemidesmosomes - 0 0

Enzymatic Trypsin 1-24 hr variable - 0 -

digestion Thermolysin 1 hr hemidesmosomes + + n.a.*

Dispase 24 hrs lamina densa + + +

Irradiation Microwave 5 min hemidesmosomes 0 n.a.* +

*n.a. = not available

24 Chapter 2

of each technique is summarized in Table 2.2, in terms of maintaining both cell viability

and tissue integrity.

2.4.1 Mechanical separation

Stratum corneum separating by cutting techniques is complicated due to the inherent

curvature of the skin. However, the thickness of the stratum corneum has little variation,

such that flattening of the skin might improve mechanical separation. It has already been

shown that the skin relief dramatically decreases when a microscope slide is placed on

top of it [70]. In the present study, topography measurements were performed on

unloaded and loaded skin with 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 slide was placed on the

upper surface of the specimen and pushed down with two weights of 100 g on each side.

Again the roughness parameters were determined. Preliminary testing showed that the

microscopic slide on top was not detected by the system and did not influence the

measurement output. A significant decrease in skin surface roughness was measured,

with a mean value of 42 μm in a loaded configuration compared with 85 μm in the

unloaded state. The latter is comparable to what can be found in literature [5].

Nonetheless, the surface roughness in the loaded state was still at least three times the

thickness of the stratum corneum.

Following the topography measurement, the sample was maintained between two plates

and stored at -80°C. In order to retain the flattened state of the skin sample, the sample

was cut using a cryotome. The surface of the stratum corneum was aligned with the

cutting system to obtain the stratum corneum using a single cut with a thickness of 20

μm. The stratum corneum sheets have some other epidermal strata attached and cavities

(Figure 2.5).

(a) (b)

Figure 2.5. 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 [8,23].

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 [71,72]. Adherent wet cells are subsequently removed with a

cotton swab such that the stratum corneum sheet remains [73]. 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.

2.4.3 Enzymatic digestion

Trypsin

The working of trypsin throughout the epidermal strata has been extensively studied

[73]. 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 [73]. 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.

26 Chapter 2

Figure 2.6. (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.

Techniques that are highlighted, are investigated in our laboratories.

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 [74], 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 [54,74,74-

76], 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

Type Method

Treatment

duration

Tissue

integrity

Tissue

viability Reproducibility

Mechanical Cutting

(cryotome)24 hrs 0 - -

Cantharidin 4.5 hrs - - -

Ionic change Ammonia 45 min - - 0

Enzymatic digestion Trypsin 2-24 hrs + + +

(a) (b)


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

[48]. 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 [77]. 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. [78] 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

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 [76]. Castagnoli et al. [79]

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.

28 Chapter 2

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 [80].

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

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 [74].

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 [78,81]. 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 [74].

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. [82] showed that the mechanical properties of human skin are not affected by

freezing as well, it might be sufficient to snapfreeze samples for mechanical

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. [83], 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.

30 Chapter 2

Table 2.3: Overview of the ease of handle and success rate of various isolation techniques.

Techniques that are highlighted, are investigated in our laboratories.

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 (Table 2.1 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.

The overview in Table 2.1 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.

Type Method

Treatment

duration

Storage

time

Tissue

integrity

Tissue

viability

epid

erm

is

short-term saline solutions none days - 0

nutrient media none weeks + +

long-term cryopreservatio

n < 1 hr years + +

snap freezing 10 s 3 months + -

SC short-term PBS none 3-5 days - +

long-term drying few days years + +


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

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,

32 Chapter 2

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.

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 aldehyde-

fuchsin 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

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, Linear shear response of the upper

skin layers, submitted.

34 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, inhomogeneous, non-linear and viscoelastic. Therefore, the most appropriate

approach seems to be to determine the mechanical properties of each individual skin

layer in 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 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 ulcers, the removal of skin

adhesives and skin-device contact such as with prosthetic limbs and shavers. All these

applications could benefit from an improved knowledge of the mechanical response of

the epidermis to shear.

As the epidermis provides the chemical and physical barrier between the human body

and its environment, it possesses extraordinary structural properties. It 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 [84]. 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 „brick-

and-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 30-

100 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 the temperature, 𝑇, and relative humidity, 𝑅𝐻.

Usually, load-bearing soft tissues are composed of a fiber network, providing strength

and elasticity to the tissue, but this is not the case for the epidermis. Its extensibility is

mainly due to the potential for smoothing 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 [5,14].

Consequently, the viable epidermis is suspected to be more rigid than other soft tissues.

In the stratum corneum, the cellular membranes are thickened, the 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.

Linear shear response of the upper skin layers 35

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 issue, the 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 offer a broad range of reliable

standard techniques used in mechanical testing. 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 accomodated.

Currently, there is a paucity of papers describing the mechanical properties of the entire

epidermis or viable epidermis only. Studies to date were either on a small-sized scale

[38,85], not reproducible [86] or included the total papillar dermis [39] and none of them

investigated the shear response. Mechanical properties of the stratum corneum have been

studied and reviewed more extensively [5,22,35,87], although very few have examined

its shear response. 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

considerably less than the broad range of tensile 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, tests with silicone rubbers are also 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 an age group between 35 and 55 years old

was used. Abdominal skin with stria, cellulite, damage due to UV exposure or

excessively hairy skin was excluded from the study.

Immediately after excision, the skin is transferred to the laboratory and processed within

4 hours. Skin slices are obtained using an electric dermatome (D42, Humeca, The

Netherlands) of which the prescribed thickness was refined for this purpose by the

36 Chapter 3

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 [5,13]. Depending on various factors such as skin surface roughness, tissue

hydration, smoothness of the cutting, some papillar dermis was observed to 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, the 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 skin, reticular dermis is also

present. For isolating the reticular dermis, the top layer of skin is dermatomed until the

white opaque dermis appears 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 a maximum of 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 prior to 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 [88-90]. The shear

stress 𝜏 and shear strain 𝛾 are then calculated from the measured torque 𝑀 and the angle

𝜃 using:

𝜏 =𝑀𝑟

2𝜋𝑟12

𝑟 − 𝑟1 2

2+

𝑟12

8

, 𝛾 = 𝜃

𝑟

ℎ,

(3.1)


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.

r

r1

M

(a) (b)

(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).

VE

SC

PD

RD

38 Chapter 3

where 𝑟 is the radius of the plate, 𝑟1is the sample radius and ℎ is the sample height. The

advantages of positioning the sample at 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 straightens. Subsequently, the droplet is removed 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 is

subjected to the normal force.

Samples are measured in a controlled environment using a home-built system, see Figure

3.3. Therefore, dry and fully hydrated air are mixed to obtain the desired RH by

regulating the flow inlets. The mixing chamber, as well as the chamber to obtain fully

hydrated air, is placed in a water bath to control the temperature. Finally, the air is

transported 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.

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 chambers in the waterbath (D). In one chamber, the air is fully

hydrated. In the next chamber, the dry and fully hydrated air are mixed to obtain the

desired RH. 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.

C

D A

B

E G

F


3.2.3 Rheological methods

Linear viscoelastic material behavior is described by a multi-mode Maxwell model:

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:

𝜏 = 𝜏𝑖 𝑛𝑖=1 . (3.3)

A frequency (𝜔) dependent input 𝛾 = 𝛾0 sin 𝜔𝑡 will lead, for linear viscoelastic

behavior, to a sinusoidal shear stress:

𝜏 = 𝐺𝑑𝛾0 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:

𝐺 ′ = 𝜏0′ 𝛾0 = 𝐺𝑖

𝜆𝑖2𝜔2

1 + 𝜆𝑖2𝜔2

;

𝑛

𝑖=1

(3.6)

𝐺 ′′ = 𝜏0′′ 𝛾0 = 𝐺𝑖

𝜆𝑖𝜔

1 + 𝜆𝑖2𝜔2

𝑛

𝑖=1

, (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 . (3.8)

The phase shift 𝛿

to 𝐺′ and 𝐺" via:

tan 𝛿 =𝐺 ′′

𝐺′ . (3.9)

𝜏𝑖 +1

𝜆𝑖𝜏𝑖 = 𝐺𝑖𝛾 ; 𝑖 𝜖 1, 𝑛 , (3.2)

40 Chapter 3

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 the combination of skin layers. In order to determine the

mechanical parameters, the linear viscoelastic strain regime defined as 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 demands that

preliminary tests are essential, i.e. the optimal experimental conditions needs to be

defined to ensure reliable results.

It is recognised that the samples, particularly involving the stratum corneum, are

extremely thin (under 20 μm). Measuring such thin samples is at the limit of the

possibilities of the apparatus. Therefore, to validate the experimental approach, a well-

defined homogeneous soft material, i.e. silicone rubber, with different thicknesses was

tested. Also, an approach using a stack of layers to increase the sample thickness was

evaluated. Furthermore, the natural wrinkling 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 magnitude

of the applied normal force was determined for various numbers of stacked stratum

corneum samples. In addition, the sensitivity of the upper skin layers to its environment

needs to be translated into conditioning times, i.e. the times required for the mechanical

behavior to be stabilized to an equilibrium.

The experimental procedures to account for each of these issues are discussed in the

order given below. For each experimental procedure, the type of sample is stated.

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 at a constant diameter of 8 mm. The shear modulus is determined for various

frequencies increasing stepwise from 1 to 100 rad/s at 0.01 strain.


Stacking

A possible way to resolve the problem of thin samples with a complex wrinkled sample

geometry is to stack a few of these samples. This approach is evaluated 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 pliable than stratum corneum, it is assumed

that the space between the parallel plates is filled and that the skin surface roughness is

negligble. A normal force of 1 g is applied to 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 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 are designed to confirm 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 were 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.

3.3 Results

For all tests, the linear viscoelastic behavior is presented in terms of the shear modulus,

𝐺𝑑 , and the phase angle, 𝛿. As 𝛿 appeared to remain 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

42 Chapter 3

loss modulus 𝐺" for samples with various thicknesses are measured. It is conluded that

the experimental approach is appropriate for measuring thin, soft materials.

(a) (b)

Figure 3.4: Frequency sweeps performed on silicone rubber of various thickness: 0.050,

0.120 and 2.00 mm. (a) The average shear modulus 𝑮𝒅; (b) The average phase angle δ.

Stacking

In this test, stratum corneum samples were examined. As shown in Figure 3.5a,

increasing the force from 1 to 3 g results in large differences in 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 assumption (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 up to a value

of almost 0.01. As the conditioning time for epidermis and stratum corneum could only

be estimated during this test, the measured value of 𝐺𝑑 might differ slightly 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 correspondingly lower

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 dermis (not shown). However, histological examination showed that the

composition of those skin samples is in agreement with the predictions.


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 will also be applicable 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 slightly decreased until about 40 min. Therefore, fully

hydrated skin samples are preferably conditioned for 45 minutes. A considerable

increase in the standard deviaton for the higher RH was noted.

(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 ().

44 Chapter 3

(a) (b)

(c) (d)

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 indicates the necessary conditioning time.

Determination of G’ and G”

The dependency on RH and temperature were measured for both the epidermis and

stratum corneum. For each RH/T combination, the tests were designed to measure 3

samples per subject. However, the test sequence could not be completed for subjects 2

and 4 within 72 hr. As only three meausurements could be performed on epidermis from

subject 4, subject 4 was totally excluded from this part of the study.

For both stratum corneum and epidermis, the modulus was found to be slightly

frequency dependent (see Figure 3.8). However, 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 the small

frequency dependency, a comparison between the different environmental conditions

was performed at one frequency only. In this case, 10 rad/s was chosen (see Figure 3.9

and Figure 3.10). The results for stratum corneum show a decrease in modulus with

increasing humidity, but no apparent change with temperature. For the epidermis, data

were more variable, especially at 20°C. The mechanical parameters appeared

independent of the two temperatures.


(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.

46 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 only 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 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 optimal

experimental conditions to ensure reliable final measurements.

In order to measure shear properties of a soft biological material, measurement methods

have been developed for muscle, brain and thrombus [88-90]. However, pre-testing was

needed to prove that the experimental approach is also appropriate for samples in the

order of a few micrometers 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 higher normal

force causes compression of the sample, which influences the measured shear modulus.

The skin surface roughness becomes less significant for other samples involving thicker

skin layers. In addition, these layers are more pliable than the 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 whithout affecting the

desmosomes. In the dermis, the shear response will be determined by the ground

substance, because the collagen and elastin fibers are mainly oriented transversally. It is

likely that this substance has a lower shear resistance than the highly organized

epidermis. By contrast, the tissue response to in-plane tensile loading will be determined

by the mechanical integrity 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 intact

skin and dermis-only was measured [91,92]. 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 of

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 those of the epidermis, although

some corrections are needed to account 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 are very similar. Currently, our shear moduli

can only be compared with in-plane tensile properties of stratum corneum from

literature. Accordingly, current values for shear moduli are one order of magnitude lower

than those in dry conditions and up to two orders of magnitude when fully hydrated,

based on the lowest reported values for tensile moduli [23,25,93]. 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 the present observations, delamination studies with stratum corneum,

which also showed the pre-failure mechanical response, showed no dependence over the

identical temperature range [94].

No clear relationship between the mechanical properties of the epidermis and RH could

be established. Time sweeps showed that moduli stabilize to an equilibrium 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 stratum

48 Chapter 3

corneum structure has revealed swollen corneocytes and water pools in the extracellular

spaces after storage in PBS [95]. Furthermore, water disrupts the lipid lamellae to

varying degrees and causes degradation of intercellular corneosomes [84,84,96]. 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 the present

samples. The prolonged time of conditioning for the sample at higher RH limited the

number of experiments that could be performed 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 experimental system. Viable epidermis could not be

measured as an isolated skin layer, but its properties can be derived from the other skin

samples. The 𝐺𝑑 for stratum corneum approximately ranges from 4 to 12 kPa, decreasing

with increasing RH. The values are considerably lower than 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 of the same order of magnitude, but

were less consistent possibly due to a less well-defined tissue composition. Therefore, it

would be interesting to combine mechanical testing with real-time imaging techniques to

monitor changes in tissue deformation. It has already been shown by histological

examination after 2 days of loading that shear forces induce cell displacement in skin,

and particularly in the epidermis [97].

Furthermore, electron microscope imaging techniques could support histological

examination in assessing tissue damage due to preparation, storage or handling. In

addition, it is important to correlate the shear response both with tensile testing and with

the effects of perpendicular loading, as accomplished with indentation or compression

testing.

Chapter 4

A new indentation method to determine

mechanical properties of the epidermis

The contents of this chapter are based on M. Geerligs, L.C.A.v. Breemen, G.W.M.

Peters, P.A.J. Ackermans, C.W.J. Oomens, and F.P.T. Baaijens, A new indentation

method to determine mechanical properties of the epidermis, submitted.

50 Chapter 4

4.1 Introduction

The outer skin layer possesses important characteristics that make it a favorable site for

pain-free drug delivery with minimal damage. Indeed it presents a rich population of

immunologically sensitive cells as well as the lack of blood vessels and sensory nerve

endings. The development of drug delivery using microneedles or microjets is a

challenging task because of the lack of 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 [12]. 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 to represent the mortar (see Figure 1.3). The viable

epidermis mainly consists of keratinocytes migrating 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 microjet 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. Penetration is preceeded by indentation of the skin

layers, especially the epidermis. So a full understanding of the delivery path requires also

the understanding of this indentation phase and, therefore, the knowledge of the

mechanical behavior of the epidermis. The modeling of tissue repair and remodeling in

future work benefit from a well determined mechanical behavior.

Recently, Kendall et al. [38] were the first to report on the 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 alone [98,99].

The authors argued that this is because of an increasing moisture content with depth. In

the viable epidermis, the storage modulus remained nearly constant. By contrast,

penetration of the 5 μm probe showed a negligible decrease in storage modulus

throughout the stratum corneum and a gradual increase in the viable epidermis, although

the values for the shear modulus were less than for the corresponding 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 [99]. A needle with a tip

radius 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 ensured that the test was

predominantly confined to the properties of the stratum corneum [34]. Measured forces

were typically of the order of 3.0 N. Recently, a limited number of nano-indentation

studies have been performed on isolated stratum corneum [32,98,100,101,101]. The tips

A new indentation method to determine mechanical properties of the epidermis 51

used varied between 1-10 μm, while corneocytes have a diameter ranging from 26-45

μm [3,102,103]. As a consequence, very local properties were determined in these

experiments. Furthermore, in some of the studies, the three-sided Berkovich tip that has

a sharp three-sided 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 [38,98,101]. The drawback of CSM is that the results are influenced by the

selected amplitude and frequency for viscoelastic materials. Combining the nano-

indentation studies on stratum corneum reveals a measured Young‟s moduli, which

varies from 10 MPa [104] for wet porcine samples up to 1 GPa for dried human samples

[98]. 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 mechanical properties of the stratum corneum

only at the tissue level is therefore also required.

The aim of the present study is to present such an indentation method and to use it to

determine the Young‟s modulus of the epidermis, i.e. the stratum corneum and viable

epidermis. The typical complex geometry, a variable thickness between 30 to 150 μm,

and the porosity of the epidermis places high demands on this mechanical

characterization. Therefore, isolated epidermis and isolated stratum corneum were tested

using equipment that is known for its accuracy and reliablity. The device is originally

designed for solid materials of which well-defined samples can be obtained and

therefore, the testing protocol needs to be adapted to epidermal samples. To validate that

the testing protocol holds for thin materials with a low stiffness, tests have been

performed with silicone rubber. Moreover, indentation experiments require a model for

the interpretation of the measured results. Next to the analytical model used, which

assumes a homogeneous linear elastic material behavior upon unloading, we also

adopted a numerical method that allows for taking into account geometrical details and

different material properties for different layers.


Skin

Indentation tests have been carried out on ex vivo abdominal skin of Caucasian women

aging 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 was brought into the laboratory and processed

within 4 hours. Epidermal sheets were obtained using a dermatome (D42, Humeca), in

which the prescribed 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

52 Chapter 4

amount of cones and ridges, samples may consist of epidermis and/or some papillar

dermis (see Figure 4.1b and c).

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.

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

VE

PD

RD

(b)

SC

VE

PD

(a) (c)

Figure 4.1. (a) Aldehyde-fuchsin stained full-thickness skin including the stratum corneum

(SC), viable epidermis (VE), papillar dermis (PD) and reticular dermis (RD); (b)

Dermatomed skin, set thickness of 100 μm, consisting of the epidermal layer only; (c)

Dermatomed skin, set thickness of 100 μm, consisting of epidermis and some fragments of

papillar dermis.

Figure 4.2. The center of the triangles, highlighted by the large red points, formed by the

glyphics was chosen as indentation location on the skin samples.

1.8 mm

SC

VE


Silicone rubber

In order to validate the experimental procedure for thin samples, a highly elastic silicone

rubber (Köraform 42 A, Alpina Silicone, 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. Thereafter, samples of about 1 cm2 were cut out.

4.1.2 Experimental procedure

Skin

The skin sample should be placed on a substrate such that in-plane tissue movement

could not occur. The large number of pores in the epidermis precluded the use of any

fixation method. It appeared, however, that the adhesive, sticky nature of the skin sample

was sufficient to no longer require any further fixation. Immediately after thawing at

room temperature, samples were spread out on an aluminum disc with the outer skin

surface facing upwards. 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 the use of

the built-in microscope of the NanoIndenter XP (MTS Systems, USA). Each location is

at least 500 μm away from each other, to avoid any influence between measurements.

The centers of triangles formed by the glyphics, the primary and secondary lines, were

chosen as the indentation locations to optimize the contact between the indenter and the

tissue [Figure 4.2].

All experiments were performed using a sapphire sphere with a radius of 500 μm. The

load and displacement resolutions are 1 nN and 0.01 nm, respectively. The maximum

load depends on the depth limit of indentation, which was set to a value which did not

exceed 10% of the sample thickness [105]. Preliminary testing demonstrated this

maximum load to be 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.

For both epidermis and stratum corneum, the protocol was repeated on three samples for

each subject. Test series were completed within 2 h. The temperature and humidity are

kept constant at 22°C and 28% RH, respectively.

Silicone rubber

The skin samples, particularly the stratum corneum samples, are extremely thin (under

20 μm). Measuring such thin samples might be at the resolution limit of the apparatus.

Therefore, to evaluate the usefulness of the protocol for thin materials, a well-defined

homogeneous soft material, silicone rubber, with different thicknesses (50-2000 μm) was

tested with the indentation protocol similar to that for the epidermis. The samples were

placed on the substrate without fixation. Indentation locations were identified

automatically, using a 3x3 grid with a distance of 500 μm between the various locations.

54 Chapter 4

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

Pharr [105], which assummes a fully elastic recovery upon unloading. From the initial

unloading slope of the load-displacement (𝑃, ℎ) curve, the reduced modulus Er is

obtained according to:

𝐸𝑟 = 𝜋

2

𝑑𝑃/dℎ

𝐴 (4.1)

where 𝐴 is the contact surface. In practice, the measured tip displacement is never equal

to the contact depth because, at the vicinity of the tip, the surface can either sink-in or

pile-up (see Figure 4.3). In that case, 𝐴 is replaced by the projected area 𝐴𝑝 , which 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

𝐸𝑟

=1 − 𝜈2

𝐸+

1 − 𝜈𝑖2

𝐸𝑖

(4.3)

where E and ν are the Young‟s modulus and the Poisson‟s ratio for the specimen and Ei

and νi for the indenter. The epidermis, stratum corneum and silicone rubber are all

assumed to approximate to incompressible materials, using a Poisson‟s ratio of 0.495.

Figure 4.3: Contact profile developed during indentation: 𝒉 is the indentation depth, 𝒉𝒄

the contact depth, and 𝒂 the radius.

Numerical model

To check if the Young‟s moduli obtained with the analytical method give reasonable

results, a finite element calculation using MSC.Marc (MSC.Software Corporation, Santa

Ana, USA) was used. An axisymmetric mesh was used to fit the experiments using a

hp

hchc

hp

a0

a

ah h

Pile-up Sink-in


Neo-Hookean model with different material parameters when modeling stratum corneum

and viable epidermis, 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, contact between the indenter and

the sample was assumed to be frictionless.

For silicone rubber, the Young‟s modulus, 𝐸𝑆𝑅 , was estimated by fitting the average

load-displacement curve of the 50 μm thick samples. This 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 assumed for the skin

samples too. First, the Young‟s modulus for the stratum corneum, 𝐸𝑆𝐶 , was derived by

fitting the average load-displacement curve of the stratum corneum samples of 20 μm.

This modulus was used to describe the experimental data of the epidermis, such that the

modulus for the viable epidermis, 𝐸𝑉𝐸 , could be derived. The thickness of the stratum

corneum was varied from 10 to 20 μm, the thickness of the viable epidermis was kept

constant at 80 μm. In order to assess the sensitivity of this fitting approach, the effect of

increasing 𝐸𝑆𝐶 or decreasing 𝐸𝑉𝐸 by a factor 2 on the maximum indentation depth was

studied.

4.2 Results

Silicone rubber

The load-displacement curves obtained for the silicone rubber samples are shown in

Figure 4.4. The results were highly reproducible for each thickness. The maximum

indentation depth decreases with a decrease in sample thickness. Consequently, the slope

of the initial unloading curve decreases, which is reflected in the average values for the

Young‟s moduli namely, 3.67±0.20, 2.22±0.10 and 1.69±0.04 MPa for a sample

thickness of 50, 120 and 2000 μm, respectively. It is evident that the assumption of a

linearly elastic half space from Oliver & Pharr [105] is not valid for these thin samples.

Using the FE model, the Young‟s modulus was estimated to be 2.16 MPa. When using

this value to describe the unloading curves for the 120 and 2000 μm thick sample, it is

shown that the unloading curves and maximum indentation depth for all thicknesses are

in good agreement to the experimental data (Figure 4.5).

Epidermis and stratum corneum

An example of the results from one subject is shown in Figure 4.6. Data that displayed

significant measurement errors or deviated from the general response, were ignored. In

generally, 2 or 3 tests out of a series of 9 measurements were excluded from subsequent

calculations. Figure 4.7 clearly shows that the average curves overlap for different

subjects. Estimates for the Young‟s moduli were derived via the analytical approach and

found to be 2.6±0.6 MPa and 1.1±0.2 MPa for the stratum corneum and epidermis,

respectively.

56 Chapter 4

Figure 4.4. All force-indentation (𝑷, 𝒉) curves for silicone rubbers with different thicknesses.

Figure 4.5: The unloading curves obtained with the FE model (x) are depicted together

with the mean curves of the experimental data.

The fitting results using the FE-model are shown in Figure 4.8. From the stratum

corneum experiments, 𝐸𝑆𝐶 was calculated to be 0.6 MPa. For a 20 μm thick stratum

corneum and 80 μm thick viable epidermis, 𝐸𝑉𝐸 equals 𝐸𝑆𝐶 with a value of 0.6 MPa.

Decreasing the thickness of the stratum corneum to 10 μm minimally affects the

unloading curve. Also increasing the stiffness of the stratum corneum did not have a

noticeable effect. However, a reduction in the stiffness of the much thicker viable

epidermis caused an increase in indentation depth, from approximately 8 to 12 μm.

4.3 Discussion

The major problem in performing indentation experiments on skin is probably the skin‟s

surface roughness. In order to have a surface as smooth as possible, we used a large

spherical indenter (ø=500 μm) such that the contact area was much larger 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

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 a

well-defined 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


Figure 4.8. Unloading curves for the epidermius obtained with the FE model are depicted

together with the experimental indentation curve. The thickness of the stratum corneum is

varied from 10 (dashed lines) to 20 μm (dotted lines).

(a) (b)

Figure 4.6. Indentation curves of subject 2 for stratum corneum (a) and epidermis (b).

Dotted curves were not included in calculating the average curve.

(a) (b)

Figure 4.7 Average indentation curves for stratum corneum (a) and epidermis (b).

58 Chapter 4

damaging the sample [106]. Using this measurement protocol, highly reproducible data

could be obtained for all subjects and the variance between the subjects was negligbly

small.

In order to obtain meaningful, reproducible data from in vitro experiments, 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 papillar dermis in the sample.

Current 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

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 homogeneous materials responding fully elastically upon unloading. In the case

of soft tissues, this assumption is not valid because of material responses like piling-up

and sinking-in. Due to piling up of the tissue, the projected contact area is bigger then

used in the calculations (see Figure 4.3). In the present study, the deviation is mild,

because the use of a large spherical indenter reduces the boundary effects.

The introduction of a numerical model should result in a better approach. The results

show that the stiffness of the viable epidermis is comparable to that of the stratum

corneum. For both epidermal layers, the stiffness of the two layers is approximately 1

MPa, which shows that the viable epidermis considerably contributes to the mechanical

response of skin at this length scale and load direction. In comparison with literature

using indentation test, current values for stratum corneum are on the low side of the

published range [35,98,104]. 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 [98,107,108].

Extending the Neo-Hookean model to a multimode Maxwell model would be a logical

step forward. However, the relaxation spectrum and corresponding low shear moduli that

were derived from rheological experiments (see Chapter 3) do 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 at higher loading rates than those used in the present experiments

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).

To conclude, the small deformation behavior of epidermis was studied. 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 length scale

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 need 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.

62 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 [39,40,44,109,110]. 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 [1,111]. 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 [18].

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 [18]. 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

[20,112]. Both morphological and metabolic differences were found between those two

layers [112-114], but it is not clear if these layers differ in terms of the mechanical

properties.

Linear viscoelastic behavior of subcutaneous adipose tissue 63

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 [42,44,45,92,115,116]. 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 [109]. Patel et al. [40] 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 [41,117].

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 [109,117], or because the authors were only

interested in a comparison of properties between human fat and a mimicking material

[40].

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.

64 Chapter 5

5.2 Methods and Materials


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 [118]. 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 [82]. Rapid freezing, which has not been demonstrated to

change the fatty acid composition compared to fresh tissues [119], 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.


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.


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:

𝛾 𝑡 = 𝛾0 sin 𝜔𝑡 , (5.1)

𝜏 𝑡 = 𝐺𝑑 sin 𝜔𝑡 + 𝛿 . (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, 1

2𝜋 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 𝛿 =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 [120]. 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 𝛿 𝜔, 𝑇 = tan 𝛿 𝑎𝑇𝜔, 𝑇0 , (5.5)

𝐺𝑑 𝜔, 𝑇 =1

𝑏𝑇𝐺𝑑(𝑎𝑇𝜔, 𝑇0) . (5.6)

5.2.3 Testing procedure

Test protocols were based on measuring the linear viscoelastic properties of other soft

biological tissues, such as brain [121], muscle [122] and thrombus [90]. 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

66 Chapter 5

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 # Specimens (# samples per specimen)

Strain Sweep 3 (4,5,3)

Frequency Sweep 3 (3,3,3)

5x repeated

Model fit 3 (3,3,3)

Effect snapfreezing 3 (3,5,3)*

Frequency/Temperature Sweep 2 (3,3)

Increasing T

Decreasing T 2 (2,1)

Stress relaxation 1 (5)

* sample number per condition

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 [123] 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

Figure 5.1 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.

68 Chapter 5

(a) (b)

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:

𝜏 = 𝐺 𝑡 − 𝑡 ′ 𝛾 𝑡′ 𝑑𝑡′𝑡

−∞, (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:

with G'(1) and p as constants [124]. The same relation is used for G''.

𝐺 ′ 𝜔 = 𝐺 ′(1)𝜔𝑝 , (5.8)


The phase angle can be expressed in terms of the exponent p [124,125]:

𝑡𝑎𝑛𝛿 =𝐺"

𝐺′=

𝑝𝜋

2. (5.9)

So the small oscillatory strain behavior is captured by an approximation with only two

constants (p,G(1)). It is known [124] 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!)

𝑝𝜋𝑠𝑖𝑛

𝑝𝜋

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 ),(),( 0TaGTG Tdd .

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 = 𝑒𝑎𝑇02+𝑏𝑇0+𝑐 , (5.12)

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.

70 Chapter 5

(a) (b)

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

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 signal-

to-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 [40]. In

comparison with Samani et al. [44], 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

72 Chapter 5

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 three-

dimensional 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

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 (2010), Does subcutaneous adipose

tissue behave as an (anti-)thyxotropic material?, Journal of Biomechanics, accepted.

74 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 [126]. 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 [41]. 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 [127]. Few related in vitro studies on

tissue behavior exist. Of these, rheological measurements demonstrated a decrease in

viscosity with increasing shear rate [40]. 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-and-

hold cycles during confined compression tests at various ramp rates [40,41]. 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 [111]. 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 [117,128].

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? 75

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 [18]. 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 [126].

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


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 [118]. 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

76 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 14-

18 weeks old at necropsy.


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 [126]. 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 [126]. 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.


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 Table 6.1.

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.

78 Chapter 6

Table 6.1: Overview of number of samples used for the experiments.

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.

(a) (b)

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,

the initial moduli remains constant at the plateau value, see Fig. 3a. After a recovery

Test # specimens (# samples per specimen)

Time sweep 4 (1,6,4,2)

Constant shear rate

increasing shear 1(3)

decreasing shear 1(2)

Constant shear 2(3,3)

Stress relaxation 2(3,3)


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.

80 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.


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

82 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

[117].

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 [89,121,129]. 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

[130]. 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 anti-

thixotropic 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 [130,131]. 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 [130]. When establishing such experiments, the large 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 loading-

unloading 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 [40]. 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 [132,133]. 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

84 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

86 Chapter 7

7.1 Introductory remarks

The mechanical behavior of skin is of utmost importance for many 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 have utilized

in vivo testing for obvious reasons. Non-invasive studies can then be applied on skin in

its natural environment at different body sites in a reasonably cost-effective manner.

However, non-invasive measurements require elegant procedures with a lot of

assumptions to simplify the models describing the experiment or numerical-experimental

procedures including inverse parameters estimations. Nonetheless, methods are quite

succesful for mechanical chararcterization of the dermis. The overall mechanical

behavior of skin is often considered to be dominated by the dermal properties.

Most clinical and cosmetic applications require more detailed knowledge about

individual layers at the skin surface, viable epidermis and stratum corneum, and the

deeper hypodermis. It is required to accurately measure displacements in each layer with

non-invasive methods like ultrasound, MRI, confocal microscopy, and Optical

Coherence Tomography. The different lengthscales, ranging from 10 μm of the stratum

corneum to the cm scale for the hypodermis, and the inverse relationship between

penetration depth and resolution of each of the techniques present a major problem. The

length scales and the range of stiffness values from the different layers also form a major

difficulty for the numerical simulation tools, as well as for the parameter algorithms [22].

These issues associated with in vivo testing have stimulated the present work in which

individual human skin layers were mechanically characterized in a reliable and

reproducible manner using an in vitro experimental system. The layers of interest were

the stratum corneum, viable epidermis and hypodermis, because their mechanical

behavior is largely unknown or has produced inconsistent literature. 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 the

skin samples. As epidermis exists as a layered structure, the small strain behavior was

determined in both in-plane and perpendicular directions under various environmental

conditions. For the hypodermis, rheological experiments were used to study both linear

and non-linear behavior. The experimental approach for the different skin layers is

depicted in Figure 7.1.

In the following sections the in vitro model (Section 7.2) and the mechanical testing

methods (Section 7.3) are appraised. Thereafter, the implications for clinical and

cosmetic treatments (Section 7.4) and recommendations for further research (7.5) are

provided.

General discussion 87

Figure 7.1: Experimental approach for each skin layer.

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, as opposed to cadaveric skin, provides higher

viability and is, moreover, available from a range of ages. 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 subcutaneous

adipose tissue has undergone changes in comparison with healthy subjects [18,134].

Therefore, a porcine model was introduced for this tissue layer, which is comparable in

structure and function to human adipose tissue [118]. In addition, its availability and

reproducibility makes it an attractive option.

After harvesting the skin tissue, the necessary skin layer must be isolated. As the time

between harvesting and mechanical testing is usually too long to maintain the tissue

viability and intregrity, means of preservation were needed. It is essential to ensure that

preparation treatments do 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 [135] are mainly guided by

cost, time effectiveness and ease of use. However, it is widely known and demonstrated

that these ways of tissue preparation influence mechanical properties [136]. In particular,

the epidermal layers are known to be highly sensitive to chemical and physical changes

in the environment. Much knowledge is already available from skin grafting techniques

for burn wounds. However, the required tissue condition is different from in vitro testing

[54,74]. Therefore, available and new techniques to isolate and preserve epidermis and

stratum corneum were assessed as to their successfulness with reference to the

maintenance of tissue integrity and viability. Furthermore, both the ease of handling and

the reproducibility of the protocol were considered.

From the numerous techniques in use to isolate the epidermis, it was concluded that

many are limited when considering the maintenance of tissue integrity and viability.

slicer

isolation preservation

PBS(or -80°C)

HHBSS(or -80°C)

PBS(or drying)

dermatome

linear + non-linear

linear

linear

hp

hc

a

h

mechanical testing

linear

linear

indentation

shear

88 Chapter 7

However, cutting using a dermatome and enzymatic digestion with dispase fulfills both

requirements and also provides ease of handling and reproducibility. Cutting result into a

better defined sample geometry, which is convenient for most mechanical tests. As

shown in Chapters 3 and 4, when fragments of papillar dermis were present in the

epidermal samples (Figure 4.1), this did not lead to a measurable influence on the results

for small strain behavior. As established in these chapters, the mechanical behavior of

stratum corneum and viable epidermis are comparable and both have a higher stiffness

than the dermal layer. Thus, it can be assumed that the influence of fragments of papillar

dermis in the samples will also be minimal in large deformation studies. Regarding the

isolation of the stratum corneum, the gold standard represents enzymatic digestion with

0.1% trypsin. Some other techniques were analyzed but none comparable performance.

For the present studies, after separation the samples were preferrentially stored in

HHBSS in an incubator at 37°C and 5% CO2. It should be noted that a well-controlled

environment is much better achievable during storage in an incubator than during

transport and the mechanical tests. Usually, temperature control is built in a device but

implementation of a humidity control sytem remains difficult because it might influence,

for example, the sensitivity of fragile load cells.

The practical problems that have to be dealt with, emphasize 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 separated epidermis sample 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 the addition of a drop water. Skin

samples including reticular dermis curl up and twist, which makes gentle treatment

challenging. With reference to adipose tissue, each touch causes geometric deformations,

which hinders the correct placement of the sample in the experimental set-up.

7.3 Mechanical methods

In this thesis, new protocols were developed for the mechanical characterization of

separate skin layers for available, reliable and accurate equipment. The standard

techniques for the in vitro mechanical characterization of skin layers are uniaxial and

biaxial testing. Uniaxial tensile tests are relatively easy to perform, cost-effective and

testing equipment is available 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 comparison 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 as well as to determine the width and

thickness of the sample due to the presence of the skin lines. In addition, it is also

difficult to define the unloaded initial configuration because of the natural pre-stress in

the skin.

Other techniques, such as indentation and rotational shear, are better able to

accommodate these issues and therefore provide 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 relatively low stiffness, and the sample thickness and the rough surface of

epidermis and stratum corneum only. The newly developed protocols that were validated

with silicone rubber resulted into a set of repoducible data for all measured skin layers.

Only linear shear properties of the epidermis showed large variations (Figure 3.10).

In general, rheological experiments aim to characterize the viscoelastic response of soft

materials, requiring relatively large homogeneous samples. To be able to obtain a more

homogeneous strain field as well as to increase the accuracy, an eccentric configuration

that was especially designed for measuring soft tissues [88], was used for the upper skin

layers (see Figure 3.2). Temperature and humidity could be well regulated by a home-

built system. The measurement chamber with controlled environment could not be

closed completely, because it would then interfere with the applied shear. Accordingly,

the temperature and humidity sensors were placed close to the sample to ensure a stable

environment in that area. In addition, for the upper skin layers, the required settings were

close to the limitations of the apparatus. As the axial resolution is 1 μm, the rheometer

cannot be used to perform compression tests on stratum corneum, which is 10-20 μm in

thickness. In addition, there is some uncertainty about the shear data for the stratum

corneum, because of the thin, undulating geometry of the sample (Chapter 3).

Nonetheless, the results were reproducible, which indicates that the measurement itself is

reliable.

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 tissue, another set of experiments

designed to examine thixotropic behavior, would be appropriate. Although thixotropic

studies have been extensively discussed, appropriate protocols for biological tissues are

not available.

Indentation methods, such as the NanoIndenter XP (MTS Systems, USA), are

increasingly used to probe the mechanical response of biological materials. Because of

the variable probe size, indentation can be used to measure the mechanical properties of

biological samples ranging from cell membranes up to the global tissue level. In

addition, the method is appropriate for very thin, small and heterogeneous samples. This

90 Chapter 7

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 obtain a good

contact during indentation. Because of the sensitivity of the load cells, it is challenging

to regulate humidity. Another related problem could be the definition of the initial

sample height, because the role of adhesive forces increases in the contact definition

problem. In addition, visualization of the experiment is not yet possible. Therefore,

alternative indentation set ups as developed by Cox et al. [137] might be more

appropriate for future work.

From a mechanical point of view, the Nanoindenter XP is a very interesting technique

for further research involving both the non-linear behavior of the upper skin layers and

failure behavior. Indeed Wu et al. [35] has already developed methods to determine

properties, such as fracture behavior, from load-displacements curves of the stratum

corneum. When a direct coupling between structure and loading is essential, other

methods need to be considered.

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 between 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 from the literature as indicated in Table 7.1. For the shear

experiments, the Young‟s modulus was derived from the shear modulus assuming Neo-

Hookean material behavior, such that 𝐸 = 3𝐺𝑑 . Although some authors have assessed

the stiffness of the (viable) epidermis in combination with intact papillar dermis

[38,39,138], the present study provides the first data obtained from epidermis within a

small strain regime. As shown in Chapter 4, analytical methods are not able to describe

the indentation experiments, such that a finite element model was used to obtain a value

for the stratum corneum and viable epidermis from the indentation experiments.

According to the highly anisotropic structure of the epidermis with the keratinocytes and

corneocytes, which change shape with depth, enormous differences in values exist

between loading directions. The differences can be further explained by the fact that

different structural components play a dominant role with various loading modalities.

The resistance of the keratinocytes mainly determine the mechanical response during

shear, while the tensile stiffness is determined by the connections between the

keratinocytes, i.e. the desmosomes. As indentation is a mixture of compression, tensile

and shear forces, it is difficult to identify which of the structural components 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 is similar 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 decreased with increasing humidity, but was minimally

influenced by temperature and frequency.

Table 7.1: Overview of Young’s moduli for all skin layers. Those determined in this work

are given in bold.[139]

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. This value is in good agreement with literature

data [40,43]. Prolonged loading results in a dramatic stiffening of the material. This

behavior is reversible with a recovery time of about 3 hours (Figure 6.3).

The studies on its non-linear behavior suggest tissue structure changes with increasing

strains. Up to 0.15 strain, the adipose tissue behaves as a Mooney-Rivlin material.

Thereafter, the stress response decreases with increasing strain until a strain of 0.3.

Higher strains result in the same maximum stress level. In addition, this behavior appears

to be reversible in nature as well.

The present data suggest that adipose tissue behaves like a thixotropic material. Before

numerical models can be developed, more experiments are required to fully describe its

non-linear behavior.

7.5 Implications for clinical and cosmetic applications

The research presented in this thesis is part of larger research programmes being pursued

jointly within Philips Research and Eindhoven University of Technology (TU/e). The

relevance of the work in this thesis has already been indicated in Chapter 1. In this

section, the implications for some of these applications are discussed.

SHEAR INDENTATION TENSILE

Eshear [kPa] Eindent [kPa] Euniaxial [kPa]

Stratum corneum 25% RH 30 600 0.04-10∙106

98% RH 10 n.a.* 6-10∙104

(Viable) epidermis 25% RH 30 600 n.a *

98% RH 10 n.a.* n.a.*

Dermis 8 1-10[100,110] 1-20∙103 [5,139]

Hypodermis 24 20-30[41] n.a.*

*n.a =not available

92 Chapter 7

In Philips Research, part of the innovation is related to consumer products that are in

contact with skin, like electric shavers. During shaving, the skin penetrates the slits of a

shaver, a process known as doming. To enhance shaving performance, the hairs must be

cut as close as possible to the skin surface without causing irritation or other damage to

the skin. The small length scale of the skin doming requires that the top layers are

included in numerical simulations. To date, however, the influence of the top layer on

doming has been difficult to incorporate. During shaving, the underlying tissue can be

soft tissues, such as adipose tissue, or 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 hydrating additives might affect the mechanical

behavior of the top layers and thereby affect skin doming.

At TU/e, an ongoing research programme is focused on the early detection and

evaluation of (deep) presssure ulcers. Pressure ulcers are defined as areas of soft tissue

breakdown that result from sustained mechanical loading, involving both compression

and shear, of skin and underlying tissues. To date, this work was mainly focused on early

markers in skin [140,141] and the mechanisms associated with muscle injury [142-145].

The poor understanding of the mechanical behavior of adipose tissue has limited its

corporation into the current research. Thus, in pressure ulcer research, the mechanical

behavior of adipose tissue has been largely ignored. The thixotropic behavior observed

in the present thesis would undoubtedly have an influence on the load distribution in soft

tissues during prolonged loading and thus it must play a role in the aetiology of pressure

ulcers.

The mechanical behavior of the top layers of the skin is important especially in the case

of high shear forces of the surface where friction also plays a role. Therefore, the

observed large differences in shear stiffness and compression stiffness are relevant. It is

evident that failure studies are required both for adipose tissue as well as epidermis.

7.6 Recommendations

Some important questions remained unanswered. The research described in this thesis

was focused on a reliable in vitro mechanical characterization of separate skin layers.

However, to fully understand the mechanical behavior of a heterogeneous sample, it is

necessary to understand how mechanical damage affect the tissue composites. For

instance, 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 give additional information for the interpretation of

measurements, such as those involving 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 limit the visualization of

the epidermis during mechanical testing. Therefore, 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 [146,147]. In

addition, confocal imaging techniques are able to track the cell nuclei with more than 10

images per second [148,149]. Both techniques have the advantage that images can be

obtained from intrinsic tissue properties only, thus making them appropriate for in vivo

imaging. Another imaging technique involves the combination of a second-harmonic

signal and 2-photon imaging as developed by Palero et al. [150]. They demonstrated

with both in vivo and ex vivo epidermal tissue from mice that the viability of cells and

the structure of the cell membranes could be measured simultaneously. In the longer

term, 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 groups of adipocytes. The relative

large structures, i.e. adipocytes have a diameter up to 70 μm, 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. If this can be

resolved, then three-dimensional techniques such as optical projection tomography can

prove useful. However, in the meantime, the geometric deformations of the adipose

tissue samples can be examined during mechanical behavior. In 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 large deformation

behavior of those layers as well. In principle, the experimental approaches presented in

this thesis can be used to develop testing appropriate for the non-linear region.

Ultimately, experimental studies on the failure behavior are necessary, which might

incorporate models for transport and structural damage. Therefore, it would also be

desirable to perform those studies on in vitro human skin.

Current tests proved that the non-linear behavior of adipose tissue is rather complex and,

as yet, can not be captured in a constitutive model. Therefore, a new set of experimental

data have to be designed which are appropriate for input into such a constitutive model.

An overview of these type of tests is described by Dullaert et al. [151,152]. Mechanical

tests in other loading directions should also be performed. Compression tests are most

relevant to clinical and cosmetic applications and can also be performed on a rheometer.

It should be noted that the present work was conducted on abdominal skin from

Caucasian women in the age group of 35-55 years. Skin with 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,

etc.

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

94 Chapter 7

investigating new prevention and treatment strategies. For applications, such as pressure

ulcers and transdermal drug delivery, it would be advisable to incorporate transport

models.

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 anisotropically in the

small strain regime and the stiffness of the viable epidermis is equivalent to that

of the stratum corneum in each loading direction,

2) the hypodermis initially shows typical small strain behavior for soft tissues, but

appears 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 experimental methods, which have been developed,

represent valuable tools for studying the mechanical properties of skin in relation to

disease and treatments in future.

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. Tot nu toe is veel

onderzoek naar de mechanische eigenschappen van de huid uitgevoerd door middel van

in vivo experimenten, waarbij werd aangenomen dat de middelste huidlaag met zijn

vezelstructuur representatief is voor de huid. Het doel van dit promotieonderzoek was

om de mechanische eigenschappen van de afzonderlijke huidlagen te karakteriseren.

Hierbij is de aandacht specifiek gericht op die huidlagen, waarvan nog nauwelijks

literatuur beschikbaar is of de resultaten in de literatuur inconsistent zijn.

Allereerst is onderzocht wat de beste methoden zijn om de verschillende huidlagen van

elkaar te scheiden en levensvatbaar te houden in een in vitro omgeving.

Aandachtspunten hierbij waren het effect van een methode op de weefselstructuur en de

levensvatbaarheid en daarnaast de betrouwbaarheid, duur en de uitvoerbaarheid. Hieruit

kon geconcludeerd worden dat voor dit onderzoek de epidermis het best geïsoleerd kan

worden met een dermatoom. Vervolgens is de epidermis bewaard in HHBSS of

ingevroren volgens een specifiek protocol. Het stratum corneum kan van de epidermis

geisoleerd worden door gebruik te maken van het enzym trypsine en vervolgens bewaard

in PBS of in gedroogde vorm.

Vervolgens zijn er verschillende methoden ontwikkeld om de mechanische reactie van

de afzonderlijke huidlagen te meten. Voor de bovenste huidlagen, de epidermis en

stratum corneum, zijn in vitro meetopstellingen gebouwd om de mechanische respons bij

kleine rekken te kunnen meten. Onder fysiologische omstandigheden 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 er geen significant verschil is tussen de mechanische

96 Samenvatting

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

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 tijd. De frequentie- en temperatuurafhankelijkheid 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 resulteert in veranderingen in de

weefselstructuur die reversibel bleken. Tot zekere schuifrekken is het gedrag van

onderhuids vet 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 power-law 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

weefselstructuur en het mechanisch gedrag te bestuderen met behulp van

visualisatietechnieken. Daarnaast zal het onderzoek uitgebreid moeten worden met

studies naar het faalgedrag van de individuele huidlagen in relatie tot klinische en

cosmetische toepassingen.

Dankwoord

Graag wil ik iedereen bedanken die (in)direct 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 deze samenwerking heb

ik gebruik kunnen maken van de faciliteiten van beide zijden 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 is 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 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!:-) Anke en

98 Dankwoord

Maria, ik blijf het leuk vinden om af 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 zeeën 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. 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 relativerende 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 geautomatiseerd 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. Vanaf 1 december 2009 is zij werkzaam bij Philips Consumer Lifestyle te

Drachten.

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, 24(2), 215-224.

[2] Plewig, G. and Marples, R. R., (1970). Regional differences of cell sizes in the human

stratum corneum. I. British Journal of Dermatology, 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. Journal of

Investigative Dermatology, 62(4), 415-422.

[4] Brody, I., (1977). Ultrastructure of the stratum corneum. International Journal of

Dermatology, 16(4), 245-256.

[5] Agache, P. and Humbert, P., (2004). Measuring the skin, Springer Verlag, Berlin.

[6] Plewig, G., (1970). Regional differences of cell sizes in the human stratum corneum. II.

Effects of sex and age. British Journal of Dermatology, 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, 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, 246(3), 271-283.

[9] Elias, P. M., (2005). Stratum corneum defensive functions: an integrated view. British

Journal of Dermatology, 125(2), 183-200.

[10] Marks, R., (2004). The stratum corneum barrier: The final frontier. Journal of Nutrition,

134(8 SUPPL.), 2017-2021.

References 101

[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. Journal of

Investigative Dermatology, 82(2), 188-194.

[12] Elias, P. M. and Feingold, K. R., (2005). Skin Barrier, Taylor & Francis Group, New

York.

[13] Sandby-Moller, 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, 83(6), 410-413.

[14] Silver, F. H., Siperko, L. M., and Seehra, G. P., (2003). Mechanobiology of force

transduction in dermal tissue. Skin Research and Technology, 9(1), 3-23.

[15] Montagna, W., Kligman, A. M., and Carlisle, K. S., (1992). Atlas of Normal Human Skin, Springer Verlag, New York.

[16] Odland, G., (1991). Structure of skin. In: Goldsmith, L. A. (eds), Physiology,

biochemistry, and molecular biology of the skin, Oxford University Press, Oxford.

[17] Briggaman, R. A. and Wheeler, C. E., Jr., (1975). The epidermal-dermal junction. British


[18] 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, 53(4), 671-683.

[19] Albright, A. L. &. S. J. S., (2005). adipose tissue, http://sportsci.org/encyc/adipose/adipose.html

[20] 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, 11(1), 5-11.

[21] Elsner, P., Berardesca, E., Wilhelm, K. P., and Maibach, H. I., (2002). Bioengineering of

the skin; Skin Biomechanics, CRC Press.

[22] Hendriks, F. M., (2005). Mechanical Behaviour of Human Epidermal and Dermal Layers.

Technische Universiteit Eindhoven.

[23] 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, 56(1), 72-78.

[24] 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, 233-

12.

http://sportsci.org/encyc/adipose/adipose.html

102 References

[25] 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, 399(1),

170-180.

[26] Wolfram, M. A., Wolejsza, N. F., and Laden, K., (1972). Biomechanical properties of

delipidized stratum corneum. British Journal of Dermatology, 59(6), 421-426.

[27] 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, 71(2), 140-144.

[28] de Rigal, J. and Leveque, J. L., (1985). In vivo measurement of the stratum corneum

elasticity. Bioengineering and the Skin, 1(1), 13-23.

[29] 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, 269(3), 221-232.

[30] 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, 269(2), 127-135.

[31] Pierard, G. E., Nikkels-Tassoudji, N., and Pierard-Franchimont, C., (1995). Influence of

the test area on the mechanical properties of skin. Dermatology, 191(1), 9-15.

[32] 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.

[33] Gardner, T. N. and Briggs, G. A. D., (2001). Biomechanical measurements in

microscopically thin stratum corneum using acoustics. Skin Research and Technology,

7(4), 254-261.

[34] 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.

(eds), Bioengineering of the skin: skin biomechanics, CRC Press LLC, Boca Raton.

[35] 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, 27(5), 785-795.

[36] Koutroupi, K. S. and Barbenel, J. C., (1990). Mechanical and failure behaviour of the

stratum corneum. J.Biomech., 23(3), 281-287.

References 103

[37] Middleton, J. D., (1969). The effect of temperature on extensibility of isolated corneum

and its relation to skin chapping. Br.J.Dermatol., 81(10), 717-721.

[38] 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, 28(33), 4968-

4977.

[39] 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, 28(3),

259-266.

[40] 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, 73A(3), 313-319.

[41] Gefen, A. and Haberman, E., (2007). Viscoelastic properties of ovine adipose tissue

covering the gluteus muscles. Journal of Biomechanical Engineering-Transactions of the

Asme, 129924-930.

[42] Samani, A. and Plewes, D., (21-9-2004). A method to measure the hyperelastic

parameters of ex vivo breast tissue samples. Phys.Med.Biol., 49(18), 4395-4405.

[43] 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. Physics in Medicine and Biology, 52(6), 1565-1576.

[44] 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, 52(6), 1565-1576.

[45] 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. Journal of Magnetic Resonance Imaging, 17(1), 72-85.

[46] 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 (eds),

Methods in Skin Research, John Wiley & Sons Ltd., New York.

[47] 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, 9(4), 321-330.

[48] 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, 32(4), 423-429.

104 References

[49] 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. Toxicology and Applied Pharmacology, 99(3), 522-533.

[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. Journal of Cutaneous Pathology, 18(1), 8-12.

[51] Steinstrasser, I. and Merkle, H. P., (1995). Dermal metabolism of topically applied drugs:

pathways and models reconsidered. Pharmaceutica Acta Helvetiae, 70(1), 3-24.

[52] Ogura, R., Knox, J. M., and GRIFFIN, A. C., (1960). Separation of epidermis for the

study of epidermal sulfhydryl. Journal of Investigative Dermatology, 35239-243.

[53] VAN SCOTT, E. J., (1952). Mechanical separation of the epidermis from the corium.

British Journal of Dermatology, 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, 26(4), 367-378.

[55] Kiistala, U., (1968). Suction Blister Device for Separation of Viable Epidermis from

Dermis. Journal of Investigative Dermatology, 50(2), 129-137.

[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. Journal of Investigative Dermatology, 65(6), 513-521.

[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., 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.

[59] Dimond, R. L., Erickson, K. L., and Wuepper, K. D., (1976). The role of divalent cations

in epidermolysis. Br.J.Dermatol., 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, 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. Journal of Investigative Dermatology, 81(2), 149-153.

References 105

[62] Baden, H. P. and Gifford, A. M., (1970). Isometric contraction of epidermis and stratum

corneum with heating. Journal of Investigative Dermatology, 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. Pharmaceutical Research, 15(1), 82-

84.

[64] Briggaman, R. A., (1982). Biochemical composition of the epidermal-dermal junction and

other basement membrane. Journal of Investigative Dermatology, 78(1), 1-6.

[65] Walzer, C., Benathan, M., and Frenk, E., (1989). Thermolysin treatment: a new method

for dermo-epidermal separation. British Journal of Dermatology, 92(1), 78-81.

[66] Liu, S. C. and Karasek, M., (1978). Isolation and growth of adult human epidermal

keratinocytes in cell culture. British Journal of Dermatology, 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. British Journal of Dermatology, 93(2), 287-290.

[68] Kitano, Y. and Okada, N., (1983). Separation of the epidermal sheet by dispase. British


[69] 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., 77(4), 183-187.

[70] 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., 42(16), 3027-3037.

[71] Ferguson, J., (1977). A method to facilitate the isolation and handling of stratum corneum.

British Journal of Dermatology, 96(1), 21-23.

[72] Humphries, W. T. and Wildnauer, R. H., (1971). Thermomechanical analysis of stratum

corneum. I. Technique. British Journal of Dermatology, 57(1), 32-37.

[73] 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. Archives of Dermatological Research, 268(1), 53-64.

[74] Kearney, J. N., (2005). Guidelines on processing and clinical use of skin allografts.

Clinical Dermatology, 23(4), 357-364.

[75] Zieger, M. A. J., Tredget, E. E., and McGann, L. E., (1996). Mechanisms of cryoinjury

and cryoprotection in split-thickness skin. Cryobiology, 33(3), 376-389.

106 References

[76] Ben Bassat, H., (2005). Performance and safety of skin allografts. Clinical Dermatology,

23(4), 365-375.

[77] 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.

Polish Journal of Veterinary Sciences, 6(1), 17-28.

[78] De, A., Mathur, M., and Gore, M. A., (2008). Viability of cadaver skin grafts stored in

skin bank at two different temperatures. Indian Journal of Medical Research, 128(6),

769-771.

[79] 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, 29(8), 759-767.

[80] Han, B. and Bischof, J. C., (2004). Engineering Challenges in Tissue Preservation. Cell

Preservation Technology, 2(2), 91-112.

[81] Lawrence, J. C., (1972). Storage and skin metabolism. British Journal of Plastic Surgery,

25(4), 440-453.

[82] Foutz, T. L., Stone, E. A., and Abrams, Jr, (1992). Effects of freezing on mechanical

properties of rat skin. American journal of veterinary research, 53(5), 788-792.

[83] Bouwstra, J. A., (2009).

[84] 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, 120(5), 750-758.

[85] 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, 37(5), 670-672.

[86] 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, 8(1), 27-36.

[87] Geerligs, M., (2007). Mechanical behaviour of human skin layers: stratum corneum,

living epidermis and hypodermis. A literature review.PR-TN 2006/00450

[88] 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, 42(3), 193-207.

References 107

[89] 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, 43(5), 623-636.

[90] 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, 43(6), 695-707.

[91] Holt, B., Tripathi, A., and Morgan, J., (28-8-2008). Viscoelastic response of human skin to

low magnitude physiologically relevant shear. J.Biomech., 41(12), 2689-2695.

[92] 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, 51(8), 980-989.

[93] 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, 23(1), 13-21.

[94] Wu, K. S., Li, J., Ananthapadmanabhan, K. P., and Dauskardt, R. H., (2007). Time-

dependant intercellular delamination of human stratum corneum. Journal of Materials

Science, 42(21), 8986-8994.

[95] 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, 106(1), 89-95.

[96] Warner, R. R., Stone, K. J., and Boissy, Y. L., (2003). Hydration disrupts human stratum

corneum ultrastructure. Journal of Investigative Dermatology, 120(2), 275-284.

[97] 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 and Regeneration, 16(1), 102-107.

[98] 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, 263(7-12 SPEC. ISS.), 1038-1043.

[99] 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. Bioengineering and the Skin, 234-240.

[100] 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, 18(15-

16), 1739-1758.

108 References

[101] 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

[102] Plewig, G. and Marples, R. R., (1970). Regional differences of cell sizes in the human

stratum corneum. I. Journal of Investigative Dermatology, 54(1), 13-18.

[103] Plewig, G., (1970). Regional differences of cell sizes in the human stratum corneum. II.

Effects of sex and age. Journal of Investigative Dermatology, 54(1), 19-23.

[104] Yuan, Y. and Verma, R., (2006). Measuring microelastic properties of stratum corneum.

Colloids and Surfaces B: Biointerfaces, 48(1), 6-12.

[105] 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, 7(6), 1564-1583.

[106] Ebenstein, D. M. and Pruitt, L. A., (2006). Nanoindentation of biological materials.

Nano Today, 1(3), 26-33.

[107] Jachowicz, J., McMullen, R., and Prettypaul, D., (2007). Indentometric analysis of in

vivo skin and comparison with artificial skin models. Skin Research and Technology,

13(3), 299-309.

[108] Boyer, G., Laquize, L., Le Bot, A., Laquize, 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), 15(1), 55-67.

[109] 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., 44(5), 208-217.

[110] Haberman, E. and Gefen, A., (2007). Viscoelastic properties of ovine adipose tissue

covering the gluteus muscles as related to pressure ulcer modeling

[111] 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, 9(3), 274-283.

[112] 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, 50(4), 425-435.

References 109

[113] 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., 9(1), 10-16.

[114] Bray, G. A., Lovejoy, J. C., Smith, S. R., Delany, J. P., Lefevre, M., Hwang, D., Ryan,

D. H., and York, D. A., (2002). The influence of different fats and fatty acids on obesity,

insulin resistance and inflammation. Journal of Nutrition, 132(9), 2488-2491.

[115] 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, 20(4),

260-274.

[116] sarvazyan, A. P., Skovoroda, A., and Emelianov, S, (1995). Biophysical bases of

elasticity imaging. Acoustic Imaging, 21223-240.

[117] 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, 40(7), 1443-1454.

[118] 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, 16(1), 45-70.

[119] 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, 42(2), 317-322.

[120] Tirrel, M., (1994). Rheology of Polymeric Liquids. In: Macosko, C. W. (eds), Rheology:

principles, measurements, and applications, Wiley-VCH, Inc, New york.

[121] 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

modeling59-69.

[122] 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, 42(3), 193-207.

[123] Verbeke, G. and Molenberghs, G., (2000). Linear Mixed Models for Longitudinal data, Springer, New York.

[124] Pipkin, A. C., (1986). Lectures on Viscoelasticity, Springer-Verlag, New York.

[125] Tanner, R. I., (2000). Engineering Rheology, Oxford University Press,

110 References

[126] Geerligs, M., Peters, G. W., Ackermans, P. A., Oomens, C. W., and Baaijens, F. P.,

(2008). Linear viscoelastic behavior of subcutaneous adipose tissue. Biorheology, 45(6),

677-688.

[127] 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., 41(2), 454-464.

[128] 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.

Technology and Health Care, 15(6), 385-398.

[129] 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., 7(2), 127-137.

[130] Mewis, J. and Wagner, N. J., (2009). Thixotropy. Advances in Colloid and Interface

Science, 147-48214-227.

[131] 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, 9(4), 222-228.

[132] Linder-Ganz, E. and Gefen, A., (2009). Stress analyses coupled with damage laws to

determine biomechanical risk factors for deep tissue injury during sitting. Journal of

Biomechanical Engineering, 131(1), 011003.

[133] Stekelenburg, A., Gawlitta, D., Bader, D. L., and Oomens, C. W., (2008). Deep tissue

injury: how deep is our understanding? Archives of Physical Medicine and

Rehabilitation, 89(7), 1410-1413.

[134] 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,

53(4), 663-670.

[135] OECD, (13-4-2009). OECD Guideline for the testing of chemicals. Skin absorption: in

vitro MethodTest Guideline 428

[136] 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, 44(1), 51-58.

[137] 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, 41(2), 422-

429.

References 111

[138] 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, 10(4), 231-

241.

[139] Bader, D. L. and Bowker, P., (1983). Mechanical characteristics of skin and underlying

tissues in vivo. Biomaterials, 4(4), 305-308.

[140] Bronneberg, D., (2007). Biochemical markers for early detection of superficial pressure

ulcers. Technische Universiteit Eindhoven.

[141] Cornelissen, L.H., (2008). Modeling the transport of biochemical markers in skin:

towards pressure ulcer assessment. Technische Universiteit Eindhoven.

[142] Gawlitta, D., (2006). Compression-induced factors influencing the damage of engineered

skeletal muscle. Technische Universiteit Eindhoven.

[143] Bosboom, E. M. H., (2001). Deformation as a trigger for pressure sore related muscle

damage. Technische Universiteit Eindhoven.

[144] Stekelenburg, A., (2005). Mechanisms associated with deep tissue injury induced by

sustained compressive loading. Technische Universiteit Eindhoven.

[145] Ceelen, K. K., (2008). Modeling the development of in vitro and in vivo pressure-induced

muscle damage, Technische Universiteit Eindhoven.

[146] Hendriks, R. and Lucassen, G., (2001). Two-photon fluorescence and confocal video

microscopy of in-vivo human skin4262287-293.

[147] Ericson, M. B., Simonsson, C., Guldbrand, S., Ljungblad, C., Paoli, J., and Smedh, M.,

(2008). Two-photon laser-scanning fluorescence microscopy applied for studies of

human skin. Journal of biophotonics, 1(4), 320-330.

[148] Branzan, A. L., Landthaler, M., and Szeimies, R. M., (2007). In vivo confocal scanning

laser microscopy in dermatology. Lasers in Medical Science, 22(2), 73-82.

[149] varez-Rom+ín, R., Naik, A., Kalia, Y. N., Fessi, H., and Guy, R. H., (2004).

Visualization of skin penetration using confocal laser scanning microscopy. European

Journal of Pharmaceutics and Biopharmaceutics, 58(2), 301-316.

[150] 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. Biophysical Journal, 93(3), 992-1007.

[151] Dullaert, K. and Mewis, J., (2005). A model system for thixotropy studies. Rheologica

Acta, 45(1), 23-32.

112 References

[152] Dullaert, K. and Mewis, J., (2005). Thixotropy: Build-up and breakdown curves during

flow. Journal of Rheology, 49(6), 1213-1230.

skin layer mechanics - materials technology - tu eindhoven · skin layer mechanics the human skin...

Documents