Biological fabrication: gradients produce failure resistant dental structures

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
BIOLOGICAL FABRICATION:
GRADIENTS PRODUCE FAILURE RESISTANT DENTAL
STRUCTURES

by

Bret Becker

A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
MASTER OF SCIENCE
(BIOCHEMISTRY AND MOLECULAR BIOLOGY)

December 2007

Copyright 2007 Bret Becker

ii

Acknowledgements

I would like to thank Dr. Malcolm L. Snead, my mentor and Committee Chair,
whose leadership and guidance have helped make this graduate program at
the University of Southern California a valuable learning experience. My
special thanks to Dr. Zoltan Tokes and Dr. Robert Stellwagen for their
continued interest and encouragement. Last but certainly not least I would like
to thank my parents for their constant guidance, support, and friendship.

iii

Table of Contents

Acknowledgements ii

List of Tables iv

List of Figures v

Abstract viii

Introduction 1

Chapter I: Enamel Versus Bone: Commonalities and Differences 2

Chapter II: Enamel Synthesis and Assembly 3

Chapter III: Amelogenesis Imperfecta 10

Chapter IV: Biomineralization at the DEJ 19

Chapter V: Functional Width of the DEJ 26

Chapter VI: Deformation Properties of Teeth 38

Chapter VII: Mechanical Gradient in Biology 48

Chapter VIII: Self-Healing Molecular Mechanisms 58

Conclusion 72

Bibliography 76

iv

List of Tables

Table 1: Hydrodynamic Radii (RH) of Different Nanosphere Components 17

Table 2: Comparison of Indentation Characteristics from Studies 36
Conducted across the DEJ

v
List of Figures

Figure 1: Schematic of the hierarchical organization of enamel 4
development

Figure 2: Enamel organization at the nanoscale and mesocale 6

Figure 3: TEM, SEM, and a cartoon images showing the enamel 9
organization for ‘wild-type’ mice.

Figure 4: Morbid anatomy of the mouse amelogenin M180 11

Figure 5: Incisors of wild type and null mice 13

Figure 6: Light microscopy of mandibular molars 14

Figure 7: Scanning electron microscopy of fractured incisors 15

Figure 8: Three-dimensional view of atomic force micrographs 18

Figure 9: Schematic representing the formation of the DEJ 21

Figure 10: SEM analysis of wild-type and myc-Dpp transgenic animals. 23

Figure 11: Incisor and molar crown anatomy for wild-type and myc-Dpp 25
transgenic animals.

Figure 12: Optical micrograph of the Vickers indents in the enamel near 28
(optical) DEJ in a human molar

Figure 13: Typical profiles of the Vickers hardness and indentation 29
toughness taken across the DEJ

Figure 14: SEM analysis showing examples of cracks from the enamel 31

Figure 15: SEM analysis showing examples of uncracked ligament bridging 32

Figure 16: AFM images of nanoscratches across dentin, enamel, & DEJ 34

Figure 17: Friction coefficients obtained by scratching across the DEJ 35

Figure 18: Schematic illustration of sample preparation and mounting 39

vi
Figure 19: Water immersed sample containing dentin, DEJ, & enamel layer 40

Figure 20: Edge view (side profile) of displacements 42

Figure 21: Pseudo-3D representation of the averaged displacement 43
maps of six tooth samples

Figure 22: Branching at ends of dentinal tubules near DEJ 44

Figure 23: Collagen fibril orientation at a distance of 50 µm from DEJ 46

Figure 24: Dentin, observed 50 µm beneath the DEJ 46

Figure 25: Radial stress, σr, in a butt joint made of two materials 49
having different stiffness

Figure 26: Schematic mussel on the half-shell with one byssal 51
thread showing the incremental steps in stiffness

Figure 27: Proposed structure of the preCOLs 53

Figure 28: Model of the hexagonal (6 + 1) bundles of bent-core trimers 54

Figure 29: Distribution of different preCOLs along a byssal thread 56

Figure 30: Fracture surface of human bone showing mineralized 60
collagen fibrils

Figure 31: Sacrificial bond cartoon and analysis graphs 64

Figure 32: Suspected calcium-mediated sacrificial bonds in bone 65

Figure 33: Schematic drawing of the basic principle of the sacrificial 67
bond-hidden length mechanism

Figure 34: Various sacrificial bonds on single strand connections 69

vii

Abstract

Unlike many aspects of the body, human teeth do not have natural regenerative
mechanisms. Instead, the multiple tissues and mechanical structures in teeth
combine to create a structural gradient that successfully absorbs a lifetime of
impact forces from routine mastication without irreversible damage. The
integration of hard and soft tissue types is advantageous for distributing loads
but the resulting interfaces are prone to fracture. The intermediate boundary
known as the dentino-enamel junction (DEJ) is a natural strategy which avoids
abrupt changes in mechanical properties. The DEJ is a self-assembling
structure which combines two dissimilar materials, enamel, and dentin. This
junction is critical to the longevity of tooth function and is recognized to fulfill
multiple roles. This thesis reviews the current state of knowledge related to the
DEJ and its surrounding structures that cooperate together to create a failure
resistant structure.

1

Introduction

Unlike many aspects of the body, human teeth do not have natural regenerative
or repair mechanisms. Instead the body relies on the mechanical structure and
composition of teeth to create a structural gradient that absorbs a lifetime of
impact forces from routine mastication without irreversible damage. Teeth are
composed from a variety of structures and tissue types that have dissimilar
mechanical properties. The integration of different hard tissue types is
advantageous for distributing impact loads but the interfaces between the
materials are prone to fracture. The intermediate boundary known as the
dentino-enamel junction (DEJ) is a natural strategy that avoids abrupt changes
in mechanical properties. The complex architecture of teeth prevents these
non-uniform tissues and junctions from experiencing structural failures so that
loads from mastication can be distributed through to the next layer.

Enamel, the outer covering of teeth is a unique biomineralized tissue in the
mammalian body that has remarkable assembly and structural characteristics.
Due to its high content of biologically formed hydroxyapatite crystallites, enamel
is the hardest and most mineralized tissue in the body. Unlike other
biomineralized structures, enamel is designed to withstand a lifetime of impacts
without provisions for repair. As a result, great emphasis is placed on the initial
assembly of this critical structure.

2

Chapter I: Enamel versus Bone: Commonalities and Differences

Enamel is the hardest material in the human body and makes up the outer most
layer of teeth. Enamel is a distinctive biomineralized tissue that is unique in its
development and ultimately its structural properties. Despite possessing similar
chemical properties for their mineral phase, bone and enamel are
fundamentally different from one another. To begin with, both tissues originate
from their distinct germ layers. Bone is derived from mesenchymal tissue while
enamel is ectodermal in origin. Importantly, bone contains living osteoblast
cells that are responsible for bone formation. Osteoblasts are embedded in a
mineralized organic matrix and these living cells enable bone to normally
undergo a process of matrix removal, remodeling, and repair. In contrast, the
ectodermal ameloblast cells that create the enamel tissue are lost at tooth
eruption and thus eliminate any chance for remodeling or repair. The assembly
of the inorganic enamel matrix from the organic precursor is a highly regulated
process. The finalized enamel structure contributes to the important function of
breaking down food for digestion. Great emphasis and energy is therefore
placed on the complex, assembly of the enamel tissue.

3

Chapter II: Enamel Synthesis and Assembly

The self-assembly of the enamel matrix occurs on several size scales including
nanoscale, mesoscale, and macroscale. At the nanoscale, enamel matrix
forming cells release proteins which interact with other enamel matrix proteins
and with adjacent hydroxyapatite crystallites, guiding the growth habit of the
mineral phase (figure 1) (Paine, White et al. 2001). During this early phase of
enamel development at the nanoscale level, developmental control over the
genes expressed and their quantities contribute to the protein-protein
interactions that guide self assembly and contribute to the formation of the
organic enamel matrix (Yamada and Kleinman 1992; Miner, Cunningham et al.
1998; Paine, White et al. 2001). Two predominant proteins found in tooth
enamel include the amelogenins and the enamelins (Paine, White et al. 2001).
Both amelogenin and enamelin proteins are expressed by inner enamel organ
epithelial cells and are secreted into the developing enamel extracellular matrix
(Paine, Luo et al. 2005). Amelogenin protein is the dominant gene expressed
by enamel cells. Amelogenin undergoes self-assembly via protein-protein
interactions at the nanoscale with the formation of 10-20 nm nanospheres
(Fincham, Leung et al. 1998). These amelogenin nanospheres interact with
hydroxyapatite (HAP) crystals and orient the crystal growth in a direction
perpendicular to the surface. It is these nanoscale protein-crystallite
interactions that guide the formation of long thin hydroxyapatite (HAP)
crystallites (Moradian-Oldak 2001). The HAP elongating crystallite chains
ultimately arrive at the mesoscale by bundling together in the formation of rods
(figure 2). Each rod can be thought of as a remnant which defines the
secretory path of a single ameloblast cell synthesizing of the enamel organic
matrix.

Fig. 1. Schematic of the hierarchical organization of enamel development,
from the initial developmentally-regulated timing of the enamel matrix
proteins through finishing at the organization of the mature tooth. (Paine,
White et al. 2001)

Enamel mesoscale structures are at the scale of cells and are built from a
combination of nanoscale structures. Two highly organized enamel mesoscale
structures include rods and interod bundles of HAP crystallites (figure 1). At the
macroscale, cellular features combine to create a structurally significant enamel
tissue which functions during mastication to withstands a lifetime of use and
abuse. The ectodermally derived ameloblast cells regulate these processes at
the mesoscale level. Behaviors that occur at the mesoscale include; secretory
events, interactions between ameloblast cells, ameloblast interactions with the

4

5
extracellular matrix, and ameloblast migration toward the surface of the tooth
(figure 1)(Paine, White et al. 2001). These mesoscale events combine and
result in the important ribbon pattern of self-assembling enamel matrix proteins.
Rods organized by the ameloblasts as collections of HAP crystallites are woven
together at 60 degree angles and are bound in a third dimension normal to the
rods with additional HAP crystallites organized as the interrods (figure 2)(Paine,
White et al. 2001). “This coalescence of interrod and rod permits the formation
of a tough fibrous continuum that has the ability to undergo slippage and
damage limitation by crack deflection” (White, Luo et al. 2001).

Fig. 2. Enamel organization at the nanoscale and mesocale reflect the
creation of ribbons (rods) of self-assembled enamel proteins secreted by
ameloblasts. Panel (a) depicts human enamel seen by the scanning
electron microscope along the cross-sectional (outer facial surface) and
long-sectional (sagittal) views of rod (R) and interred (IR). Rods are seen
to be largely surrounded by distinct interrod, but the apical surface of
each rod blends into interrod to form a continuum, producing the ‘fish
scale’ appearance. The presence of interrod crystallites running at
approximately 60˚ to the long axis of the rods is especially evident,
because the rod crystallites have been preferentially etched away (etched
rod). Panel (b) is a schema of the proposed enamel hierarchical structure
model that includes several important points. First, the paths of the
Tomes’ processes (TP) create a staggering between adjacent layers of
rod, governing the possible spatial and temporal relationships among
forming rods, and their possible connections as the ameloblast migrate
toward the surface of the tooth. (Paine, White et al. 2001)

6

7
Ameloblast cells undergo several identifiable steps during their differentiation,
including presecretory, secretory, and maturation. In the presecretory stage,
the basement membrane separates the dental epithielium from the
mesenchymal preodontoblasts (Thesleff and Hurmerinta 1981; Adams and
Watt 1993). The breakdown of the basement membrane is a critical step which
enables odontoblasts to begin the continued secretion of the dentin matrix.
The secretory stage of the ameloblast cells starts when the cells contact the
dentin matrix. Here the ameloblasts produce and secrete proteins which
contribute to the formation of the organic enamel matrix. Interestingly, the
secretion and assembly of the dentin-enamel organic matrix takes place in the
space that previously corresponded to the position of the basement membrane.
This anatomic orientation is unlike other secretory epithelia which usually
contact the basement membrane and secrete away from this source of polarity.
During the ameloblast maturation stage, the organic protein matrix molecules
are systematically replaced with the final inorganic phase formed from by
mineralized hydroxyapatite (HAP) crystallites.

Ameloblast cells secrete amelogenin, the most abundant enamel matrix forming
protein. Gene replication is responsible for regulating the production of
ameloblast cells. Although indirect, gene replication ultimately influences the
amount of amelogenin secretion during tooth development and enamel
maturation. At the nanoscale level, amelogenin molecules are secreted into
the extracellular space and self-assemble to form 15-20nm spheres.

8
The precise function of the amelogenin proteins during enamel formation is not
well defined however, these proteins are known to account for 90% of the
organic matrix remnant found in enamel (Gibson, Yuan et al. 2001). The
organic enamel matrix composed of secreted proteins is characterized by a
high degree of organization achieved by self-assembly. This ordered structure
which precedes the inorganic enamel phase suggests that the organic
extracellular matrix proteins secreted during the early stages of amelogenesis
must regulate the organization and crystal habit of the HAP (figure 3)(Paine and
Snead 1997).

Fig. 3. Transmission electron microscopy (TEM, panel A), scanning
electron microscopy (SEM, panel B). images and a cartoon (panel C)
showing the enamel organization for ‘wild-type’ mice. The cartoon is
inspired by the accompanying TEM and SEM images. The cartoon
illustrates how the amelogenin nanospheres may influence enamel
formation both at the nanoscale, and at the mesoscale levels. The cartoon
in panel C thus depicts wild-type amelogenin self-assembling to form
nanospheres. The nanospheres align as they interact (double-headed
Arrows) with the retracting Tomes’ processes of the ameloblasts through
the A-domain (represented by a serrated nanosphere periphery) which
includes a lectin-like binding avidity. The intact B-domain (represented by
a cross inside the nanosphere) stabilizes nanosphereassemblies for
subsequent proteolytic processing by enamelysin(see text). The physical
properties of organizing nanospheres participate inpreserving rod-
interrod boundaries at the mesoscale level. The organization of the
resulting wild-type enamel is depicted in the SEM image (panel B). Scale
bars in nanometers (nm, panel A) and micrometers (µM, panels B and C)
are included. (Paine, White et al. 2001)

9

10
Chapter III: Amelogenesis Imperfecta

Amelogenesis imperfecta is an X-linked genetic disorder which has provided
evidence for the critical role played by the protein amelogenin in the formation
of the enamel matrix. Patients with this genetic disorder have a defect located
on the gene coding for amelogenin. This error causes various degrees of
reduced expression of amelogenin. In 1985, Snead and colleagues identified
two highly conserved coding regions on the M180 amelogenin gene (Paine and
Snead 1997). These two regions were located at the amino-terminal residues
1-45 and on the carboxy-terminal residues 157-173. Subsequent research
focusing on the mouse amelogenin M180 protein found the amino terminal-
residues 1-42 and carboxy terminal-residues 157-173 to be self-assembling
(Paine and Snead 1997). These highly conserved and self-assembling regions
are respectively referred to as domain “A”, and domain “B”. The missence or
point mutations causing amelogenesis imperfecta was found to occur on amino
terminal 22 and 41 of domain “A” (Lench and Winter 1995; Collier, Sauk et al.
1997). The highly conserved nature of these regions suggests physiological
importance.

Fig. 4. Morbid anatomy of the mouse amelogenin M180, including the
signal peptide. Self-assembly domains (A) (solid line) and (B) (broken
line) are underlined. Intron exon boundaries are identified by an arrow,
and exon domains numbered. The amino-acids mutated in amelogenesis
imperfecta phenotypes are identified by a box. The carboxyl-terminal
tryptophan (Trp), of the tyrosine-rich amelogenin polypeptide (TRAP), is
identified by an arrow. (Paine and Snead 1997)

11

12
Patients with amelogenesis imperfecta have phenotypic abnormalities in the
enamel. These include hypoplastic defects (thin pitted or grooved) and also
hypomineralization resulting in decreased mineral content (Gibson, Yuan et al.
2001). Studies have found correlations between these variations and the sites
of different mutations located on the amelogenin gene. These observations
suggest that the location of the mutated protein domain ultimately determines
which defects will be present in the developed enamel layer (Paine and Snead
1997).

Fig 5. Incisors of wild type and null mice. Photograph of incisor
teeth from wild type (A) and null mice (B). Scanning electron micrograph
showing a smooth enamel surface for wild type (C) and marked
enamel hypoplasia characterized by a furrowing of the enamel in the
incisor from a null mouse (D). (Gibson, Yuan et al. 2001)

Mice which completely lack amelogenin were found to exhibit hypoplastic
defects in this animal but the elemental mineral composition was consistent
with normal teeth (Gibson, Yuan et al. 2001).

13
This observation demonstrates that mineral crystal initiation is not dependent
on amelogenin production. Instead, it is the architecture of these minerals, a
prism pattern whose formation is guided by amelogenin protein. These
discoveries also infer that the prism pattern is a critical component for
producing an enamel layer of proper thickness (figure 6 & 7)(Gibson, Yuan et
al. 2001). These findings are further supported by Aoba and colleagues (Aoba,
Moreno et al. 1989) who demonstrated that the carboxy-terminal of amelogenin
binds hydroxyapatite crystals. This discovery is a likely mechanism which
explains how amelogenin facilitates proper orientation along the crystallites.

Fig. 6. Light microscopy of mineralized thin sections through
mandibular molars from 16-week-old-mice. A, wild type dentition
shows a normal and relatively even enamel layer over the tooth crown. B,
dentition from the null mouse has a markedly reduced enamel thickness.
Paired arrowheads indicate thickness of the enamel layer. (Gibson, Yuan
et al. 2001)

14

Fig. 7. Scanning electron microscopy of fractured incisors from 16-week-
old wild type and null mice. The enamel (E) and junction with dentin (D)
are shown. A, wild type mouse. B, the enamel from the null mouse does
not have a normal prismatic structure and is markedly reduced in
thickness compared with that of the wild type mouse shown at the same
magnification as A. C, higher magnification of the enamel layer from the
null mouse. Arrowheads indicate enamel thickness. Bars in A and B = 10
µm; bar in C = 1 µm (Gibson, Yuan et al. 2001)

To further differentiate between the functions of the two highly conserved
domains, (Paine and Snead 1997) created recombinant amelogenin proteins.
Bacteria were used to express an engineered amelogenin cDNA which lacked
domain “A” or domain “B”. This recombinant protein was tested with dynamic
light scattering and atomic force spectroscopy (Moradian-Oldak, Paine et al.
2000). Analysis of the engineered proteins confirmed the prediction that the
sizes of the nanospheres were affected by the loss of either domain from the
recombinant amelogenin protein (table 1 & figure 8).

Subsequent experiments using transgenic mice revealed that the organic
matrix failed to effectively direct the mineral replacement beginning at the
nanoscale level (Paine, White et al. 2001). Results of this study indicated that
domains “A” and “B” are required for directing the self-assembly of

15

16
macromolecular structures associated with amelogenin (Paine and Snead
1997).
Additionally, domain “A” was shown to possess a self-assembly property even
in isolation. Domain “B” however did not maintain its assembly characteristics
in isolation.

Protein
Radius
(nm)
%
Scatt
er
%
Ma
ss

rM179 10.1 9 30

14.1 40.2 50.4

19.8 39.3 17.7

27.8 11.5 1.9
Rp(H)M180 14.1 8 22.4

19.8 68.8 69.2

27.8 23.2 8.4
rp(H)M180DA 2.7 4.7 77.6

3.7 3.4 20.1

5.2 0.0 0.0

14.0 8.1 0.9

19.8 21.9 0.9

27.8 27.2 0.4

38.8 22.7 0.1

54.3 12.0 0.0
rp(H)M180DB 27.8 31.0 64.6

38.8 36.3 27.7

54.3 24.8 6.9

76.0 7.9 0.8
rp(H)M180T21 –I 19.8 25.1 61.2

27.8 32.1 28.3

38.9 26.90 8.7

54.4 15 1.8

76.1 1 0
rp(H)M180P41 –
T
27.8 20.1 57.2

38.9 28.4 29.7

54.5 26.7 10.2

76.2 18.1 2.5
rp(H)M180TI,PT 27.8 16.6 47.9

38.8 35.5 37.6

54.3 31.6 12.2

76.0 16.4 2.3

TABLE I
Hydrodynamic Radii (RH) of Different Nanosphere Components as Estimated
from the Measurements Presented in Table I, Only Using a “Regularization
Data Log” (a Multicomponent Fitting System Calculated by
Dynamics)(Moradian-Oldak, Paine et al. 2000)

17

Fig. 8. Three-dimensional view of atomic force micrographs (500 by 500
nm field) in tapping mode in air of the recombinant histidine-tagged
amelogenins adsorbed onto a mica surface. (A) rp(H)M180; (B)
rp(H)M180DA; (C) rp(H)M180DB; (D) rp(H)M180T21 3 I; (E) rp(H)M180P41 3
T; (F) rp(H)M180 T21 3 I,P41 3 T. Note that the height scale is 10 nm for A
and B and 25 nm for C, D, E,and F. (Moradian-Oldak, Paine et al. 2000)

18

19

Chapter IV: Biomineralization at the DEJ

In many aspects the dentin enamel junction (DEJ) is a unique biological layer.
Its unordinary structure translates into mechanical properties which
successfully transmits impact loads from the hard and wear resistant enamel
layer to a softer underlying dentin layer without failure. In addition, cracks that
do occur in the enamel are often arrested or limited by the DEJ. Cracks are
therefore usually prevented from extending into the dentin, a property that
provides toughness to the interface. The structural properties used to
distinguish dentin from enamel can be attributed in large part to the varying
percent composition of mineral hydroxyapatite (HAP) crystallites found within
the final tissue. The hard outer enamel contains more than 95% mineral
crystals along with a minute percentage of protein remnants (Fincham, Belcourt
et al. 1982). In comparison, the dentin layer contains approximately 70%
mineral crystals (Kim, Simmer et al. 2006). As these two dissimilar materials
approach the DEJ interface both their compositions and physical properties
transition so that the materials are more similar when they intersect at the DEJ.
The DEJ is a complex boundary created from proteins derived from both the
dentin and enamel layer.

20
Current research has suggested that the DEJ survives as a precursor of
proteins that existed in a transient state between dissimilar germ layers
(figure 9). A transient protein state affords an opportunity for the mixing of
protein materials from either germ layer component (Paine, Luo et al. 2005).
This pattern of formation provides the capacity to impart unique structural
characteristics to bonds between tissues and across germ layers (Simmer and
Fincham 1995). Normally it would be expected that the interface joining two
dissimilar materials would be prone to delamination. In this case however, the
tissues are intimately connected. Despite this connection the hard outer
enamel does not transmit mechanical stresses to the underlying dentin. Rather
than promoting a delaminating structure, the DEJ functions to absorb stresses
that might otherwise cause delamination. The DEJ produces a unique, crack
resistant mechanism that allows a hard mineralized tissue like enamel to exist
over a softer, more compressible material like dentin.

Fig. 9. Schematic representing the formation of the specialized zone of
enamel and dentin that contributes to the dentino-enamel junction. The
expression by ameloblasts of proteins commonly thought of as dentinal
in origin, namely Dsp and Dpp, and the simultaneous expression by
odontoblasts of proteins commonly thought of as enamel in origin,
namely ameloblastin (Ambn) and enamelin (Enam), participate in the
formation of the specialized enamel and dentine approximating the DEJ
by creating a unique blending of extracellular matrix molecules in this
zone (panel A). These matrix molecules direct the properties of the
mineral phase and contribute to the unusual material properties of the
DEJ that provides a link between a hard, wear-resistant enamel with a
softer, but tough underlying dentine layer (panel B). Other abbreviations
used are: dentin, D; mantle dentin, MD; specialized aprismatic enamel,
SAE; prismatic enamel, PE; and amelogenin, Amel. (Paine, Luo et al.
2005)

Portions of the DEJ contribute cells that originate from both the dentin and
enamel layer. Dentin found in teeth is produced mainly by odontoblast cells
while ameloblast cells are largely responsible for the formation of the enamel
layer. When it comes to the formation of the DEJ, it is interesting to note that
some overlap in protein production has been found. For example, ameloblasts
participate in making a portion of the DEJ by contributing proteins that are

21

22
normally expressed by odontoblasts (Paine, Luo et al. 2005). Similarly,
ameloblasts transiently produce proteins at the DEJ that are normally
synthesized by odontoblast cells in all other developmental periods.

One of the most important proteins that both odontoblasts and ameloblasts
express during the formation of the DEJ is dentin sialophosphoprotein (DSPP).
The DSPP protein has been shown to influence the DEJ structure and
properties, potentially uniting these dissimilar tissues (figure 10) (Paine, Luo et
al. 2005). The sharing of proteins is thought to occur between the enamel and
the dentinal segments of the DEJ. This transient sharing of DSPP during the
formation of the DEJ has been suggested to contribute to the unique properties
of the DEJ. One can postulate that the non-covalent interactions between
DSPP and collagen could functionally provide an additional pathway for
toughness at the DEJ (Paine, Luo et al. 2005).

Fig. 10. SEM analysis of wild-type and myc-Dpp transgenic animals.
Showing is a fracture line through the transition zone of incisor teeth
from wild-type (panels A and B), myc-Dpp line 53 (panels C–E), and myc-
DPP line 47 (panels F and G). Panels A, C, and F show surface enamel
(top), and the entire enamel thickness, through the transition out zone,
with dentin seen in the bottom of each. Panel B shows wild-type enamel
prismatic structure close to and contacting the DEJ. Panel D shows
enamel close to the surface in the myc-Dpp line 53 animal, and panel E
shows enamel close to and contacting the DEJ in this same animal. Panel
G shows enamel close to the DEJ in the myc-Dpp line 47 transgenic
animal. Panels A, C, and F are 500 magnification with the scale bar in
panel A to 50 µm; panels B, D, E, and G are 2,000 magnification with the
scale bar in panel B at 20 µm. (Paine, Luo et al. 2005)

23

24
The failure of these non-covalent bonds would likely provide stress relief to the
covalently bonded framework and contribute to crack-deflection during periods
of mechanical stress. This proposed mechanism at the DEJ interface would
absorb a large amount of energy prior to catastrophic failure and irreversible
damage.

DSPP generates two distinct extracellular proteins: dentin sialoprotein (DSP)
and dentin phosphoprotein (DPP). These two proteins are primarily expressed
in the dentin matrix. More than 90% of the dentin matrix is collagen! DPP is a
highly acidic protein which makes up a large portion of the remaining
noncollagenous component of the dentin matrix. DSP is a glycoprotein found in
small amounts. Paine and colleagues (Paine, Luo et al. 2005) studied the
functional effects of DSP and DPP on enamel formation by creating transgenic
mice which over expressed either DSP or DPP under the control of the
amelogenin promoter. It was hypothesized that over expression of these
proteins would result in differing physical and functional properties when
compared to a wildtype mouse. The teeth of the transgenic mice proved to be
distinct from that of the wildtype mouse. Over expression of DPP in mice
resulted in an enamel layer which was pitted, chalky, and of non-uniform
thickness. The damaged enamel layer due to high levels of DPP was found to
be increasingly prone to wear (figure 11)(Paine, Luo et al. 2005). The severity
of this defect was also transgenic dose dependant. In contrast, animals over
expressing DSP in enamel resulted in an elevated rate of mineralization. From
these observations DSP mice may possess superior hardness properties of the
final bulk enamel. “These results support the notion that the dentin proteins
expressed by presecretory ameloblasts contribute to the unique properties of
the DEJ.” (Paine, Luo et al. 2005)

Fig. 11. Incisor and molar crown anatomy for wild-type and myc-Dpp
transgenic animals. Six-week-old wild-type (panels A and B) or myc-Dpp
line 47 transgenic animal (panels C and D) mandibular molar (top) and
mandibular incisor (bottom) teeth are shown. Molar teeth are viewed from
lingual, and incisor teeth are viewed from lateral. Severe enamel pitting is
clearly evident for all transgenic molars (panel C), whereas for the incisor
teeth of this transgenic animal line, the bulk of enamel has fractured away
from the underlying dentin.(Paine, Luo et al. 2005)

25

26

Chapter V: Functional width of DEJ

Originally the DEJ was thought to be a sharp boundary separating two different
germ layers, an image conveyed by its optical properties. The DEJ is
comprised of secreted proteins from both dentin producing dentinoblasts and
enamel producing ameloblasts. This mixing drives a biomineralization process
that results in a DEJ tissue possessing tougher and harder structural
characteristics than the corresponding parent tissues. This mechanism
enables stresses imparted on the outermost enamel layer of teeth to be
displaced over a greater distance thus reducing the overall strain to any one
specific layer. The entire tooth can be thought of as an energy absorbing
mechanical gradient. The DEJ is an essential component of this gradient which
transmits loads not absorbed by the previous layer to next in order to preserve
the overall structure. The toughness property is a means by which a great
deal of damage can be absorbed at the interface prior to failure. This allows
the enamel and dentin layers to work as a unit without failure thus enabling
teeth to last a lifetime.

Several independent studies have focused on the structural and mechanical
properties of the DEJ. It is fascinating to note that these particular studies of
the DEJ have required a combination of skills from several disciplines including

27
material sciences, engineering, restorative dental sciences, structural biology,
and physics. Some studies required the development of novel instruments and
techniques to gain a greater understanding for the complex nature of the DEJ.
These techniques include microscale and nanoscale indentation,
nanoscratching with atomic force microscopy (AFM), and backscatter radiation.

Imbeni and colleagues studied the hardness, toughness, and crack propagation
properties at the DEJ by creating numerous micro-indentations at varied
positions with respect to the DEJ (figure 12). The forces required to create
these indentations and the resulting penetration depth were measured. This
information was then used to create a Vickers hardness and indentation
toughness profile of the DEJ (figure 13)(Imbeni, Kruzic et al. 2005). The profile
indicates that the mechanical hardness property of enamel falls quite rapidly
within a millimeter of the microscopically observable sharp interface of the DEJ.
Additionally, the mechanical properties of enamel vary over a distance of
several hundred micrometers across the interface with the enamel toughness
rising dramatically over the final 500 µm into the interface.

Fig. 12. Optical micrograph of the placement of Vickers indents in the
enamel within~20-50 µm from the (optical) DEJ in a human molar, which
were used to create cracks that impact upon the DEJ. Inset shows optical
micrograph with Nomaski interference contrast of one such indentation
with cracks, which emanate from the intent corners, propagating into the
scalloped interface. (Imbeni, Kruzic et al. 2005)

28

Fig 13. Typical profiles of the Vickers hardness and indentation
toughness taken normal to, and across, the DEJ from the enamel to the
dentin in a human molar. Hardness indentations were made with a load
range between 3 and 5 N to minimize brittle fracture damage but to still
form cracks around the indents to enable toughness measurements.
Lines of indents were performed on three different teeth (each from a
unique patient), with three series for each tooth. The indentation
toughness, Kind,c, was determined from the indentation load P, and the
average crack lengths, c, emanating fromthe indent corners, according to
Kind,c = χP/c3/2, where χ is the residual indentation coefficient (taken as
0.076 for enamel). Such measurements could only be made in the enamel
as inelasticity in the dentin suppresses the formation of indent cracks.
These profiles show that cracks in the enamel experience a region of
decreasing hardness yet increasing toughness as they approach the DEJ.
(Imbeni, Kruzic et al. 2005)

The physical cracks created by the nano-indentations were studied using
scanning electron microscopy (SEM). 172 indentation cracks were examined
and it was found that more than 75% of these cracks actually penetrated the
DEJ a short distance before being arrested 10µm or less into the mantle dentin.

29

30
Beyond the crack tip numerous uncracked-ligament ‘bridges’ are observable.
These uncracked ligaments span the crack tip and contribute to the prevention
of further crack propogation (figure 14). This bridging of ligaments has been
recognized as a toughening mechanism in structural materials. The bridging
effectively deflects the energy contributing to the crack-driving force. Bridging
of ligaments may ultimately prove to be the mechanism responsible for the low
incidence of delamination along the interface of the DEJ.

31
Fig 14 & 15. Scanning electron micrographs (taken using a conventional
SEM) showing examples of cracks from the enamel which are normally
incident on the DEJ and are arrested after propagating less than ~10 µm
beyond the interface into the mantle dentin. Behind the arrested crack tip,
numerous uncracked-ligament “bridges” can be seen; these are regions
of uncracked material that oppose the opening of the crack and sustain
load that would otherwise be used for crack growth. Such bridging, which
is a form of crack-tip shielding and is prominent toughening mechanism
in dentin and bone acts to reduce the effective driving force for crack
extension, thereby arresting the crack. Cracking can also be seen near,
and nominally parallel, to the DEJ. However, by comparing these images
with corresponding images in the environmental SEM (at 8 torr water
pressure), such “delamination” cracking was found to be an artifact
caused by dehydration in vacuo in the conventional SEM.(Imbeni, Kruzic
et al. 2005)

32

33

In another study, Habelitz and colleagues performed nanoscratching
experiments to characterize the functional width of the DEJ. Scratch testing is
a widely used technique for characterizing the mechanical resistance of the
coating substrate interface for large scale applications. “The development of
nanoindentation instruments enabled measurements of mechanical properties
of layers and structural entities on the nanometer scale”(Habelitz, Marshall et
al. 2001). The nanoscratch test was thought to be a superior technique when
compared to the nano-indentation technique used by Imbeni. Unlike the
incremental indentation technique, the nanoscratching technique can record a
continuous measurement of frictional properties.
This continuous method leads to improved resolution of the measurement. The
nanoscratching instrument consists of a nanoscratch tester attached to an
atomic force microscope (AFM). This instrument measures the frictional
coefficient while generating small scratches across the DEJ interface. These
frictional coefficients reflect the hardness and elastic modulus properties of the
material it crosses. Accordingly, this instrument is able to measure where one
material ends and another begins.

During this study it was found that differences in elastic modulus caused the
dentin to deform more under an applied pressure when compared to enamel.
The average frictional coefficient for dentin and enamel was determined to be
0.31 and 0.14 respectively at loads of 50-300µm for dentin and 50-600µm for
enamel (figure 16). The increased coefficient for dentin is attributed to the
higher content of organic phases (Habelitz, Marshall et al. 2001). A very small
transition zone of frictional coefficients was also measured. The transition zone
corresponding to the DEJ was measured to be 2µm. This value is much lower
than the measurement taken from the nano-indentation studies.

Fig. 16. AFM images of nanoscratches across (a) intertubular dentin, (b)
enamel rods, and (c) dentino-enamel junction of human third molars.
Numbers indicate the load applied in micronewtons. The color schemes
show the z range of the images, with brighter colors indicating increased
topographic height in a linear scale. (Habelitz, Marshall et al. 2001)

34

Fig 17. Analysis of a course of friction coefficients obtained by scratching
across the DEJ at a normal load of 300 mN. The width of the slope relates
to the width of the DEJ. (Habelitz, Marshall et al. 2001)

Researchers have identified that the structural and mechanical properties of the
DEJ transition over a small distance that provides additional toughness.
Between the distinct enamel and dentin layers the DEJ promotes bonding,
stress distribution, and fracture prevention. Interestingly, several studies have
concluded that the functional width of the DEJ to varies from 1-100um (table 2).
This relatively large discrepancy has been attributed to the size, force, angle,
and depth of the scratching instruments used in each experiment. In addition,
the scalloped pattern of the DEJ creates challenges for researchers when trying
to maintain a consistent scratch orientation. The changing properties of enamel
and dentin as they approach the functional width of the DEJ are further
evidence for a structural motif designed to cushion impacts.

35
(Habelitz, Marshall et al. 2001)
TABLE II
Comparison of Indentation Characteristics from Studies Conducted across the DEJ
Indentation
depth
Indentation
size

Type of indenter Load on human
dentin
At surface Step
size
Width
DEJ
Vickers
microindentationa
147
µN 3.4 µm 17 µm 50 µm
27–
100µm
Berkovich
nanoindentationb
1 µN 100 nm 850 nm .2 µm
15–25
µm
Cube corner
nanoindentationc
500
µN
300 nm 800 nm 1–2 µm
10–13
µm
Spherical
nanoscratchtesterd
300
µN
25 nm 1.4 µm 0 µm 1–3 µm
Note. Dimensions refer to indentations on
the dentin side.

a White et al. (2000).
b Fong et al. (2000).
c Marshall et al. (2001).
d This study.

Recent studies rely on mature teeth for analysis. A mature tooth can be
thought of as a fossilized record. The final crystal organization, mineral
composition, types of defects, and geometry interfaces are all initially controlled
by cellular and by biological processes. Proteins secreted by ameloblasts and
dentinoblasts ultimately influence formation during the mineral phase. Much of
the current knowledge has been revealed through reverse analysis. A greater
understanding of these developmental processes would provide further
explanations for the formation of dentin, enamel, and the bordering DEJ. The
development of novel scientific techniques will answer questions that could not
previously be answered by analysis of fully formed tissue alone.

36

37
The DEJ is now known to be a transitioning boundary of material and
mechanical properties. The advent of improved instrumentation and novel
techniques has furthered our understanding of this complex layer. Aspects of
the DEJ layer including mechanical hardness and crack propogation properties
have been a focal point of this research. Modern instruments like
nanoscratching, nanoindentation, SEM, and AFM have provided details that
help explain how the DEJ avoids delamination. Future studies using
instrumentation of improved sensitivity will be useful in refining our
comprehension of the molecular interactions occurring at this complex and
essential junction.

38

Chapter VI: Deformation properties of teeth

The teeth of most vertebrates are composed of a thin yet hard outer enamel
layer that surrounds and protects a tough, dentin inner layer. The material
composition of these distinct layers explains the differing structural
characteristics. Enamel is composed of more than 95% Hap mineral crystal
while dentin is composed of 50% mineral, 30% organic matrix, and 20% water
(Zaslansky, Friesem et al. 2006). When overstressed, the brittle enamel layer
fractures in response to strain. Unlike enamel however, the tougher dentin
layer is able to compress. This compression response enables dentin to relieve
strain without irreversible structural deformation. These two functionally
different materials must cooperate together to resist mechanical,and thermal
stresses on a daily basis.

Current research of the DEJ seems to support the findings of Wang and Weiner
who described the DEJ zone as a functioning cushion or gasket that enables
two structurally distinct layers to work in unison (Zaslansky, Friesem et al.
2006). It has been observed that the hardness of the DEJ drops sharply away
from the external enamel surface but then recovers in a 200-300 µm zone
beyond the DEJ in the dentin (Zaslansky, Friesem et al. 2006). To gain further
understanding of the interface it has been necessary to study and measure the
distortion of these materials under compression.

Fig. 18. Schematic illustration of sample preparation and mounting. (A) A
representative sketch of a slice through the lingual–buccal orientation of
an upper 1st human premolar tooth (root removed for convenience).
Lingual (L) and buccal (B) samples were cut from each of three teeth from
within regions approximately outlined by the dark lines on both sides. All
samples contained both enamel and dentin with the DEJ oriented
approximately perpendicular to the long sample axis. (B) Each sample
was mounted for testing in an upright orientation with the enamel and
dentin loaded in series. The samples were held in place by embedding the
lower and upper edges in a soft dental polymer light-cured composite
(bottom) which was cured after positioning.(Zaslansky, Friesem et al.
2006)

39

Fig. 19. Water immersed mounted sample. The enamel, DEJ, and dentin
were optically visible on each of the samples. The sample is held in
between compression anvils by fully cured composite layers. The
approximate position of the DEJ is indicated by the arrow. (Zaslansky,
Friesem et al. 2006)

Simulating and measuring compressive stresses on teeth is both difficult to
perform and interpret. Zalansky and colleagues (Zaslansky, Currey et al. 2005)
have developed a novel method for replicating and measuring compressive
stresses on teeth. This method combines electronic speckle pattern-correlation
interferometry (ESPI) with a mechanical compression apparatus. The
resolution of this tool exceeds the results previously obtained by Wang and
Weiner in “Enamel Matrix Protein Interactions” (1998). All tooth samples
consisted of an enamel, DEJ, and dentin portion (figure 18).
These samples were positioned between two adjustable aluminum anvils which

40

41
provided the highly precise compressive force (figure 19). A laser light was
required to illuminate the sample’s surface. The ESPI apparatus optically
measures the scattering of the laser light. This scattering of light occured when
the surface of the samples became displaced due to the compression loads of
the anvil. Six samples were tested and the data recorded by the ESPI
apparatus of the entire sample’s surface was used to create displacement
maps.

The maps created from ESPI data reflect the differing gradients of relative
displacement for the enamel, DEJ, and dentin regions of the sample. The
gradient is the result of a displacement or movement in the sample’s surface.
Since stiff materials are less likely to deform or become displaced, the stiffness
and displacement gradient are inversely proportional. Of the three regions, the
DEJ possessed the greatest deflection volume. The relative displacement of
the DEJ region increased as it traveled away from the top of the sample (figure
19). The distance of the upward slope in figure 20 suggests the DEJ region to
be 200-300um thick (Zaslansky, Friesem et al. 2006). The enamel portion of
this map is shown as a relatively flat segment (low gradient) and its value of
relative displacement is low when compared with the DEJ and the dentin
segments. The large gradient which characterizes the DEJ indicates this zone
to be less stiff when compared to the enamel and dentin regions.
Comparatively, less stiffness of the DEJ enables this region to take up much of
the overall load through increased deformation.

Fig. 20. An edge view (side profile) of displacements. A qualitative
interpretation of the magnitudes and gradients in displacements can be
observed along one sample shown here. A graph of the averaged
displacements across the surface determined for each row along the
sample is superimposed on the displacement profile (black line), and
three regions can be identified: (A) enamel, in which the displacements
follow a plateau with no obvious gradient in displacements, (B) DEJ zone,
where a marked gradient in displacements can be identified, and (C) bulk
dentin revealing a smaller gradient. The stiffness is inversely related to
the gradient and consequently region (B) has the lowest stiffness. Red
arrow points to the approximate location of the DEJ. And yet, due to the
noise, a quantitative estimate of strain cannot be derived from the trends
in the average displacement values in this manner.(Zaslansky, Currey et
al. 2005)

42

43

Fig. 21. Pseudo-3D representation of the averaged displacement maps of
six tooth samples. Each panel represents the average displacement map
calculated from repeated compression-cycle speckle images. The maps
are orientated such that the displacements of the enamel at the upper end
of the physical sample appear on the right side of each panel (note
indications of ‘top,’ ‘bottom’ and distances obtained from measurements
of the sample dimensions, which are depicted beneath each displacement
map). The displacements were calculated relative to a point located in the
center of the enamel compared to which positive (upward) displacements
appear in the dentin (left side of each panel). The averaged stress for
each buccal (B) or lingual (L) sample is indicated below each
corresponding averaged displacement map. All axes are scale in
micrometers. Note however that the base axes represent lateral
coordinates of the sample surface, whereas the vertical axis represents
displacement. A single color scale was used to facilitate visual
comparison between samples. Note that the displacement magnitudes are
well below 1µm.(Zaslansky, Currey et al. 2005)

Fig. 22. Branching at ends of dentinal tubules near DEJ. Very little
peritubular lining can be seen at a depth of about 10 µm below the DEJ
(black arrow). Enamel on the left. (Zaslansky, Currey et al. 2005)

SEM images of samples for this study have revealed structural changes
occurring within a 200-300 µm thick dentin region which lies immediately below
the DEJ. Within a distance of 10µm of the optical DEJ, dentinal tubules can be
seen terminating in the form of small branches (figure 22). In this region very
little peritubular-dentin (PTD) is observed and the tubules appear as empty
channels which penetrate the intertubular dentin (ITD)(Zaslansky, Friesem et
al. 2006). These empty channels contribute to a reduction in the mineral
density of dentin and ultimately allow this region to deform more easily when
loaded.

44

45
Observations based on SEM images describe structural dentinal changes
which support the cushion or gasket of dentin, producing a phenomenon first
described by Wang and Weiner.

SEM images taken more than 20µm from the optical DEJ show a gradual
thickening of PTD until an average PTD thickness of 1µm is achieved (Figure
23). This increase in thickness is accompanied by an increase in tubule density
per unit volume (Zaslansky, Friesem et al. 2006). The convergence of tubules
combined with the thickening of the PTD contributes to the increased stiffeness
of the dentin at a distance from the DEJ. SEM images have also demonstrated
changes in the fiber orientation of ITD. The ITD at a distance of 20-50µm from
the DEJ forms a reticulate structure with the fibers traveling in random
orientations relative to the tubule. In contrast, the ITD fibers found deeper in
the dentin run in an orthogonal pattern to the tubules.

Fig. 23. Collagen fibrils are oriented at all angles relative to the tubules at
a distance of 50 µm from enamel (white arrow). An increase in PTD
thickness can be seen (black arrows). (Zaslansky, Currey et al. 2005)

Fig. 24. Dentin, observed 50 µm beneath the DEJ. The ITD has a porous
fibrous reticulate texture.(Zaslansky, Currey et al. 2005)

46

47
SEM was used to study the enamel, DEJ, and dentin in concert with
displacement maps. It was observed through these imaging techniques
coupled to ESPI that the stiffness and reinforcement of the dentin layer
increases with increasing distance from the DEJ. “Specifically we observe that:
(i) the peritubular-dentin (PTD) increases (with increasing distance from DEJ) to
form less bendable tubules having 1um thick wall; (ii) tubules become more
densely packed, having less intertubular-dentin (ITD) volume between them;
(iii) collagen fibers in the ITD found near the DEJ are oriented in all directions
relative to the tubule long axis”(Zaslansky, Friesem et al. 2006).
These finding are consistent with the displacement maps which observe
continually changing stiffness profiles across the DEJ zone.

48

Chapter VII: Mechanical Gradients in Biology

Successfully joining dissimilar materials is a challenge which is not only
restricted to biological structures. Man made structures for example often
combine diverse materials such as wood, metal, concrete, and glass. When
exposed to mechanical stresses, man made structures built from distinct
materials and biological structures containing both hard and soft tissues face
similar obstacles. Biological structures like the DEJ are of great interest in part
to study how nature has resolved a reoccurring design challenge.

Many of earth’s earliest life forms such as amoebae, slime molds, and jellyfish
are composed only of soft tissue. Refinement of these structures over time led
to higher life forms with increased organization and complexity. Hard tissues
such as bones, shells, and tendons provided a framework for the softer tissue.
While these stiffer frameworks were advantageous, it is likely that complex
parallel adaptations were also necessary to accommodate refined structures
that integrated dissimilar tissues types (Waite, Lichtenegger et al. 2004).
“Contact deformation and damage” was described by (Suresh 2001) as the
primary mechanical challenge faced by early biological systems with hard and
soft tissues. This describes a local condition that exists specifically at the
junction of soft and hard tissue.
“Contact deformation and damage” explains that the soft tissue directly in
contact with hard tissue becomes damaged due to the comparatively increased
stiffness associated with the hard tissue. The DEJ is just one example of many
scaffold structures found in nature which have successfully overcome the
obstacle of “contact deformation and damage”.

Fig 25. Radial stress, σr, in a butt joint made of two materials having
different stiffness (Ei). Graph A shows how σr increases as the stiffness
of B decreases relative to A, fixing the Poisson ratio of both at 0.4. Graph
B shows how σr increases as the difference between Poisson ratios of νA
and νB increases at a fixed stiffness ratio of 1. A nominal axial stress σz
=1 is assumed. (Waite, Lichtenegger et al. 2004)

The differing stiffness between two contacting materials determines the extent
to which contact deformation occurs. The stiffness of a material is technical
referred to as Young’s modulus or initial modulus and is represented as Ei.
Stiffness is a property that can be determined for any material in tension,
compression, or shear from the linear portion of a stress-strain curve.

49

50
This curve is defined by Hooke’s law, Ei = σ/ є, where σ is stress or force
divided by cross sectional area, and є is strain or change in length divided by
initial length (referenced in)(Waite, Lichtenegger et al. 2004). The poisson ratio
is an important characteristic of materials subjected to load. The poisson ratio
describes the lateral deformation of a material subjected to uniaxial loading
(Waite, Lichtenegger et al. 2004). For example, a cylindrical material under
tension has a poisson ratio of 0.4 denoting the fact that a 40% reduction in
diameter translates into a 100% increase in length. When two contacting
materials have the same stiffness and poisson ratio, the radial stress, σr will be
zero. The first graph in figure 25 demonstrates that when the poisson values of
two materials are fixed at 0.4, a change in the stiffness value of one material
relative to another results in a linear change in radial stress. In contrast, the
second graph demonstrates the effects of altering the poisson value of one
material relative to another while maintaining fixed stiffness values. This graph
distinguishes the poisson value from the stiffness value by highlighting the fact
that a change in poisson value results in an exponential adjustment in radial
stress. The case described in figure 25 is a simplistic example demonstrating
the interrelationship existing between radial stress, the stiffness ratio, and
poisson ratio between two contacting materials.

Biological systems and human manufacturing approaches have independently
adopted similar techniques which prevent the failure at interfaces joining
dissimilar materials. In manufacturing there are two methods of reducing
interfacial failure according to Waite and colleagues. The first and most widely
used technique is to increase the surface area of the contact zone.
By increasing the bonding area, the first technique effectively enhances the
threshold for failure. The second and less obvious technique is the formation of
a functional gradient. In this design, two distinct materials gradually transform
over a distance to become more alike at the site of contact. This strategy
effectively reduces the likelihood for failure by reducing the mismatch in
material properties at the interface.

Fig 26. Schematic mussel on the half-shell with one byssal thread
showing the incremental steps in stiffness, Ei, from the retractor muscles
to the rock. Note the 10-fold decrease in stiffness between the distal and
proximal portions of the thread. (Waite, Lichtenegger et al. 2004)

51

52
Byssus is a non-repairable connective tissue utilized by mussels. Byssus
provides a secure attachment for mussels by mediating contact between itself,
a soft tissue, and other stiff materials. This biological system employs a
functional gradient to securely bind materials such as rock and coral (figure 26)
(Waite, Lichtenegger et al. 2004). Mussel byssus is composed of hundreds of
threads which vary in length from 2 to 4cm. The thread bundles are
strategically arranged in a nonuniform manner. This concentration of thread
bundles at one end of the connective tissue causes the distal end attaching the
stiff material to itself be stiffer in comparison to the proximal end which binds
the soft tissue (Waite, Lichtenegger et al. 2004). Mytilus galloprovincialis is a
mussel which specifically employs byssal thread for attachment to rock. These
threads undergo a 10 fold change in stiffness when comparing the distal to the
proximal end (Waite, Lichtenegger et al. 2004).

The fundament unit of structure in byssal thread is preCOL, a collagen
containing precursor. The proposed structure of the preCOL is a trimeric block
domain structure. This structure is composed of a central collagen domain, a
flank region, and two histidine rich regions located at the N- and C-terminal
ends. PreCOL has been found in three variants in which the postscripts P, D,
and NG denote proximal, distal, and nongrade respectively (figure 27 & 28).
These distinctions are found in the “flank” region of the preCOL and are likely
responsible for differences which ultimately enable the structure to vary in
stiffness. The preCOL-D has the highest estimated stiffness constant of 10,000
MPa, followed by preCOL-NG with a constant of 150 MPa, and lastly preCOL-P
with a constant of 2 MPa (Waite, Lichtenegger et al. 2004).

Fig 27. Proposed structure of the preCOLs making up the bulk of each
mussel byssal thread. The block domain structure in 2D of a trimer is
shown in panel A; the bent-core analogue of a trimer is shown in panel B;
amino to carboxy terminal orientation is top-to-bottom. Summary of
sequence features of each flank type and estimated stiffness constants
(panel C) are as described in text. X represents any amino acid except
glycine and alanine. Stiffness represents the initial modulus determined
for fully hydrated biopolymers including the alanine-rich fibroin of
Anaphe silk

53

Fig 28. Model of the hexagonal (6 + 1) bundles of bent-core trimers in the
flower and banana configurations (top) and AFM image of a smectic array
of preCOL-D bundles in the proximal portion of the thread stretched by at
least 100% (bottom). A pair of overlaid model molecules related by a C2
point symmetry is shown in black with the N-termini in blue. (Waite,
Lichtenegger et al. 2004)

It is currently unknown how preCOLs might assemble to create a higher ordered
structure which transitions in stiffness. Two models have been proposed to explain
how preCOL P, D, and NG might assemble (figure 5).

54

55
Model A proposes that preCOL-NG alternates between preCOL-P and preCOL-
D to form an alternating repeat pattern. An example of this sequence would be,
D-NG-D-NG-P-NG-P-NG (Waite, Lichtenegger et al. 2004). The gradient in this
model would result from the different positions of P-onset in the packed
microfibrils. In Model B, the NG fibers are separated from the P and D fibers
and the sequences could be, NG-NG-NG-NG-NG-NG and D-D-D-P-P-P (Waite,
Lichtenegger et al. 2004). In this model the gradient would be formed by the
different positions at which D changes to P in the packed D/P microfibrils.
Topographical maps of these byssal preCOL threads were developed using
AFM. Preliminary AFM images suggest a lateral uniformity in the thread
structure which supports model B over model A.

Fig 29. Distribution of different preCOLs along a byssal thread. Gradients
of preCOL-D (D) and preCOL-P (P) in the thread (top) indicate that
preCOL-D predominates in the distal portion, whereas preCOL-P
predominates in the proximal portion. PreCOL-NG is uniformly present all
along the thread. Possible relationship between preCols in the
transitional region according to two assembly models A and B (bottom) is
represent by PreCol-D (black), preCol-NG (gray), and preCol-P (white).
The elastic modulus calculated according to the Voigt equation appears
below each increment. (Waite, Lichtenegger et al. 2004)

Gradients allow for smoother stress relief, reduced concentration of stress,
improved bonding strength, and increased fracture toughness for load bearing
structures. The advantages for gradients used in mechanical structures are
overwhelming but few studies have sought to understand how biological

56

57
structures develop bioscaffolds at the molecular level. Mussel byssus is a
simplistic model of a gradient used in biology (Waite, Lichtenegger et al. 2004).
The mussel byssus is a biomolecular example which employs a gradient to
attach a soft living structure to one that is hard and non-living. Waite, and
colleagues have studied this model to gain insight into structural bioscaffolds.
Analysis of these structures at the molecular level requires research tools
designed for material science, biophysics, and physical chemistry (Waite,
Lichtenegger et al. 2004). AFM is currently a commonly used tool in this
research. The absence of analytical tools with greater resolution than AFM
continues to be the major hinderance toward further discoveries.

58
Chapter VIII: Self-healing molecular mechanisms

Molecular mechanisms whose bonds are able to break under strain and yet
reform to regenerate the original shape are considered to be self-healing. The
majority of the tooth is a mineralized tissue known as dentin which provides
support for the overlying enamel and serves to protect the underlying pulp. By
weight, dentin is composed of 70% mineral, 20% organic matrix, and 10%
water (Kim, Simmer et al. 2006). The bulk of dentin is made up of collagen
which provides a scaffold for intra- and interfibrillar mineralization. Collagen is
present in numerous tissues including tendons, bones, ligaments, and
basement membranes. In recent years AFM has enabled researchers to
investigate the mechanical properties of various interacting proteins. Using
AFM, researchers have been able to successfully study the adhesion properties
of collagen fibrils by removing single molecules from rat tail tendons
(Gutsmann, Fantner et al. 2004). As a result of these studies, Hansma and
colleagues have proposed a model in which certain molecules provide a
reversible toughening mechanism which increases the overall energy required
to irreversibly deform and or break a material. This toughening mechanism
involves sacrificial bonds and hidden lengths to transfer more of the energy into
enthalpy while minimizing increases in entropy.

59
Sacrificial bonds and hidden lengths were discovered using atomic force
microscopy while studying the toughness of natural fibers, composites and
adhesives (Smith, Hansma et.,al 1999). Sacrificial bonds are bonds that break
prior to the failure of the main structural link (Smith, Hansma et.,al 1999).
These bonds are frequently weaker than the covalent bonds which maintain the
molecular backbone. Below a certain threshold of external force, sacrificial
bonds can hide the true length of the molecule. Hidden lengths are defined as
open spaces in the molecule that are immobilized due to the presence of
sacrifical bonds. During the failure of these sacrificial bonds, energy is also
absorbed through the movement of these hidden lengths.

Fig. 30. Fracture surface of human bone showing mineralized collagen
fibrils. b, Individual collagen fibrils are held together with cross-links
between fibrils and other non-collagenous proteins (arrows), which might
resist separation of the filaments. c, AFM image of a fractured surface
also showing filaments (arrows) between neighboring fibrils. d,
Schematic showing how these filaments could resist the separation of
fibrils. Initially, the mineralized collagen fibrils are bound together by non-
collagenous proteins. e, When a force is applied, the non-collagenous
proteins resist the separation of the fibrils and filaments form between
the mineralized fibrils. (Fantner, Hassenkam et al. 2005)

60

61
Using SEM images, Fantner, Hansma and colleagues have observed
mineralized collagen fibrils from the surface of fractured human trabecular
bone. From these images it was found that collagen fibrils in their natural
environment can be packed together as tightly or loosely spaced molecules.
The spaces between collagen fibers are sometimes bridged by small filaments.
These bridging filaments are initially a thin layer between the fibril but these
filaments stretch under load like a glue to resist separation. Previous AFM
studies have also predicted the presence of spaces between the mineralized
collagen fibrils. These studies concluded that both crack formation and bone
fracture occur in these spaces which separate the collagen fibrils. This is an
example where providing nothing results in an empty space which provides an
improvement in function.

62
To study the molecular adhesion forces of mineralized collagen fibrils,
researchers created a system based off single-molecule force spectroscopy.
Using an AFM cantilever, two small pieces of bone in solution were cycled
between phases of contact and separation while varying ion concentrations as
well as the molecular restoring time. The “molecular restoring time” (figure 31
Panel A) is the time lapse between cycles which enables connections between
mineralized collagen fibrils to partially reformed. After being pressed together,
the external force and the associated stretching required to separate the bone
was measured and graphed (figure 31 Panel B). This study concluded that
more energy was dissipated in the presence of calcium ions compared with
sodium when an external force was applied. (figure 31 Panel C&D). This
finding supports the previous observations of Thompson and colleagues who
initially proposed a calcium ion dependant self-healing mechanism found in
bone (Thompson, Kindt et al. 2001). In addition there was an increase in
energy dissipation when the molecular restoring time exceeded 10 seconds
(Gutsmann, Hassenkam et al. 2005). The non-fibrillar organic matrix was
interpreted to act like a glue and the mineralized collagen fibrils are thought to
contain sacrificial bonds which require several seconds for reformation (figure
31 Panel C). The restorative property between these bonds enabled
subsequent separation cycles to require similar external forces for bone
separation when compared to the earlier cycles. This “healing” mechanism
contributes to the toughening mechanism of bone and other similarly organized
biological structures like dentin which also contains mineralized collagen fibrils.

63

The types of bonds involved in the mechanism of self-healing of mineralized
collagen fibril toughening have not been identified but the influence of calcium
on this system does provide some insights. It has been hypothesized that this
system contains calcium mediated sacrificial bonds (Fantner, Hassenkam et al.
2005). These sacrificial bonds could be found in the non-fibrillar organic matrix,
the collagen fibrils, or even the mineral plates (figure 32 Panel B). The
formation of these sacrificial bonds could occur between each of these
structures and it is likely that multiple sacrificial bond configurations exist.

Fig 31. A piece of bone glued to an AFM cantilever is pressed onto
another piece of bone. As these are subsequently pulled apart, filaments
exert forces that resist the separation of the mineralized fibrils. b,
Representative pulling curves. Upper curve: not all filaments were
broken; lower curve: all filaments were broken. The restorative forces are
large relative to the forces involved in single polymer chain. c, The total
energy dissipation involved in separation of the bone fibrils is greater if
calcium ions are present (data averaged over several hundred pulls). Red
curve: with calcium and sodium ions in the buffer; blue curve: with
sodium ions but without calcium ions in the buffer. The error bars are
standard deviations. d, Filaments hold on for a longer pulling distance if
Ca ions are present. The average lengths at which the filaments break
completely (see the black lines) are 2.7 0.06 m and 1.9 0.09 m for Ca
and Na buffer respectively, with a Student's-t-test uncertainty of >0.001.
(Fantner, Hassenkam et al. 2005)

64

Fig 32. Glue filaments could resist the separation of mineralized fibrils. b,
The suspected, calcium-mediated sacrificial bonds in the bone could form
between (1) two binding regions on one polymer, (2) two polymers or (3) a
polymer and a mineral plate or a combination of these. For all cases the
sacrificial bond might involve multiple weak bonds in parallel. (Fantner,
Hassenkam et al. 2005)

The basic model of a sacrificial bond and hidden length is one in which a single
polymer chain loops back onto itself, creating a single sacrificial bond and
hidden length feature (figure 33). Initially the external force (shown in black)
contributes exclusively to the stretching and breaking of the intact sacrificial
bond which typically requires a force of 300nN. Following this initial phase,
subsequent external forces contribute to the extension and stretching of the
newly released hidden length (shown in red). This movement causes an

65

66
increase in entropy at the immediate portion of the polymer containing the
hidden length. At the same time the overall force measured from both ends of
the polymer has been reduced due to the movement of the hidden length.
Eventually the entire polymer and hidden length become fully extended (figure
33). The amount of energy required to extend the hidden length of a molecule
is much greater than the energy required to break even a strong bond!
Additional external forces at this time cause the entire polymer to stretch, thus
increasing the entropy of the entire molecule. When the force relaxes, the
molecule could potentially revert back to its initial conformation with a hidden
length and reformed sacrificial bond. The ability of this self-healing mechanism
to retain its initial shape would explain how energy could be absorbed without
irreversible damage being incurred to the overall structure.

Fig 33. Schematic drawing of the basic principle of the sacrificial bond-
hidden length mechanism. Before a sacrificial bond is broken, only the
black length of the molecule contributes to entropic spring and therefore
to the force with which the molecule resists the stretching. The red length
of the molecule is hidden from the applied force by the sacrificial bond.
When the bond breaking force is reached, only a small amount of energy
is needed to break the sacrificial bond. After that, the whole length (black
plus red) contributes to the entropy of the molecule. In total, the energy
that has to be put in to break the molecule is increased by the shaded
area under the first peak. Additionally, the initial slope of the force curve
is steeper, which indicates that the molecule is initially stiffer. (Fantner,
Oroudjev et al. 2006)

The model in figure 33 depicts the formation of a sacrificial bond within a single
molecule. Unlike this simple model, the formation of sacrificial bonds and
corresponding hidden lengths can also become more complex and occur
between distinct molecules. To portray these interactions, several models have
been proposed by Hansma and colleagues which describe the potential
patterns of resulting sacrificial bonds (figure 34). Sacrificial bonds between
various molecules differ in their breaking forces.

67

68
The sacrificial bonds indicated by A, B, & C in the following examples have 200,
300, and 800nN forces respectively. Case 1 in figure 34 is very similar to the
example in figure 33 where a single molecule forms sacrificial bonds. Case 1
differs in the fact that this molecule contains multiple sacrificial bonds with
varying breaking forces. In the presence of an external force each sacrificial
bond experiences an equal load which causes the molecule to stretch. The
sacrificial bonds break in the order of their weakest breaking force. The
location or position of the sacrificial bond does not influence the order of
breakage since the load is equally distributed. In this particular case the
successive breaking peaks will always be greater than the preceding peaks.

Fig 34. Various sacrificial bonds on single strand connections. (1) The
molecule binds to itself with sacrificial bonds; the hidden length is the
distance between the two binding sites on the molecule. When a force is
applied and molecules are stretched, the sacrificial bonds will break
according to their relative bond strengths. (2) The molecule binds to the
surface with sacrificial bonds. The hidden length equals the distance
between two neighboring sites on the molecule that bind to the surface.
When a force is applied and the molecules are stretched, the sacrificial
bonds break in the sequence that they are in on the molecule. (3) One
molecule binds to a second molecule with complementary binding site
sequence. The hidden length equals the distance between neighboring
sacrificial bonds on one molecule plus the distance between the
corresponding binding sites on the second molecule (in the figure this
length is twice the distance between A and B). When a force is applied
and the molecules are stretched, the bonds break in the sequence that
they are in on the molecules. (4) One molecule bonds to a second
molecule with a reversed complementary binding site sequence. The
sacrificial bonds are loaded in parallel; therefore they will all break once a
force is reached larger than the combined binding force. In this case, no
hidden length is set free. (Fantner, Oroudjev et al. 2006)

69

70

The graph reflecting force versus pulling length from case one in figure 34
demonstrates that sacrificial bonds in that configuration break in the order of
their lowest to their highest failure force. This order is due to the fact that the
external forces are equally distributed between each of the sacrificial bonds.
The external force in case two however is not equally distributed throughout the
sacrificial bonds. The different overall configuration in this case causes the
bonds to break in the order which they become exposed to the external force.
In the example of case 2, only one sacrificial bond can be exposed to an
external force at a given time. The failure force is determined by the type of
sacrificial bond but here their arrangement on the molecule determines the
order for bond failure. The failure of one sacrificial bond causes the
subsequent sacrificial bond to now bear the external force. As a result, the
order of breaking points in graph 2 is determined by the arrangement of the
sacrificial bonds on the molecule and not by the failure force.

The arrangement in cases 3 and 4 depict the formation of sacrificial bonds
between separate molecules. The order of bond failure in the third case is
analogous to the second case where the determining factor is molecular
arrangement of the sacrificial bonds. In this configuration, the strength of the
failure force does not determine the order of breakage and thus the failure force
of subsequent sacrificial bonds can be lesser than or greater than the
preceding bond. The bonds in case number 4 are arranged in a parallel

71
configuration. Here the external force is equally distributed between the bonds
in the parallel configuration. This configuration is unique and results in a
structure whose sacrificial bonds fail simultaneously. The result is a single,
large failure force which represents the failure of all sacrificial bonds. In this
configuration there are no hidden lengths which contribute to the absorption of
energy. The configuration in case 4 has been observed in DNA molecules
(Krautbauer 2003, Bockelmann 2002, Lubensky 2000).

72
Conclusion

From a biochemical perspective, teeth are a fascinating organic structure.
Teeth are composed from several tissues layers which are distinct in material
composition and structural properties. Due to the non-repairable nature of
teeth, it is critical that this organ develops in a precise manner. When
accurately assembled, teeth generally are successful at dissipating a lifetime of
strain from mastication.

The biological paradigm presented in fig. 1 demonstrates how developmental
events influence and guide the assembly of an organ with unique
biomechanical properties(Paine, White et al. 2001). Early gene transcription
and translation events control the quantity and timing for the production of
certain cells and proteins. This regulation mechanism is critical during
developmental processes which require the production of certain cells and
proteins which ultimately produce larger cellular structures.

Amelogenesis Imperfecta is an X-linked disorder which demonstrates how
errors occurring during gene translation can ultimately influence the final
structure and mechanical properties of fully developed teeth. Amelogenin is a
critical protein that properly orients HAP crystals in the developing enamel
layer. Amelogenesis Imperfecta is a disorder which prevents the proper
translation and synthesis of the amelogenin protein.

73
As a result, the fully developed enamel layer is of reduced thickness, grooved,
and pitted (Gibson, Yuan et al. 2001). Due to these structural flaws the enamel
layer does not posses the strength and longevity characteristically present in a
normally developed enamel layer.

The mechanical properties which enable developed teeth to last a lifetime are
the result of the unique material composition and architecture. The hard outer
enamel layer combined with the soft inner dentin tissue creates a complex yet
advantageous structure that can dissipate stress. The integration of these
dissimilar layers however presents additional structural challenges at the
interface that increases the chances of delamination due to strain. Biology has
however developed techniques which help prevent bound tissue layers from
failing due to external stresses. One common technique is to increase the
surface area of the binding surface. Increasing the bonding area enables the
junction to sustain a greater load of energy without failure (Waite, Lichtenegger
et al. 2004). Another less obvious strategy is the formation of a transition zone
between the separate layers. When exposed to stress, this zone acts as a
buffer to create a functional gradient. Reducing the mismatch in material
properties at the interface effectively reduces the likelihood for failure (Waite,
Lichtenegger et al. 2004). The dentin enamel junction (DEJ) not only maintains
the structural integrity of this interface but it also functions as a cushion or
gasket to absorb strain between the two distinct layers.

74
Current research has revealed a pattern of protein development at the DEJ
which might explain factors that influence the final biomechanical properties.
The development of the DEJ begins with the breakdown of the basement.
Ameloblast cells from the enamel layer are and Odontoblast cells from the
dentin layer both begin secreting proteins into the space which was previously
the basement membrane (Paine, Luo et al. 2005). At this critical period it is
interesting that both these cells have been found to secrete dentin
sialophosphportein (DSPP). The simultaneous secretion of DSPP by these
cells suggests a potential mechanism that could increase the molecular
interactions occurring between the enamel and dentin layer at the DEJ(Paine,
Luo et al. 2005). Additional bonding due to the presence DSPP is likely to
increase the threshold for strain required to cause irreversibly damage to the
structure. The ability of DSPP to positively influence the final biomechanical
properties of teeth reinforces the critical importance of biological fabrication.

Studies by Zaslansky and colleagues have observed the displacement patterns
of enamel, dentin, and the DEJ when exposed to an external load. This
research demonstrated (figure 20) that the DEJ experiences the most
compression of the three layers (Zaslansky, Currey et al. 2005). The ability to
compress is inversely proportional to stiffness. Scanning electron images
(SEM) of dentin at varying distances from the DEJ revealed trends in the
material composition which likely influences the deformation properties of teeth.

75
The high mineral density of peritubular-dentin (PTD) has been found to
contribute to the increased stiffness of dentin. At a distance of 10µm from the
DEJ, SEM images show very little PTD in the adjacent dentin(Zaslansky,
Currey et al. 2005). As a result, this region lacks significant stiffness and
deforms easily when loaded. The abundance and thickness of PTD in the
dentin layer increases with increasing distance from the DEJ (Zaslansky,
Currey et al. 2005). This trend causes the dentin to increase in stiffness away
from the DEJ. This is one mechanism which demonstrates why the DEJ region
compresses to a greater extent when compared to the overall dentin layer. The
transition of PTD in this region developed as a result of cellular fabrication
processes and ultimately created a functional gradient of mechanical
properties.

Self-healing molecular mechanisms have recently been identified as an
additional strategy that contributes to the toughness and failure resistance of
certain materials (Smith, Hansma et.,al 1999). This strategy works like Velcro,
but at the molecular level. Sacrificial bonds are weak bonds that fail under
strain but ultimately reform once the energy has been dissipated. The failure of
sacrificial bonds releases segments of proteins known as hidden lengths which
stretch to act as an additional energy absorbing mechanism. In abundance,
sacrificial bonds and hidden lengths significantly impact the ability of certain
materials to resist failure, thus contributing to the overall toughness of a
structure (Fantner, Oroudjev et al. 2006).

76
This self-healing mechanism breaks prior to the failure of the main structural
framework and reduces the load which is exposed to these non-repairable
covalently bound structures.

The discussed developmental and mechanical aspects of teeth will assist in the
production of more effective tooth restorations. Understanding the biological
fabrication process of the DEJ will lead to the production of materials which
bond to teeth more effectively. These materials will also incorporate functional
gradients including sacrificial bonds and hidden lengths. Taking into account
the way each layer including the DEJ deforms under load will enable materials
to be developed with similar material composition and structural mechanisms.
The knowledge of certain identifiable mechanisms in teeth such as functional
gradients, crack deflection, sacrificial bonds, and hidden lengths suggests how
this organ is able to be withstand a lifetime of external stress through
mastication.

77

Bibliography

Adams, J. C. and F. M. Watt (1993). "Regulation of development and
differentiation by the extracellular matrix." Development 117(4): 1183-98.

Aoba, T., E. C. Moreno, et al. (1989). "Possible roles of partial sequences at N-
and C-termini of amelogenin in protein-enamel mineral interaction." J Dent
Res 68(9): 1331-6.

Collier, P. M., J. J. Sauk, et al. (1997). "An amelogenin gene defect associated
with human X-linked amelogenesis imperfecta." Arch Oral Biol 42(3): 235-
42.

Fantner, G. E., T. Hassenkam, et al. (2005). "Sacrificial bonds and hidden
length dissipate energy as mineralized fibrils separate during bone fracture."
Nat Mater 4(8): 612-6.

Fantner, G. E., E. Oroudjev, et al. (2006). "Sacrificial bonds and hidden length:
unraveling molecular mesostructures in tough materials." Biophys J 90(4):
1411-8.

Fincham, A. G., A. B. Belcourt, et al. (1982). "Comparative protein biochemistry
of developing dental enamel matrix from five mammalian species." Calcif
Tissue Int 34(2): 182-9.

Fincham, A. G., W. Leung, et al. (1998). "Does amelogenin nanosphere
assembly proceed through intermediary-sized structures?" Connect Tissue
Res 38(1-4): 237-40; discussion 241-6.

Gibson, C. W., Z. A. Yuan, et al. (2001). "Amelogenin-deficient mice display an
amelogenesis imperfecta phenotype." J Biol Chem 276(34): 31871-5.

Gutsmann, T., G. E. Fantner, et al. (2004). "Force spectroscopy of collagen
fibers to investigate their mechanical properties and structural organization."
Biophys J 86(5): 3186-93.

Gutsmann, T., T. Hassenkam, et al. (2005). "Sacrificial bonds in polymer
brushes from rat tail tendon functioning as nanoscale velcro." Biophys J
89(1): 536-42.

78

Habelitz, S., S. J. Marshall, et al. (2001). "The functional width of the dentino-
enamel junction determined by AFM-based nanoscratching." J Struct Biol
135(3): 294-301.

Imbeni, V., J. J. Kruzic, et al. (2005). "The dentin-enamel junction and the
fracture of human teeth." Nat Mater 4(3): 229-32.

Kim, J. W., J. P. Simmer, et al. (2006). "Mutational analysis of candidate genes
in 24 amelogenesis imperfecta families." Eur J Oral Sci 114 Suppl 1: 3-12;
discussion 39-41, 379.

Lench, N. J. and G. B. Winter (1995). "Characterisation of molecular defects in
X-linked amelogenesis imperfecta (AIH1)." Hum Mutat 5(3): 251-9.

Miner, J. H., J. Cunningham, et al. (1998). "Roles for laminin in embryogenesis:
exencephaly, syndactyly, and placentopathy in mice lacking the laminin
alpha5 chain." J Cell Biol 143(6): 1713-23.

Moradian-Oldak, J. (2001). "Amelogenins: assembly, processing and control of
crystal morphology." Matrix Biol 20(5-6): 293-305.

Moradian-Oldak, J., M. L. Paine, et al. (2000). "Self-assembly properties of
recombinant engineered amelogenin proteins analyzed by dynamic light
scattering and atomic force microscopy." J Struct Biol 131(1): 27-37.

Paine, M. L., W. Luo, et al. (2005). "Dentin sialoprotein and dentin
phosphoprotein overexpression during amelogenesis." J Biol Chem
280(36): 31991-8.

Paine, M. L. and M. L. Snead (1997). "Protein interactions during assembly of
the enamel organic extracellular matrix." J Bone Miner Res 12(2): 221-7.

Paine, M. L., S. N. White, et al. (2001). "Regulated gene expression dictates
enamel structure and tooth function." Matrix Biol 20(5-6): 273-92.

Simmer, J. P. and A. G. Fincham (1995). "Molecular mechanisms of dental
enamel formation." Crit Rev Oral Biol Med 6(2): 84-108.

Suresh, S. (2001). "Graded materials for resistance to contact deformation and
damage." Science 292(5526): 2447-51.

Thesleff, I. and K. Hurmerinta (1981). "Tissue interactions in tooth
development." Differentiation 18(2): 75-88.

79

Thompson, J. B., J. H. Kindt, et al. (2001). "Bone indentation recovery time
correlates with bond reforming time." Nature 414(6865): 773-6.

Waite, J. H., H. C. Lichtenegger, et al. (2004). "Exploring molecular and
mechanical gradients in structural bioscaffolds." Biochemistry 43(24): 7653-
62.

White, S. N., W. Luo, et al. (2001). "Biological organization of hydroxyapatite
crystallites into a fibrous continuum toughens and controls anisotropy in
human enamel." J Dent Res 80(1): 321-6.

Yamada, Y. and H. K. Kleinman (1992). "Functional domains of cell adhesion
molecules." Curr Opin Cell Biol 4(5): 819-23.

Zaslansky, P., J. D. Currey, et al. (2005). "Phase shifting speckle interferometry
for determination of strain and Young's modulus of mineralized biological
materials: a study of tooth dentin compression in water." J Biomed Opt
10(2): 024020.

Zaslansky, P., A. A. Friesem, et al. (2006). "Structure and mechanical
properties of the soft zone separating bulk dentin and enamel in crowns of
human teeth: insight into tooth function." J Struct Biol 153(2): 188-99.

Abstract (if available)

Abstract Unlike many aspects of the body, human teeth do not have natural regenerative mechanisms. Instead, the multiple tissues and mechanical structures in teeth combine to create a structural gradient that successfully absorbs a lifetime of impact forces from routine mastication without irreversible damage. The integration of hard and soft tissue types is advantageous for distributing loads but the resulting interfaces are prone to fracture. The intermediate boundary known as the dentino-enamel junction (DEJ) is a natural strategy which avoids abrupt changes in mechanical properties. The DEJ is a self-assembling structure which combines two dissimilar materials, enamel, and dentin. This junction is critical to the longevity of tooth function and is recognized to fulfill multiple roles. This thesis reviews the current state of knowledge related to the DEJ and its surrounding structures that cooperate together to create a failure resistant structure.

Linked assets

University of Southern California Dissertations and Theses

Asset Metadata

Creator Becker, Bret (author)

Core Title Biological fabrication: gradients produce failure resistant dental structures

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Biology

Degree Conferral Date 2007-12

Publication Date 12/07/2007

Defense Date 09/07/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag amelogenesis,biological gradients,dentin enamel junction,enamel synthesis,OAI-PMH Harvest,sacrificial bonds,tooth deformation

Language English

Advisor Snead, Malcolm L. (committee chair), Stellwagen, Robert H. (committee member), Tokes, Zoltan A. (committee member)

Creator Email bjbecker@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m970

Unique identifier UC1348499

Identifier etd-Becker-20071207 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-594402 (legacy record id),usctheses-m970 (legacy record id)

Legacy Identifier etd-Becker-20071207.pdf

Dmrecord 594402

Document Type Thesis

Rights Becker, Bret

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu