Integrative analysis of gene expression and phenotype data

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content THE ENDOCYTOSIS OF ENAMEL MATRIX PROTEIN DERIVATIVES

by
Jason Lee Shapiro

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CRANIOFACIAL BIOLOGY)

August 2009

Copyright 2009 Jason Lee Shapiro

ii
Dedication

This dissertation is dedicated to my wife Camerin and daughter Adelyn whose
love and support throughout my studies made this body of work possible.

iii
Acknowledgements

I wish to acknowledge Michael Paine for his incredible creativity, support, and
patience, as well as for gently prodding me to complete this process. I am a
better scientist for having trained under such a great mentor.

To Malcolm Snead, thank you for your critical insight into my research as well as
being a great listener and offering helpful advice. I truly appreciate all that you
have helped me accomplish though this process.

To the other Committee members, Casey Chen, Curtis Okamoto, and Charles
Shuler, thank you for all your contributions. My experiences at USC have been
much richer with all of your help

iv
Table of Contents

Dedication ii
Acknowledgements iii

List of Tables vii

List of Figures viii

Abstract x

Preface xii

Chapter 1. Cellular Uptake of Amelogenin and its Localization to CD63 1
and LAMP1 Positive Vesicles
Materials and Methods 6
1. Cell lines and culture conditions 6
2. Plasmid constructs 7
3. Porcine enamel matrix proteins 9
4. Antibodies 9
5. Stable transfection of pAmel-DsRed into 10
MDCK cells and crude fluorescent-protein extracts
6. Immunohistochemistry 10
7. Immunofluorescence 11
8. Transient transfection and microscopy 12
9. Time-lapse confocal microscopy 13
10. RT-PCR 13
Results 14
1. Internalized red fluorescent amelogenin localizes 14
to the perinuclear region of osteoblasts
2. CD63 is expressed by ameloblasts at all stages of 15
amelogenesis, but is more highly expressed in
late-secretory and post-secretory ameloblasts
3. CD63 and amelogenin colocalization 17
4. LAMP1 is expressed by ameloblasts at all stages of 18
amelogenesis, but is more highly expressed in early
secretory stages
5. LAMP1 and amelogenin colocalization 20
6. The transfer of extracellular amelogenin into the cell 21
cytoplasm involves the direct passage of amelogenin
into LAMP1 positive vesicles

v
7. Ameloblast-like LS8 cells express CD63 and LAMP1 24
Discussion 24

Chapter 2. The Transcriptional Responses of a Non-amelogenin 31
Producing Cell to Exogenous Enamel Matrix Proteins
Materials and Methods 32
1. Cells 32
2. Primer design 32
3. Real-time PCR 33
Results 34
1.Real-time PCR 34
Discussion 35

Chapter 3. AP3B1 Null Mice, a Model of Hermansky Pudlak Syndrome, 37
have an Abnormal Dentition
Materials and Methods 40
1. Animals 40
2. Reverse transcriptase PCR of adaptor protein β1-4 42
subunits in ameloblast like (LS8) cells
3. Genotyping 42
4. Scanning electron microscopy 43
5. Micro-computerized tomography 43
6. Statistical analysis of tissue area using 44
micro-computerized tomography
7. Linear enamel thickness 45
8. Light microscopy imaging of AP3B1 null and 45
C57BL/6J mice
9. Characterization of the enamel surface of mouse 46
teeth by FTIR spectroscopy
10. Study of the periodontal ligament (PL) in AP3B1 nulls 47
and C57BL/6J animals
Results 47
1. RT-PCR of β1-4 in ameloblast-like cells 47
2. Genotyping of animals 48
3. Age and weight comparison 49
4. Examination of the dentitions in AP3B1 null and 49
C57BL/6J animals
5. Enamel thickness comparisons between AP3B1 null and 50
C57BL/6J mice
6. Peridontal ligament analysis 52
7. Examination of the molars in AP3B1 nulls and 54
C57BL/6J mice
8. Quantification of mineral and peptide content in teeth 56
from AP3B1 null and C57BL/6J mice
Discussion 57

vi
Chapter 4. An Amelogenin Minigene to Study Alternative Splicing 64
Materials and Methods 67
1. Mouse amelogenin (mAmel) minigene construction 67
2.Minigene transfection, RT-PCR and DNA sequencing 68
3. Nested PCR 69
Results 70
1. Two novel amelogenin isoforms identified using 70
the mouse amelogenin minigene
2. Differential splicing patterns are observed for the 72
amelogenin minigene in different cell lines
Discussion 73

Conclusions 76

References 80

Appendix Characterization of the enamel surface of mouse teeth 92
by FTIR spectroscopy

vii
List of Tables

Table 1. Primer pairs for LAMP1, CD63, AP-3 sigma and GAPDH 33
real-time PCR experiments

Table 2. Two novel mouse amelogenin isoforms identified using 71
the mouse amelogenin minigene

viii

List of Figures

Figure 1. MDCK cells stably transfected with amelogenin-red 15
fluorescent protein construct as seen under
fluorescent microscopy

Figure 2. Immunolocalization of CD63 to ameloblast cells in 16
a 3-day old mouse mandibular incisor

Figure 3. CD63 and exogenously added porcine amelogenin 17
colocalization in the perinuclear region of the mouse
preosteoblast cell line MC3T3-E1

Figure 4. Immunolocalization of LAMP1 to ameloblast like LS8 19
cells in a 4-day old mouse mandibular incisor

Figure 5. Colocalization of LAMP1 and exogenously added 21
porcine enamel matrix proteins in the perinuclear
region of mouse pre-osteoblasts MC3T3-E1 and
ameloblast-like LS8 cells

Figure 6. The movement of Alexa-fluor 594 labeled Emdogain® 23
into mouse pre-osteoblasts (MC3T3-E1) involves direct
passage into pre-established LAMP1-coated vesicles

Figure 7. RT-PCR of CD63 and LAMP1 gene transcripts in 24
ameloblast-like LS8 cells

Figure 8. Human fibroblasts do not increase transcription 35
of LAMPs or AP-3 when treated with exogenous enamel
matrix proteins

Figure 9. Targeting strategy to produce AP3B1 null mice 41

Figure 10. All adaptor protein subtypes are present in 48
ameloblast-like LS8 cells

Figure 11. Confirmation of genotypes 48

Figure 12. Weight and age comparisons of animals used in the 49
genotypic and phenotypic analysis

ix
Figure 13. Dentition of 12-month old AP3B1 null animals 50
compared to age-matched C57BL/6J control animals

Figure 14. µCT measurements for mature enamel thickness and 51
enamel, dentin and pulpal area in a cross section
perpendicular to the enamel surface just after eruption,
but fully mature region

Figure 15. Enamel incisor morphology of AP3B1 null and 52
C57BL/6J mice in the mature region, from 6-week
animals fractured perpendicular to the labial surface

Figure 16. Study of the periodontal ligament (PL) in AP3B1 null 53
and C57BL/6J animals

Figure 17. Analysis of the mesiodistal dimensions of the crowns 55
of first and second maxillary molars of 12-month old
AP3B1 null and C57/BL/6J animals

Figure 18. Comparison of the percentage of peak areas for the 56
absorbances of functional groups typical for hydroxyl
apatite and proteins

Figure 19. Intracellular trafficking of LAMPS 59

Figure 20. Features of the mAMEL minigene 68

Figure 21. Amino-acid sequence alignment for M180, M156, 72
M59 (LRAP), rM134 and rM18

Figure 22. Cell-specific alternative spliced amelogenin gene 73
product profiles

x
Abstract

Dental enamel is the hardest mineralized tissue in the body and initially
forms from a protein matrix. During enamel formation, ameloblasts secrete
unique proteins into the extracellular space to direct mineralization of enamel
crystallites. This specialized organic matrix contains the proteins amelogenin,
ameloblastin, and enamelin. Amelogenin is the major enamel matrix protein
secreted, comprising ~95% of the total matrix, yet after maturation and
completion of amelogenesis, only ~1% of the initial organic matrix remains in
mineralized enamel. Ameloblasts, being the only cells in close proximity to the
organic matrix, most likely are responsible for the removal of proteinaceous
matrix debris for proper enamel maturation. The rapid removal of enamel matrix
proteins is poorly understood in the literature, principally because clathrin (the
prototypical pathway for endocytosis) is expressed at very low levels in
ameloblasts. Most of the historical record of endocytosis is focused on clathrin
mediated endocytosis whereby molecules are endocytosed into clathrin coated
vesicles, pinched off from the plasma membrane into the cytoplasm, and fused
with early endosomes. From the early endosome, these molecules are targeted
to the late endosome and eventually to the lysosome where they are digested by
a myriad of proteolytic enzymes.
Because ameloblasts contain low levels of clathrin, receptor mediated
endocytosis was not believed to play a role in the removal of enamel protein
matrix debris. Recently, clathrin-independent endocytotic pathways have been

xi
described in the literature, and these pathways may explain the rapid removal of
matrix proteins required for mineralization of enamel. The hypothesis is that there
is a “receptor mediated” endocytotic pathway involved in enamel matrix
degredative products uptake, which involves lysosomal specific proteins and
mediators.

xii
Preface

Enamel organic matrix assembly, and the subsequent process of
biomineralization, occurs outside of the cell in the extracellular space. Enamel
matrix assembly follows the paradigm of basement membrane assembly,
however, unlike the basement membrane, enamel does not remodel. Rather,
cells that produce enamel (ameloblasts) retract away from the forming matrix
with concomitant mineral deposition (Simmer and Fincham, 1995). Once enamel
has matured, the ameloblasts remain dormant until the tooth erupts, and at this
time these cell are shed from the enamel surface. The proteins that comprise the
enamel matrix are amelogenin (Snead et al., 1983), ameloblastin (Fong et al.,
1996; Lee et al., 1996) and enamelin (Hu et al., 1997). Amelogenin, ameloblastin
and enamelin represent those structural proteins whose expression remains
relatively unique to the mineralized tooth structure (Begue-Kirn et al., 1998; Lee
et al., 2003; Nanci et al., 1998). It is apparent that the developmental timing in
which these gene products are presented to the enamel matrix must be finely
controlled and regulated. Because of the limited tissue expression of these
enamel proteins, it appears that their removal from the enamel matrix during
enamel maturation has required specific proteases whose spatiotemporal
expression must also be exquisitely regulated. Proteases found in the developing
enamel organ are the serine protease kallikrein-4 (KLK4) (Hu et al., 2000; Nelson
et al., 1999; Ryu et al., 2002) and the matrix metalloproteinase (MMP20)
(Caterina et al., 2002).

xiii
Much of our understanding of the removal of the organic enamel matrix
during the events of amelogenesis is derived from rodent studies using a
radiolabel amino acid “pulse” administered to the animal and studied at various
time intervals following administration (Smith, 1979, 1998; Smith et al., 1992;
Smith et al., 1989). The pathways discussed to explain the removal of the
enamel proteins favor a passive diffusion (resorptive) process such as
pinocytosis into ameloblasts (Reith and Cotty, 1967b), or, as described for
rodents, pinocytosis of the enamel proteins into odontogenic processes
(Josephsen and Warshawsky, 1982; Warshawsky and Josephsen, 1981), and
have generally excluded ameloblast-directed endocytosis because: 1) the levels
of clathrin protein within the cytoplasm of secretory ameloblasts is negligible
(Nakamura et al., 1994), and the presence of clathrin has historically been the
hallmark of receptor-mediated endocytosis (Robinson, 2004); and 2) it is hard to
reconcile that the apical face of polarized ameloblasts could simultaneously
manage both an extreme secretory activity, and a resorptive activity (Smith,
1998). In the past decade the concept of passive diffusion (pinocytosis) has been
challenged with the discoveries of four unique adaptor protein complexes, and
the acknowledgment that receptor-mediated, clathrin-independent endocytotic
pathways exist (Janvier and Bonifacino, 2005; Ohno, 2006a, b, c; Robinson,
2004).
Today it is well-known that receptor-mediated clathrin-independent
endocytosis does occur (Dell'Angelica et al., 1998; Johannes and Lamaze, 2002;
Peden et al., 2004; Robinson, 2004; Robinson and Bonifacino, 2001; Simpson et

xiv
al., 1996), and our previous understanding of pinocytosis, or the “passive”
cellular uptake of degraded matrix proteins is over simplistic or just simply not
correct (Le Roy and Wrana, 2005; Nichols and Lippincott-Schwartz, 2001).
There must also be care in equating the events of amelogenesis in
rodents to other animal species. Amelogenesis in rodent teeth differs from
amelogenesis in primate teeth in a number of ways. In rats the full thickness of
incisor enamel is produced in a 7.5-day period (Smith and Nanci, 1996), but the
half-life of enamel proteins in vivo is estimated at 8.9-days (Smith et al., 1992).
Thus, in rodents this would suggest that the enamel most proximal to the dentin-
enamel junction (DEJ) is still very immature during the enamel maturation stages,
and could allow for passage of degraded enamel proteins into odontoblastic
processes as previously suggested (Josephsen and Warshawsky, 1982;
Warshawsky and Josephsen, 1981). In primates, if we assume the same half-life
for the enamel proteins, we must than assume the DEJ is fully
mineralized/mature weeks after it is initiated, yet secretory ameloblasts and
amelogenesis in primates continues for more than one year (Lacruz and
Bromage, 2006). Thus in primates the only options for removal of the enamel
organic matrix, days after the initiation of amelogenesis, are either directly
through the apical face of ameloblasts (pinocytosis or endocytosis), or
intercellular movement (either passive or orchestrated) along the basolateral
surfaces of ameloblasts. While a passive diffusion (pinocytosis) of the enamel
matrix debris into secretory or maturation-stage ameloblasts may explain enamel
matrix removal, alternative explanations are also reasonable. One alternative

xv
explanation would be that as ameloblasts secrete enamel proteins into the
enamel matrix, they simultaneously employ a clathrin independent endocytotic
pathway to allow for the removal of organic enamel debris that results from
enzymatic degradation. This explanation is feasible with recent in vitro data that
has demonstrated that enamel matrix proteins interact with a class of
ubiquitously expressed integral membrane-bound proteins (Paine et al., 1998;
Wang et al., 2005). The members of this protein class are called lysosomal-
associated membrane protein (LAMP)-1, LAMP2 and CD63/LAMP3 (Tompkins et
al., 2006; Wang et al., 2005; Zou et al., 2007).
Adaptor protein (AP) complexes are heterotetramers that cycle between the
cell cytosol and membranes, and they mediate the sorting of membrane proteins
in the secretory and endocytic pathways. AP complexes frequently, but not
always, involve in the formation of clathrin-coated vesicles by recruiting clathrin.
Four distinct AP complexes have been identified; these are AP-1, AP-2, AP-3
and AP-4. Widely accepted has been the notion that AP-2 and AP-3 mediate
endocytosis from the plasma membrane, and AP-1, AP-3 and AP-4 each play a
role in the endosomal/lysosomal sorting pathways (Janvier and Bonifacino, 2005;
Robinson, 2004). AP-1 and AP-2 mediated endocytosis have an absolute
requirement for clathrin, while AP-4 mediated endocytosis is clathrin-independent
(Janvier and Bonifacino, 2005; Ohno, 2006a, b, c; Robinson, 2004). Over the
past decade the molecular mechanisms of non-clathrin associated endocytotic
pathways have been investigated (Dell'Angelica et al., 1998; Johannes and
Lamaze, 2002; Le Roy and Wrana, 2005; Nichols and Lippincott-Schwartz, 2001;

xvi
Peden et al., 2004; Robinson, 2004; Robinson and Bonifacino, 2001; Simpson et
al., 1996).
The intracellular trafficking of plasma membrane-bound LAMP1, LAMP2
and CD63 involves the recruiting of, and binding to, the AP-3 complex
(Berditchevski and Odintsova, 2007; Peden et al., 2004; Robinson and
Bonifacino, 2001; Rous et al., 2002). Data suggest that AP-3 related intracellular
trafficking may proceed independently of any association with clathrin, or
alternatively have minimal clathrin requirements. Although in mammals a clathrin
consensus-binding motif is part of the AP-3  subunit (coded by the AP3B1
gene), there is little or no physiological evidence to support the association of
clathrin and AP-3 in vivo (Dell'Angelica et al., 1998; Peden et al., 2004; Peden et
al., 2002; Robinson and Bonifacino, 2001; Shi et al., 1998; Simpson et al., 1996).
For example, while significant amounts of AP-1 and AP-2 are a feature of
biochemically purified clathrin-coated vesicles, only trace amounts of AP-3 can
be characterized in these same vesicles (Peden et al., 2004; Robinson and
Bonifacino, 2001; Simpson et al., 1996). This lack of association between AP-3
and clathrin-coated vesicles is also supported by the fact that in yeast AP3B1
has no clathrin consensus-binding motif, and yeast AP-3 related trafficking is
clathrin-independent (Robinson and Bonifacino, 2001).
Plasma membrane-bound CD63 trafficking to the late
endosome/lysosome is mediated through a direct interaction with µ subunit
(coded by the AP3M1 gene) of AP-3 (Nakatsu and Ohno, 2003; Robinson and

xvii
Bonifacino, 2001). AP-3 facilitates the budding of vesicles from the Golgi
membrane and may be directly involved in protein sorting to the
endosomal/lysosomal system. The trafficking of CD63 from the cell membrane to
the lysosome is initiated by a direct protein-protein interaction between CD63,
through a lysosomal targeting motif (GYXX ; where X is any amino acid and 
is a bulky hydrophobic amino acid) located at the C’-terminus, and AP3M1 (Di
Pietro and Dell'Angelica, 2005; Rous et al., 2002). This signature AP3M1-binding
motif is also present at the C’-termini of LAMP1 and LAMP2 (Rous et al., 2002),
suggesting that CD63, LAMP1 and LAMP2 all employ the same mechanism in
their trafficking from the plasma membrane to the lysosome (Zou et al., 2007).
In humans, mutations in the  subunit (AP3B1) of AP-3 have been
associated with the Hermansky-Pudlak syndrome (HPS), a genetic disorder
characterized by defective lysosome-related organelles. One animal model that
has been described resembles the Hermansky-Pudlak syndrome, and this is the
“Pearl” mice with mutations to the AP3B1 gene (  subunit)(Yang et al., 2000).
The homozygous pearl phenotype includes hypopigmentation of hair and eyes,
and a prolonged bleeding time associated with platelet storage deficiencies. The
pearl animal has a considerably shortened lifespan. To date, the enamel
morphology of pearl mice has not been described.
In the past two decades the human and mouse genome projects have
been completed. With the identification of clathrin-independent endocytosis, and
our current understanding of protein-protein interactions occurring between

xviii
enamel matrix proteins and plasma membrane bound proteins, I have set out to
re-evaluate ameloblast endocytotic capabilities.

1
Chapter 1.

Cellular Uptake of Amelogenin and its Localization to CD63, and LAMP-1
Positive Vesicles

The structural proteins relatively unique to the enamel matrix are
amelogenin, ameloblastin and enamelin (Bartlett et al., 2006; Fukumoto et al.,
2004; Hu et al., 2000; Hu et al., 2001; Lee et al., 2003; Margolis et al., 2006;
Paine et al., 2001). Structural proteins of the enamel matrix manifest specific
protein-protein interactions required to produce a matrix capable of directing the
highly ordered structure of the enamel crystallites (Bartlett et al., 2006; Margolis
et al., 2006; Paine et al., 2001). Protein-protein interactions also occur between
the secreted enamel proteins and proteins of the plasma membrane of the
enamel producing cells (ameloblasts) (Fukumoto et al., 2004; Iwasaki et al.,
2005; Moffatt et al., 2006; Zhu et al., 2006). Implied protein-membrane
interactions between ameloblasts and the enamel organic matrix have been
discussed for ameloblastin (Fukumoto et al., 2004), and also for amelotin (Moffatt
et al., 2006). Such protein-membrane interactions may be required to establish
short-term order of the forming matrix, to mediate feedback signals to the
transcriptional machinery of these cells, and to remove matrix protein debris
during amelogenesis (Bartlett et al., 2006; Fukumoto et al., 2004; Xu et al.,
2006b; Zhu et al., 2006). Plasma membrane-bound proteins identified in
ameloblasts, shown to directly interact with the structural enamel proteins,

2
include CD63 (cluster of differentiation 63 antigen; also known as lysosome-
associated membrane glycoprotein 3 and melanoma 1 antigen) and Lamp1
(lysosomal-associated membrane glycoprotein 1) (Duffield et al., 2003; Tompkins
et al., 2006; Wang et al., 2005; Zhu et al., 2006). CD63 and LAMP1 are also
integral to the late endosome/lysosomal membrane (Duffield et al., 2003).
Clathrin adaptor protein (AP) complexes are protein heterotetramers that cycle
between the cytosol and membranes, and mediate the sorting of membrane-
bound proteins in the secretory and endocytic pathways. Four distinct AP
complexes have been identified (AP-1, AP-2, AP-3 and AP-4) and at-least three
of these AP complexes have isoforms that show tissue-specific expression
(Nakatsu and Ohno, 2003; Robinson and Bonifacino, 2001). AP-3 has been
directly associated with the intracellular trafficking of plasma membrane-bound
CD63 and LAMP1 (Duffield et al., 2003).
CD63 is a member of the transmembrane-4 glycoprotein superfamily,
which is also known as the tetraspanin family (Stipp et al., 2003; Yunta and Lazo,
2003). Family members are cell-surface proteins that are characterized by the
presence of four transmembrane domains (Stipp et al., 2003; Yunta and Lazo,
2003). Tetraspanins mediate signal transduction events that play a role in the
regulation of cell development, activation, growth and motility (Mantegazza et al.,
2004; Yunta and Lazo, 2003). In particular, as a cell surface glycoprotein, CD63
and other tetraspanins are known to complex with integrins (Berditchevski, 2001;
Yunta and Lazo, 2003). Recent studies relating the tetraspanin-integrin protein
interactions suggest that as a class of proteins, the tetraspanins act as

3
organizers of membrane microdomains and signaling complexes (Yunta and
Lazo, 2003). CD63 resides not only in the plasma membrane of most cell types,
but it also resides in late endosomes, lysosomes and secretory vesicles, and
CD63 traffics among these different compartments (Duffield et al., 2003). This
has led to the suggestion that CD63 may play a role in the recycling of
membrane components, and the uptake of degraded proteins from the
extracellular matrix (Duffield et al., 2003).
At steady-state, LAMP1 is a transmembrane protein highly expressed in
late endosomes and lysosomes and is often used as a marker for these two
organelles (Cook et al., 2004). LAMP1 is involved in endocytosis, pinocytosis, or
phagocytosis (Cook et al., 2004). The movement of LAMP1 from the rough
endoplasmic reticulum and Golgi to the lysosome membrane has been
documented, and this pathway can be independent of movement through the
plasma membrane (Cook et al., 2004). However, LAMP1 immunoreactivity is
also observed at the plasma membrane of most cell types (Holcombe et al.,
1993; Kannan et al., 1996; Silverstein and Febbraio, 1992), and it can also be
observed in early endocytic compartments (Kannan et al., 1996). The presence
of LAMP1 on the plasma membrane suggests that LAMP1 can act as a cell
surface intermediary, and can traffic directly to the late endosome/lysosome
during endocytotic events (Duffield et al., 2003).
In eukaryotic cells, receptor-mediated endocytosis frequently occurs with
the involvement of clathrin-coated pits and vesicles, and is distinguished from
other types of endocytosis or pinocytosis that are independent of clathrin (Alberts

4
et al., 1994; Peden et al., 2004; Robinson, 2004). Over the past decade the
molecular mechanisms of non-clathrin associated endocytotic pathways have
been investigated (Dell'Angelica et al., 1998; Johannes and Lamaze, 2002;
Peden et al., 2004; Robinson, 2004; Robinson and Bonifacino, 2001; Simpson et
al., 1996). The intracellular trafficking of plasma membrane-bound CD63 and
LAMP1 involves the recruitment of the AP-3 complex (Peden et al., 2004;
Robinson and Bonifacino, 2001; Rous et al., 2002), and the data suggest that
AP-3 related intracellular trafficking may proceed independently of any
association with clathrin. Although in mammals a clathrin consensus-binding
motif is part of the AP-3 3 subunit, there is little or no physiological evidence to
support the association of clathrin and AP-3 in vivo (Dell'Angelica et al., 1998;
Peden et al., 2004; Peden et al., 2002; Robinson and Bonifacino, 2001; Shi et al.,
1998; Simpson et al., 1996). For example, while significant amounts of AP-1 and
AP-2 are a feature of biochemically purified clathrin-coated vesicles, only trace
amounts of AP-3 can be characterized in these same vesicles (Peden et al.,
2004; Robinson and Bonifacino, 2001; Simpson et al., 1996). This lack of
association between AP-3 and clathrin-coated vesicles is also supported by the
fact that in yeast the AP-3 3 subunit has no clathrin consensus-binding motif,
and AP-3 related trafficking is totally independent of clathrin (Robinson and
Bonifacino, 2001).
The data outlined above has directed the experimental strategy described
in this current investigation. Amelogenin is the most abundant protein in the

5
developing mammalian enamel extracellular matrix, accounting for greater than
90% of the total matrix protein content (Deutsch et al., 1991; Margolis et al.,
2006; Robinson et al., 1998; Termine et al., 1980). Mature enamel is almost
entirely inorganic, thus the process of amelogenesis must include an efficient
mechanism for the removal of the organic matrix component and related organic
debris. With the identification of two secreted, enamel-specific proteinases
(matrix metalloproteinase-20 and kallikrein-4) (Bartlett et al., 2006; Bartlett and
Simmer, 1999; Bartlett et al., 1996; Simmer et al., 1998), it is conceivable that the
enamel organic matrix is completely degraded extracellularly and then removed
from this environment by macropinocytosis (fluid-phase endocytosis) (Smith,
1998). However, if a mechanism for enamel matrix protein removal involves a
significant uptake of partially degraded proteins (i.e. where the epitopes are
recognizable), and their subsequent trafficking to late endosomes and
lysosomes, then the presumption must be that ameloblasts absorb secreted
proteins (and the debris of these secreted proteins) during amelogenesis
(Franklin et al., 1991; Nanci et al., 1996; Nanci et al., 1987; Reith and Cotty,
1967b). This type of resorption would then be described more as a receptor-
mediated endocytosis.
Polarized ameloblasts form a continuous monolayer over the forming
enamel, with only their Tomes’ processes (secretory surface) in direct contact
with the extracellular matrix. If endocytosis or macropinocytosis is a functional
and significant characteristic of secretory ameloblasts, then uptake is likely to
occur at, or proximal to, the ameloblast apical membrane (Franklin et al., 1991).

6
One of the challenges in studying the removal of the enamel matrix in vivo is to
be able to distinguish between macropinocytosis and endocytosis and the
relative contributions of each pathway to amelogenesis. Because of previously
identified protein-interactions between LAMP1 and amelogenin (Tompkins et al.,
2006), and CD63 and amelogenin (Wang et al., 2005), we sought to identify a
possible pathway for the endocytosis of amelogenin in cell lines of varying
phenotypes, and relate our observations to the events of amelogenesis.

MATERIALS AND METHODS

Cell lines and culture conditions:
The following cell lines have been used in this study: canine kidney MDCK
cells (ATCC catalogue # CCL-34); human osteoblast hFOB_1.19 (ATCC
catalogue #CRL-11372); mouse pre-osteoblast MC3T3-E1 (ATCC catalogue #
CRL-2595); and mouse ameloblast-like LS8 cells (Dhamija and Krebsbach,
2001; Shapiro et al., 2006; Zhou and Snead, 2000). Multiple cell lines have been
included in this study to demonstrate that the cellular absorption of amelogenin
epitopes is not limited to ameloblasts or ameloblast-like cells, but that this activity
is a more general phenomenon seen in many cell types. In addition, it was
necessary to exclude crinophagy (Nanci et al., 1985; Smith, 1998) as an
explanation for the observations of amelogenin in the lysosomes. Crinophagy is

7
the digestion of the contents of secretory granules following their fusion with
lysosomes.
LS8 cells were originally derived from the first molar enamel organ
epithelium cells of newborn mice (Chen et al., 1992) and have previously been
used to study amelogenin (Xu et al., 2006c; Xu et al., 2006d; Zhou et al., 2000;
Zhou and Snead, 2000) and ameloblastin (Dhamija and Krebsbach, 2001;
Dhamija et al., 1999) gene expression in vitro.
Canine kidney MDCK and mouse ameloblast-like LS8 cells were
maintained in Dulbecco’s modification of Eagles medium (DMEM) with high
glucose (4.5g/L) supplemented with 10% (vol/vol) fetal calf serum (FCS). Human
osteoblast hFOB_1.19 cells were maintained in a 1:1 mixture of Hamm's F12
medium and Dulbecco's modified Eagle's medium with 2.5 mM L-glutamine, 0.3
mg/ml G418, and supplemented with 10% (vol/vol) fetal calf serum (FCS).
Mouse pre-osteoblast MC3T3-E1 were maintained in Alpha minimum essential
medium with ribonucleosides and deoxyribonucleosides, 2 mM L-glutamine, 1
mM sodium pyruvate, and supplemented with 10% (vol/vol) fetal calf serum
(FCS).

Plasmid constructs:
pAmel-DsRed: Mouse amelogenin M180 cDNA (Snead et al., 1985;
Snead et al., 1983) (NCBI reference sequence NM_009666), including the signal
peptide region, was PCR amplified from a plasmid template using the following
primers: forward 5’- GAATTCAAGAAATGGGGACCTGGATT, and reverse 5’-

8
GGATCCACTTCTTCCCGCTTGGT. The forward primer included an Eco RI
restriction site sequence (underlined), and the reverse primer a Bam HI
restriction site (underlined) to allow for efficient, in-frame subclonings in
subsequent steps. This PCR product was subsequently sub-cloned into a TA
cloning vector (pCR

2.1; Invitrogen Corporation), released with restriction
enzymes Eco RI and Bam HI, and subsequently subcloned into the vector
pDsRed1-N1 (Clontech, Palo Alto, CA) at the Eco RI / Bam HI multicloning site.
A two-step cloning strategy was used for this plasmid construct to ensure
efficient subcloning of the PCR product, prior to its sequencing and movement
into the red fluorescent vector that is not a TA cloning vector.
pLAMP1-GFP: Human LAMP1 cDNA (NCBI reference sequence
NM_005561) was purchased from Origene (Rockville, MD), and amplified by
standard PCR techniques using the following primers: forward 5’-
GATATCCTCGGCATGGCGCCCCGC, and reverse 5’ –
GATAGTCTGGTAGCCTGCGTGACTCC. PCR amplified LAMP1 cDNA was
subcloned into vector pcDNA3.1/CT-GFP-TOPO (Invitrogen Corporation,
Carlsbad, CA), and orientation was determined by restriction mapping.
For both plasmid constructs, the fluorescent tag follows the entire coding
regions for amelogenin or LAMP1. In both cases the entire coding regions were
sequenced to identify correct orientation, to ensure that the correct open-reading
frame was established, and that no PCR or cloning errors had occurred during
their synthesis.

9
Porcine enamel matrix proteins:
Emdogain® (previously owned by Biora AB, Malmö, Sweden but now
distributed by Straumann, Basel, Switzerland; www.straumann.com) is a porcine
derived enamel matrix product and the principle component is the enamel matrix
protein amelogenin (Maycock et al., 2002). Lyophilized porcine enamel protein
(Emdogain®) was diluted to a final concentration of 50 µg/ml in sterile culture
medium, and added freshly to cells.

Antibodies:
LAMP1: A rat monoclonal antibody to mouse LAMP1 (Developmental
Studies Hybridoma Bank or DSHB, University of Iowa; catalogue # 1D4B) was
used in this study.
CD63: A rabbit anti-peptide polyclonal mouse CD63 antibody was
generated against a unique CD63 peptide region (amino acids 177-190; N-
terminal CGNDFKESTIHTQG) by Zymed Laboratories Inc. (South San
Francisco, CA, USA).
Mouse amelogenin: Chicken egg-derived polyclonal antibody prepared
against recombinant mouse amelogenin rp(H)M180 (Moradian-Oldak et al.,
2000) described previously (Zhou and Snead, 2000; Zhu et al., 2006), has been
used in this study.

10
Stable transfection of pAmel-DsRed into MDCK cells and crude
fluorescent-protein extracts:
Plasmid pAmel-DsRed was successfully integrated into the genome of
canine kidney MDCK cells using Lipofectamine2000 (Invitrogen) and standard
laboratory methodologies (Sambrook and Russell, 2001). Cells were passaged
multiple times and continued to express significant quantities of the red-
fluorescent amelogenin protein as observed with fluorescent microscopy.
Red fluorescent MDCK cells were grown to confluence in a 60 mm cell
culture dish (Corning Incorporated, Corning, New York), harvested, pelleted,
resuspended in a minimal volume of fresh culture medium and then was
subjected to repeated freeze-thawing cycles in 0.25M Tris-Cl pH 7.8. The
complete lysate was then filtered through a 0.45 µm membrane (Corning Inc.) to
prepare a crude protein extract that included the red-fluorescent amelogenin.
Crude protein extract, plus Emdogain® to bring the total concentration of
exogenously added protein to 50 µg/ml, was then added to subconfluent human
osteoblast hFOB_1.19 cells also in a 60 mm cell culture dish, and left for 3-hours.
Cells were subsequently washed 3 times in phosphate-buffered saline at room
temperature (RT) prior to their visualization using a fluorescent microscope.

Immunohistochemistry:
Antibodies against LAMP1 and CD63 were used to examine the
spatiotemporal expression of these two proteins in mandibular incisor teeth in 3-
day and 4-day old mice. Tissues were prepared as previously (Ausubel et al.,

11
1990; Luo et al., 2004; Paine et al., 2000a; Snead et al., 1996). Slides were
counterstained with hematoxylin prior to making a digital photographic recording.

Immunofluorescence:
For CD63 and amelogenin colocalization experiments mouse MC3T3-E1
cells were grown on glass cover-slips to approximately 70% confluence.
Emdogain® was added to the media to a final concentration of 50 µg/ml and
incubated for 3-hours. Cells were then washed with phosphate buffered saline
(PBS), fixed with 2% paraformaldehyde/0.2% TritonX-100 for 10-minutes at room
temperature (RT), washed with PBS, and blocked 30-minutes with goat serum
(Zymed Laboratories, San Francisco, CA). Cells were incubated overnight at
4°C with chicken anti-mouse amelogenin antibody and rabbit anti-mouse CD63
(at dilutions of 1:2,500 and 1:200 respectively). After PBS washes, goat anti-
chicken Alexa Fluor 594 and goat anti-rabbit Alexa Fluor 488 (Invitrogen
Corporation; both at dilutions of 1:250) were used for fluorescent detection of
amelogenin and CD63 respectively. Cells were left to incubate for 1-hour at
37°C. Finally, cells were washed with PBS for three times and then mounted
with Pro-Long anti-fade reagent (Molecular Probes). Cells were visualized under
confocal microscopy at 60x magnification using a Nikon PCM2000 confocal
system.
For LAMP1 and amelogenin colocalization studies two experimental
approaches were used; either unlabeled Emdogain® was added to cultured
MC3T3-E1 cells, or Emdogain® that had been conjugated to Alexa Fluor 594

12
(Invitrogen Corporation, Catalogue # A30008) was added to cultured LS8 cells,
and studied in an identical manner to that described immediately above.
Visualization of LAMP1 was by the monoclonal anti-rat LAMP1 primary antibody
and goat anti-rat Alexa Fluor 488 used at dilutions of 1:100 and 1:250
respectively. Visualization of intracellular Emdogain® in MC3T3-E1 cells
followed primary and secondary antibody labeling. Visualization of Alexa Fluor
594-conjugated Emdogain® in LS8 cells required no additional antibodies.

Transient transfection and microscopy:
For transient transfection assays, MC3T3-E1 or LS8 cells were grown on
either glass cover slips in 3.5 cm cell culture dishes or on 2-well Lab-
Tek chamber slides (Nalge Nunc International, Rochester, New York, USA).
Cells were transfected with pLAMP1-GFP using Lipofectamine Plus (Invitrogen
Corporation) according to the manufacturer’s instructions. The day after
transfection, 50 µg/ml of Emdogain® conjugated to Alexa Fluor 594 (Invitrogen
Corporation, Catalogue # A30008) was freshly added to the culture medium for
3-hours. Cells were then washed with PBS, fixed with 4% paraformaldehyde,
washed again with PBS and mounted with Pro-Long anti-fade reagent (Molecular
Probes). Images were taken from a 63x Carl Zeiss Plan Apo objective of an
Olympus IMT-2 microscope and/or 60x lens of a Nikon PCM2000 confocal
system, respectively.

13

Time-lapse confocal microscopy:
MC3T3-E1 cells transfected with pLAMP1-GFP were grown to 70%
confluency on 35mm glass bottom culture dishes MatTek Corporation (Ashland,
MA., Catalogue #P35G014C), and transfected with pLAMP1-GFP using
Lipofectamine Plus (Invitrogen Corporation) according to the manufacture’s
instructions. After 24-hours, Emdogain® conjugated to Alexa Fluor 594
(according to manufacturer’s instructions; Invitrogen Corporation) was freshly
added to the culture medium to reach a final concentration of 50 µg/ml. Images
were taken from a 100x lens of a Nikon PCM2000 confocal system at 7-second
intervals.

RT-PCR:
Messenger RNA (mRNA) was isolated from LS8 cells, and used as
template RNA for a RT-PCR reaction using the Titanium
TM
One-step RT-PCR kit
(Clontech Laboratories Inc, Mountain View, California) and murine-specific CD63
(exon 3 forward 5’- CATTGGTGTAGCGGTTCAGGTTG and, exon 6 reverse 5’-
CATTCCCACAGCCCACAGTTATG) and LAMP1 (exon 4 forward 5’-
GTGACCGTTGTGCTCCGGGATGCC and, exon 8 reverse 5’-
CCCTGCCAGGGCACCGCCCAC) primers. A primer set for murine -actin
(exon 3 forward 5'- CTGGCACCACACCTTCTACAATG and, exon 4 reverse 5’-
GATGTCACGCACGATTTCCCTC) was also included as a control for RT-PCR. A

14
DNA ladder (Promega, Madison, WI: catalogue #G210A) was used to confirm the
amplified DNA size in base pairs (bp). Subsequently the entire PCR-generated
products for CD63 and LAMP1 were sequenced to confirm their exact identities.

RESULTS

Internalized red-fluorescent amelogenin localizes to the perinuclear region
of osteoblasts:
Exogenously added red-fluorescent amelogenin, produced in vivo from
MDCK cells that had been stably transfected with pAmel-DsRed (Fig. 1A), was
effectively endocytosed by the human osteoblast hFOB1_19 cells. This uptake
was fairly rapid and could be visualized in hFOB1_19 cells after 3-hours
exposure to amelogenin. A 3-hour time course for this, and the subsequent
experiments, was chosen based on a similarly designed and previously
published study (Reseland et al., 2006). The majority of the fluorescent-tagged
amelogenin protein detected in the hFOB1_19 cells was localized to the
perinuclear region (Fig. 1B), a region consistent with the passage of amelogenin
from the culture medium to the late endosomes and lysosomes.

15
Figure 1. Panel A: MDCK CELLS
STABLY TRANSFECTED WITH
AMELOGENIN-RED FLUORESCENT
PROTEIN CONSTRUCT AS SEEN
UNDER FLUORESCENT
MICROSCOPY. Inset: schematic of
hybrid fluorescent protein (pAmel-
DsRed) involving the amelogenin
signal peptide (s.p.), mouse
amelogenin (mAmel) and red
fluorescent protein (DsRed1). Panel
B: ENDOCYTOSIS OF THE
AMELOGENIN-RED FLUORESCENT
PROTEIN IN HUMAN OSTEOBAST
CELLS. Human osteoblast cells
(hFOB_1.19) were exposed to a crude
preparation of protein isolated from the
cells MDCK cells. The fluorescent
protein preparation was from an
approximately equal number of cells to
which the exogenous protein is to be added. In addition, porcine enamel matrix
protein was included to bring the total concentration of exogenously added
protein to 50µg/ml. Cells were left for 3 hours prior to observing the endocytosed
amelogenin proteins. Cells were washed 3 times in phosphate-buffered saline at
RT, and the amelogenin red-fluorescent hybrid protein localized to the cell
cytoplasm (arrows), but is excluded from the nucleus. Scale bar 10µm.

CD63 is expressed by ameloblasts at all stages of amelogenesis, but is
more highly expressed in late-secretory and post-secretory ameloblasts:
Immunohistochemistry has been used to examine the distribution CD63 in
developing mice mandibular incisor teeth. Data shows that CD63 is immuno-
localized to the Tomes’ processes of ameloblast cells, and is also unevenly
distributed throughout the cytoplasm of ameloblast cells and all surrounding
tissues (Fig. 2). Higher levels of CD63 expression are seen in late-stage
amelogenesis (ameloblasts associated with maturing enamel) when compared to

16
secretory or pre-secretory ameloblasts. The presence of CD63 in ameloblasts at
specific locations is suggestive of a functional role for CD63 in enamel formation.

Figure 2. IMMUNOLOCALIZATION OF CD63 TO AMELOBLAST CELLS IN A
3-DAY OLD MOUSE MANDIBULAR INCISOR: Panel A: No primary antibody
control section. No staining is seen in the control section. Panels B-F: A rabbit
anti-peptide polyclonal CD63 antibody was generated against a unique peptide
region of mouse CD63 (amino acids 177-190; N-terminal CGNDFKESTIHTQG)
by Zymed Laboratories Inc. (South San Francisco, CA, USA) and used at a
dilution of 1:30. Panels C-F: Magnification of sections approximating those
boxed regions identified in panel A. The dominant location of immuno-staining
(brown chromophore) within ameloblasts is cytoplasmic with immuno-staining
also apparent at Tomes’ processes (arrows). Pre-secretory ameloblasts are
located at the incisor’s growing end, secretory ameloblasts are within and beyond
the transition zone up to the post-secretory ameloblasts. Post secretory
ameloblasts are located at the mature end of the incisor teeth. All sections
counterstained with hematoxylin prior to photographing. Abbreviations:
ameloblasts (Am); odontoblasts (Od); stratum intermedium (Si); mature end of
incisor (m); transition zone ameloblasts (tz); growing end of incisor (ge); and
dental pulp (p). Panels A and B are taken at 4x magnification, while panels C-F
are taken at 40x magnification.

17
CD63 and amelogenin colocalization:
This experiment was performed to determine the intracellular spatial
relationship of endogenous CD63 to exogenously added amelogenins that had
been phagocytosed and/or endocytosed by MC3T3-E1 cells. Emdogain® was
added to the culture medium of mouse MC3T3-E1 cells to a final concentration of
50 µg/ml, and left for 3-hours prior to microscopic observation using antibodies
specific to mouse CD63 and mouse amelogenin. Immunoreactive amelogenin
epitopes could be recognized in the perinuclear region of MC3T3-E1 cell (Fig. 4).
This data demonstrates that recognizable amelogenin epitopes survive the
trafficking process from the culture medium to the cell perinucleus.

Figure 3. CD63 (GREEN) AND
EXOGENOUSLY ADDED
PORCINE AMELOGENIN (RED)
COLOCALIZATION IN THE
PERINUCLEAR REGION OF
THE MOUSE PRE-
OSTEOBLAST CELL LINE
MC3T3-E1: CD63 is distributed
throughout the cytoplasm, but
more highly concentrated in the
perinuclear region. Amelogenin
(red) is localized to the
perinuclear region.
Colocalization of CD63 and
amelogenin is seen as yellow
(arrows). White, interrupted line defines the nuclear membrane of a single cell.
Image taken at 60x magnification. A double arrow-head indicates expression of
CD63 at the plasma membrane.

18
LAMP1 is expression by ameloblasts at all stages of amelogenesis, but is
more highly expressed in early secretory stages:
Immunohistochemistry has been used to examine the distribution LAMP1
in developing mice mandibular incisor teeth. Data shows that LAMP1 is immuno-
localized to the region of Tomes’ processes of pre-secretory ameloblast cells,
and LAMP1 is also fairly evenly distributed throughout the cytoplasm of
ameloblast cells and surrounding tissues (Fig. 4). Higher levels of LAMP1
expression are seen in the early-stages amelogenesis (non-polar and pre-
secretory ameloblasts) when compared to secretory or post-secretory
ameloblasts. The presence of LAMP1 in ameloblasts is suggestive of a
functional role for LAMP1 in enamel formation, but this statement is made with
the proviso that LAMP1 has a ubiquitous tissue expression profile (Andrejewski
et al., 1999; Bonifacino and Traub, 2003; Eskelinen et al., 2002) and that the
presence of LAMP1 in ameloblasts was anticipated.

19

Figure 4. IMMUNOLOCALIZATION OF LAMP1 TO AMELOBLAST-LIKE LS8
CELLS IN A 4-DAY OLD MOUSE MANDIBULAR INCISOR. Panel A: No
primary antibody control section. No staining is seen in the control section.
Panels B-F: Antibody to rat LAMP1 used at a dilution of 1:50. Panels C-F:
Magnification of regions approximating those boxed regions identified in panel A
(from left to right). Pre-secretory ameloblasts (F) are located at the incisor’s
apical end, and the secretory ameloblasts (E, D) are prior to the transition zone.
Post-secretory ameloblasts (C) are located at the incisal end of the tooth.
LAMP1 is clearly and evenly expressed in the cytoplasm of ameloblasts at all
stages of amelogenesis (brown chromophore), and most highly expressed in the
secretory ameloblasts within the transition zone of enamel formation (panel E).
Immuno-staining is apparent at Tomes’ processes (arrows). High expression of
LAMP1 is also noted at the basal poles of secretory ameloblasts (asterisk; panel
E). Abbreviations: ameloblasts (Am); odontoblasts (Od); stratum intermedium
(Si); incisal end of incisor (inc); transition zone ameloblasts (tz); apical end of
incisor (ap); and dental pulp (p). All sections are counterstained with hematoxylin
prior to photographing. Scale bars are included in panel B (for A and B) and
panel F (for C-F).

20
LAMP1 and amelogenin colocalization:
This experiment allowed us to determine the intracellular spatial
relationship of endogenous LAMP1 to exogenously added amelogenins that had
been phagocytosed and/or endocytosed by MC3T3-E1 (Figs. 5A-C) and LS8
(Fig. 5D) cells. Unlabeled Emdogain® was added to the culture medium of
mouse MC3T3-E1 cells to a final concentration of 50 µg/ml, and left for 3-hours.
The cells were then immediately processed for immunofluorescent microscopic
observation using antibodies to LAMP1 and amelogenin (Figs. 5A-C).
Emdogain® conjugated to Alexa Fluor 594 (Invitrogen Corporation) was added to
LS8 cells previously transfected with pLAMP1-GFP. Data from MC3T3-E1 and
LS8 cells are similar and demonstrate that LAMP1 expression is limited to the
cell cytoplasm, and more specifically to the perinuclear regions (Fig. 5). This
localization is the case for both endogenous LAMP1 (Figs. 5A-C) or transfected
human LAMP1 (Fig. 5D). At higher magnification the bulk internalized
amelogenin was seen completely contained within the vacuole space of LAMP1-
positive vesicles (see inserts contained within Figs. 5B-D shown at higher
magnification). These data demonstrate that exogenously derived amelogenin
traffics to the cell perinucleus, and more specifically is localized to late
endosomes and lysosomes.

21

Figure 5. COLOCALIZATION OF LAMP1 AND EXOGENOUSLY ADDED
PORCINE ENAMEL MATRIX PROTEINS IN THE PERINUCLEAR REGION OF
MOUSE PRE-OSTEOBLASTS MC3T3-E1 AND AMELOBLAST-LIKE LS8
CELLS. Panel A: Combined differential interference contrast light microscopic
and immunofluorescent image of MC3T3-E1 cells showing both the LAMP1
(green) and amelogenin (red) localization is exclusively to the perinuclear region.
Panels B and C: Immunofluorescent confocal images of endogenous LAMP1
(green) and exogenously-derived amelogenin (red) localized in MC3T3-E1 cells.
Arrows point to amelogenin-containing LAMP1 coated vesicles. Panel D:
Confocal images of a LS8 cells transfected with pLAMP1-GFP, and with the
addition of Alexa Fluor 594 labeled Emdogain® Panels B-D: The arrows point to
amelogenin-containing (B, C) and porcine Emdogain® containing LAMP1 coated
vesicles identified as late endosome/lysosome compartments. Colocalization of
LAMP1 and amelogenin is seen as yellow. The white line is used to define the
plasma membrane and the white interrupted line is used to define the nuclear
membrane. Scale bar is included in panel D (for all panels). Inset images are
enlarged 2.5x.

The transfer of extracellular amelogenin into the cell cytoplasm involves
the direct passage of amelogenin into LAMP1-positive vesicles.
Fluorescently labeled enamel matrix proteins (Emdogain® conjugated to
Alexa Fluor 594) could be seen entering directly into LAMP1-positive vesicles
that were located immediately subjacent to the plasma membrane. A significant
aspect of this internalization process occurs in seconds as seen by rapid
movement of the enamel matrix proteins into LAMP1-positive vesicles. For
example, figure 6 shows a LAMP1-positive vesicle approximately 90-minutes

22
after the addition of Emdogain® (Panels B and J, frame #24). Emdogain® is
seen entering this LAMP1-positive vesicle during a 14-second time-lapse (Panels
C and K, frame #26). By observing this same vesicle for an additional 406-
seconds (approximately 7-minutes), this vesicle contracted to 40% of its original
diameter (Panels H and L, frame #84).
We can conclude from this data that the shrinking of the Emdogain®
containing LAMP1 vesicles (immediately following Emdogain® uptake) is likely to
relate directly to the rate at which the enamel matrix proteins are degraded.

23

Figure 6. THE MOVEMENT OF ALEXA FLUOR 594 LABELED EMDOGAIN®
INTO MOUSE PRE-OSTEOBLASTS (MC3T3-E1) INVOLVES DIRECT
PASSAGE INTO PRE-ESTABLISHED LAMP1-COATED VESICLES. Time-
lapse confocal microscopy was used to study the movement of exogenously
added enamel matrix proteins into MC3T3-E1 cells expressing pLAMP1-GFP.
Approximately 90-minutes after the addition of the Alexa Fluor 594-conjugated
enamel matrix proteins into the culture medium, a single cell was observed at 7-
second intervals for a total of 588seconds (Panels A-H, where the numbers seen
in the top right indicate the frame number). In panel A the cell nucleus is
highlighted with an interrupted white line, and a single boxed region containing a
LAMP1-coated vesicle (also arrowed; panels B and C) subjacent to the plasma
membrane is identified. This boxed region is enlarged 10x in panels I-L, and
shows the same vesicle (3 parallel arrows; panel I) taken from frames 13, 24, 26
and 84 respectively. The two parallel arrows in panel I identify the plasma
membrane positive for LAMP1. No red fluorescence is noted in the LAMP1-
coated vesicle in frame 24, but is present in frame 26 which is taken 14-seconds
after frame 24. An evenly timed, and significant shrinkage of this red-fluorescent
LAMP1-coated vesicle occurred in an approximate 7-minute interval (from frame
26 to 84). Scale bars are included in panels H (for A-H) and L (for I-L).

24
Ameloblast-like LS8 cells express CD63 and LAMP1 as determined by RT-
PCR:
RT-PCR was used to validate the use of LS8 cells to study amelogenin
trafficking from the extracellular space to intracellular domains, and then relate
this activity to the spatiotemporal profiles of CD63 and LAMP1 in vitro. CD63
and LAMP1 gene transcripts were identified in LS8 cells initially by gel-
electrophoresis (Fig. 7), and their identity confirmed by the sequencing of these
PCR-generated products (data not shown).
Figure 7. RT-PCR OF CD63 AND
LAMP1 GENE TRANSCRIPTS IN
AMELOBLAST-LIKE LS8 CELLS:
Murine-specific CD63, LAMP1 and
-actin primers were used to amplify
cDNA prepared from LS8 cells. Lane
4 is CD63 amplified cDNA at 449bp.
Lane 5 is LAMP1 amplified cDNA at
672bp. Lane 6 is -actin amplified
cDNA at 382bp. Lanes 2 and 3 are control lanes using CD63 and LAMP1 primer
sets (respectively) with no added DNA template for the PCR. Lanes 1 and 7 is a
100bp DNA ladder.

DISCUSSION

Amelogenin is the most abundant protein in the developing mammalian
enamel extracellular matrix, accounting for greater than 90% of the total matrix
protein content (Deutsch et al., 1991; Margolis et al., 2006; Robinson et al., 1998;
Termine et al., 1980). Mature enamel is almost entirely inorganic, thus the

25
process of amelogenesis must include an efficient mechanism for the removal of
the organic matrix component and related organic debris. One potential pathway
for their removal is through endocytosis and post-endocytotic degradation. In
this study, evidence from tissue and cultured cells was presented that support a
role for LAMP1 and CD63, membrane markers for late endosomes and
lysosomes, in the endocytosis and post-endocytotic processing of amelogenin.
Based upon immunohistochemical staining, LAMP1 and CD63 appear to be
expressed at significant levels in ameloblasts in vivo. Moreover, both proteins
localize to Tomes’ processes of ameloblasts, placing these proteins in proximity
with extracellular amelogenin, the putative cargo for endocytosis. In cultured
ameloblast-like LS8 cells, PCR analysis confirmed the expression of LAMP1 and
CD63. Based upon the results of the endocytosis studies performed here on
these LS8 and MC3T3-E1 cells, they may represent a good model system to
characterize the endocytosis and processing of amelogenin in a LAMP1-
dependent and CD63-dependent manner. We have shown here that
endocytosed amelogenin can be detected and appears to be concentrated in a
subset of either endogenous LAMP1-positive and CD63-positive membranes, as
well as in LAMP1-positive and CD63-positive membranes in cells transfected
with either of these proteins. In addition, the intracellular distribution of
endocytosed amelogenin in select vesicles is consistent with specific uptake and
post-endocytotic processing of amelogenin, rather than endocytosis by a fluid-
phase mechanism, since uptake of fluid-phase markers typically results in a more

26
uniform labeling of endosomal and lysosomal membranes after extended periods
of endocytosis performed here.
A previous study has presented evidence supporting a receptor-like function
of LAMP1 for amelogenin, by demonstrating binding of amelogenin to LAMP1 in
biochemical assays and in cell surface binding assays (Tompkins et al., 2006).
The data here confirm and extend those findings by showing that endocytosed
amelogenin accumulates in LAMP1-positive organelles, sometimes within a very
rapid time frame. Taken together, these two studies suggest that LAMP1 may
serve as a cell surface receptor for amelogenin, and LAMP1 may be involved
with the trafficking of amelogenin to late endosomes or lysosomes. To our
knowledge, these data are the first evidence that LAMP1 may function as a
specific, endocytosing ligand-binding receptor at the cell surface.
We have shown previously by yeast two-hybrid analysis that CD63 interacts
with amelogenin (Wang et al., 2005). Here, we have shown that endocytosed
amelogenin also accumulates in CD63-positive organelles. Together, these data
suggest that CD63 may serve a role similar to that of LAMP 1 in the endocytosis
or post-endocytotic processing of amelogenin. CD63 may function either
cooperatively with LAMP1, for example, as essential proteins in the biogenesis of
endocytotic organelles responsible for amelogenin uptake and processing.
Alternatively, CD63 may function independently of LAMP1 in this process. Either
scenario could account for the overlapping, yet slightly distinct distributions of
LAMP1 and CD63 in ameloblasts observed by immunohistochemistry of tissue
sections.

27
CD63 has been implicated in the regulation of trafficking of other
transmembrane proteins (Duffield et al., 2003; Jung et al., 2006b), although
LAMP1 is not one of them, as mutations in CD63 that affect its localization do not
appear to affect LAMP1 localization (Rous et al., 2002). However, to our
knowledge, the data presented here would be the first to suggest that CD63 itself
may be a cell surface ligand-binding receptor, but further functional
characterization of the endocytosis of amelogenin in a CD63-dependent fashion
will be required before CD63 can actually be defined as a cell surface receptor
for amelogenin.
LAMP1 and CD63 typically recycle between the Golgi and late
endosomes/lysosomes (Eskelinen et al., 2003). However, a fraction of both of
these proteins have been also shown to be present at the cell surface. It has
been proposed that these two proteins may traffic through the plasma membrane
as part of their normal itinerary, on their way to their ultimate steady-state
distribution to late endosomes and lysosomes, although no function has been
ascribed to their presence at the plasma membrane. It would be interesting to
define whether LAMP1 and CD63 in Tomes’ processes in ameloblasts are
actually at the plasma membrane, perhaps by immunolocalization at the
ultrastructural level. Some of the subcellular machinery involved in the trafficking
of LAMP1 and CD63 has also been characterized, particularly with respect to
trafficking dependent upon clathrin and associated proteins (Di Pietro et al.,
2006; Eskelinen et al., 2003). Both appear to interact with AP-2 clathrin adaptors
at the plasma membrane to regulate their endocytosis and with AP-3 adaptors at

28
endosomes to regulate their trafficking to late endosomes/lysosomes (Bonifacino
and Traub, 2003; Eskelinen et al., 2003). Thus, if LAMP1 and CD63 indeed
serve a receptor-like function, the putative machinery for their endocytotic
trafficking should also be present at significant levels in ameloblasts and LS8
cells. A corollary to this line of investigation is that there are mice strains
deficient in AP-3, such as mocha (Kantheti et al., 1998) and AP3B1 null(Feng et
al., 1999), and humans that lack AP-3 resulting in Hermansky-Pudlak syndrome
(Dell'Angelica et al., 1998; Di Pietro and Dell'Angelica, 2005); it would be
interesting to determine whether amelogenesis is defective in these mice and
humans. A more direct test of whether LAMP1 and CD63 are involved in
amelogenesis would be to eliminate genetically either or both of these proteins.
LAMP1 has been deleted in mice (Andrejewski et al., 1999), but there appears to
be some functional redundancy between LAMP1 and the related protein LAMP2
(Andrejewski et al., 1999; Eskelinen et al., 2004). LAMP2-deficient mice show a
more marked phenotype, and the double-knockout of LAMP1 and LAMP2 is
embryonic lethal (Eskelinen et al., 2003). It is not known whether LAMP2 can
bind to amelogenin or to compensate in vivo for LAMP1 in amelogenesis.
There are no reports of organisms with CD63 deletions, but this knock-out may
also prove lethal.
There are other possibilities with respect to the role of LAMP1 and CD63 in
amelogenesis. With the identification of two secreted, enamel-specific
proteinases (matrix metalloproteinase-20 and kallikrein-4) (Bartlett et al., 2006;
Bartlett and Simmer, 1999; Hu et al., 2002; Woessner, 1998), it is conceivable

29
that fragments of partially-degraded amelogenin are endocytosed by LAMP1
and/or CD63; however, these endocytosed fragments would be restricted to
those containing the binding site for LAMP1 or CD63. Another possibility is that
the enamel organic matrix is completely degraded extracellularly and then
removed from this environment by macropinocytosis or fluid-phase endocytosis
(Reith and Cotty, 1967a; Smith, 1979). However, currently there is no evidence
that CD63 or LAMP1 are involved in regulating macropinocytosis or fluid-phase
uptake, and our data suggest that the uptake into cultured ameloblasts is not
uniformly distributed among the various endocytotic organelles, as would be
expected for fluid-phase uptake. Macropinocytosis is potentially a mechanism by
which the internalization of amelogenin may occur directly into relatively large
LAMP1-positive and CD63-positive vesicles. A precedent for this type of
internalization is provided by the macropinocytosis of the epidermal growth factor
receptor and the platelet-derived growth factor receptor (King and Cuatrecasas,
1981; King et al., 1980; McNiven, 2006; Orth et al., 2006).
If LAMP1 and CD63 are endocytotic receptors for amelogenin, they could
also be operating in a novel fashion to initiate a signal transduction cascade
downstream of amelogenin binding. While the endocytosis of amelogenin has
been shown to regulate amelogenin gene expression in ameloblasts (Xu et al.,
2006a; Xu et al., 2006b), there are no reports of LAMP1 or CD63 regulating
signal transduction pathways. This system could provide an ideal opportunity to
test this hypothesis. In summary, we have provided evidence for a novel
association of LAMP1 and CD63 in the endocytosis and post-endocytotic

30
processing of amelogenin. These proteins may be critical to the early formation
of enamel by regulating the endocytosis of enamel matrix proteins.

31
Chapter 2.

The Transcriptional Responses of a Non-amelogenin Producing Cell to
Exogenous Enamel Matrix Proteins

As shown in the previous chapter, enamel matrix proteins are rapidly and
effectively endocytosed into LAMP positive vesicles of many cell types.
Historically, LAMPs were thought to only contribute to maintaining structural
integrity of the lysosomal membrane, but because our data shows colocalization
of LAMP coated vesicles and Emdogain®, we hypothesized that “upon treatment
with Emdogain®, cells would upregulate LAMPs and/or adaptor proteins for
effective clearing of enamel proteins”. Receptor up-regulation is a well known
process where cells increase transcription of a receptor in response to signals,
for example, epidermal growth factor receptor is up-regulated in cervical
cancers(Rajesh S. Mathur, 2000), and CD63 up-regulation in melanoma is
indicative of tumor progression(Hotta H, 1989).
To test the hypothesis, I decided to treat human fibroblasts, which rapidly
endocytose enamel matrix proteins, with 50 µg/ml of Emdogain® in serum free
media for 2, 3, 4, 6, and 24 hours. After the specified time period, the relative
transcription rates of LAMPs and AP-3 would be determined.
I chose to measure the σ subunit of AP-3, because its function is to
stabilize the entire AP-3 complex(Dell'Angelica EC, 1997; Simpson et al., 1997),

32
and therefore would be a good marker for any changes in transcription rates of
the whole AP-3 complex.

MATERIALS AND METHODS

Cells:
Human Fibroblast cells were obtained (ATCC # CRL-1613,
Manasass,VA), seeded into 6 well culture plates, and grown in Eagle’s Minimum
Essential Media supplemented with 20% fetal bovine serum. At 70% confluency
cells were treated with 50 µg/ml of Emdogain® for 2, 3, 4, 6, and 24 hours. Three
separate cell cultures were examined for each time-point.

Primer design:
LAMP1, CD63 and AP-3 sigma primers were designed by Beacon
Designer 5 software using the SYBR® Green Assay Design
(http://www.premierbiosoft.com ). GAPDH primer sequence was taken from the
RT primer database (RTPrimerDB ID : 972). Single melting curves for LAMP-1,
CD63 and AP-3 sigma were confirmed using plasmid templates.

33

Table 1. PRIMER PAIRS FOR LAMP1, CD63, AP-3 SIGMA AND GAPDH
REAL-TIME PCR EXPERIMENTS.

Real-time PCR:
Total RNA from each sample was extracted using RNAqueous (Ambion,
Austin TX) according to the manufactures protocol. cDNA was synthesized using
iScript cDNA synthesis kit (Bio-Rad, Hercules California). Quantitative analysis of
LAMP1, CD63 and AP-3 sigma subunit expression was measured in an iCycler
using iQ SYBR Supermix (cat #170884,Bio-Rad, Hercules, CA) according to the
manufactures protocol. Each sample was run in triplicate.
Three critical thresholds for each sample at a given time period were
averaged and normalized to a value of 1 for GAPDH at time=0, using the
following equation: 2^-((T-G)-Tu) , where T=treated , G=GAPDH, Tu=untreated.
LAMP1 CD63 AP-3 sigma GAPDH
Forward GCCACAGT
CGGCAATT
CCTA
AGCAGATG
GAGAATTAC
CC
GGACACAA
TCCAAATCA
TAAACCT
GTGAAGGT
CGGAGTCA
ACG
Reverse CTGAAAAC
GCCTTCGT
GACA

CTCCCAATC
TGTGTAGTT
AG

CACATGTTG
TCAGCTTTA
ACCTTC
TGAGGTCA
ATGAAGGG
GTC

34
The GAPDH normalized values for all three separate experiments were averaged
to give a final outcome.

RESULTS

Real-time PCR:
The transcriptional response of human fibroblasts to Emdogain®
treatment was measured to determine if there is an increase or decrease in
mRNA of LAMP1, CD63, or AP-3 at 2, 3, 4, 6, or 24 hours (Fig. 8). These
particular time-points were chosen to examine early and late responses. The
data shows no significant measurable change in the transcription of LAMP1,
CD63, or AP-3 sigma upon Emdogain® treatme

35
Human Fibroblast Response to Emdogain Treatment
0
0.5
1
1.5
Time
Normalized gene expression ratio
Lamp1
CD63
AP3 Sigma
Lamp1 0.95 0.91333333 0.93666667 1.03333333 0.99
CD63 0.89333333 0.89333333 0.98333333 1.04666667 0.93666667
AP3 Sigma 0.78333333 0.91333333 0.81666667 0.89 0.91
2h 3h 4h 6h 24h

Figure 8. HUMAN FIBROBLASTS DO NOT INCREASE TRANSCRIPTION OF
LAMPS OR AP-3 WHEN TREATED WITH EXOGENOUS ENAMEL MATRIX
PROTEINS: There is no significant increase or decrease in the expression of
LAMPs or AP-3 in human fibroblasts upon Emdogain® treatment. These data
points are the average of three samples, each run in triplicate and normalized to
a value of 1 for GAPDH at t=0 (not shown).

DISCUSSION

Cells routinely regulate expression of proteins in response to their
environments. We hypothesized that treating cells with a high concentration of
Emdogain® would cause an increase in the expression of the LAMPs as well as
the associated AP-3 needed for effective clearing of the excess enamel matrix
proteins. We chose a human fibroblast cell line that does not express any
endogenous enamel matrix protein but rapidly endocytoses Emdogain® (data

36
not shown) to ensure that any changes in expression were due to our
experimental conditions.
Our data shows that human fibroblasts do not change their expression of
LAMPs or AP-3 when treated with Emdogain®. This finding is not altogether
surprising as there is an abundance of all three of these proteins in cells and
thus, cells may utilize protein already present to clear the excess enamel matrix
proteins. An alternative explanation for our results is that cells have excess
transcripts of these proteins pre-made in the cytoplasm and upon being
challenged, translate these transcripts to increase protein levels.
Cells utilizing proteins or transcripts already present in the cytoplasm to
clear enamel matrix proteins would not be detected in a real time PCR
experiment, because real time PCR only detects changes in expression levels.
But it is clear that there is no obvious or significant change in expression levels of
LAMP proteins or AP-3 upon Emdogain® treatment.

37
Chapter 3.

AP3B1 Null Mice, a Mouse Model of Hermansky Pudlak Syndrome, have an
Abnormal Dentition

Hermansky Pudlak Syndrome (HPS) (Hermansky and Pudlak, 1959) is an
autosomal recessive disease linked to the following gene loci: HSP1,
AP3B1/HSP2, HSP3, HSP4, HSP5, HSP6, DTNBP1/HPS7 and BLOC1S3/HPS8,
and gene products from all 8 loci are involved with some aspect of endosome-
lysosome biosynthesis. AP3B1 is involved in protein trafficking to lysosomes or
specialized endosomal-lysosomal organelles. Patients with HPS have albinism,
bleeding problems, inflammatory bowel disease, and defects in many
cytoplasmic organelles including melanosomes, platelet-dense granules and
lysosomes. Although a rare disease worldwide, HPS is not uncommon in Puerto
Rico with a frequency of approximately 1:800 (or ~ 3,700 affected individuals),
giving a carrier frequency of 1:21. Significant populations of HPS sufferers are
documented in Switzerland and also Japan. The Medical Genetics Branch of the
NIH estimates that one in every 400-2,000 people worldwide are carriers of a
mutation causing HPS, and it is currently estimated that 1 in every 500,000-
1,000,000 people worldwide are sufferers of HPS (Gahl, 2007; Markello, 2008;
Witkop et al., 1990a; Witkop et al., 1990b). The biggest dental concern is
periodontal disease, often left untreated because of the excessive gingival
bleeding, and medical support and management is needed in patients for dental

38
surgeries (Feliciano et al., 2006). There is a report in the literature noting “severe
dental decay” in addition to “aggressive periodontitis” in two related individuals
(consanguineous pedigree) with HPS type II resulting from a mutation to the
AP3B1 gene (Jung et al., 2006; Kotzot et al., 1994). In addition, there is a
reference from the “emedicine” website (Izquierdo, 2008) describing HPS
patients as having abnormal enamel (a description that is unreferenced). I have
been unable to find any additional literature describing the tooth enamel and
dentin phenotypes in these patients.
HPS is an example of how mutations in a component of the sorting
machinery causes miss-sorting of integral lysosomal membrane proteins leading
to disease(Dell'Angelica et al., 1999). HPS patients have increased localization of
LAMP1, CD63 and LAMP2 on the plasma membranes of their cells(Dell'Angelica
et al., 1999) as compared to non affected patients because of the mutation to
AP-3.
LAMPs normally after modification in the Golgi apparatus, exit the trans-
Golgi and move to sorting endosomes from which they are targeted to the
lysosome by AP-3 interacting with the GYXXΦ sorting sequence found on the
LAMP cytoplasmic tails (Dell'Angelica et al., 1999). In HPS, abnormal amounts of
LAMPs are shunted from sorting endosomes to the plasma membrane via
recycling endosomes due to the absence of AP-3 directed trafficking.
Mouse models of HPS-like characteristics are available for study, most
notably the “pearl” mice with mutations to the AP3B1 gene locus. This locus
encodes the B1 protein contained within the adaptor protein-3 (AP-3) complex,

39
which is involved with receptor-mediated endocytosis. By all accounts pearl
animals have a relatively normal life and require no special dietary or caging
needs. Although pearl animals produce small litters and have a shortened
lifespan.
Enamel matrix proteins (EMPs) bind LAMPs, and all LAMPs directly
interact with the AP-3 complex to initiate receptor-mediated endocytosis. In
yeast, AP-3 associated endocytosis is clathrin-independent, while in higher
organisms the requirement of clathrin is unclear. AP-3 is a protein complex with
four subunits referred to as , ,  and . Three other adaptor protein complexes
(AP-1, AP-2 and AP-4), each having four similar but unique subunits, are also
involved with endocytosis and are either clathrin-dependent (AP-1 and AP-2) or
clathrin-independent (AP-4). Under normal circumstances, all-four adaptor
protein complexes operate independently of each other. Only AP-3 appears to
have the ability to interact directly with plasma membrane-bound LAMPs. It is the
 subunit of AP-3 that interacts directly with the internalized C-terminus of
LAMP1, LAMP2 and CD63 through a peptide motif common to LAMP1, LAMP2
and CD63. As mentioned above, mutations to the AP3B1 (or ) subunit can
result in HPS in humans. If endocytosis of the degraded EMPs plays a significant
role in enamel maturation, then AP-3 dysfunction should also result in enamel (or
dental) defects.

40
MATERIALS AND METHODS

Animals:
The animals used in this study, AP3B1 null (B6.Cg-Ap3b1
tm1.1Sms
/J, stock
#006253) are described in figure 9. Genetically matched control C57BL/6J (stock
#000664) mice were also purchased and maintained for these studies (Jackson
Laboratory, Maine, USA). Both AP3B1 null and C57BL/6J were maintained in
identical environments and fed the same diets.

41

Figure 9. TARGETING STRATEGY TO PRODUCE AP3B1 NULL MICE: A
targeting vector containing a loxP site flanked neomycin resistance and
phosphoglycerate kinase selection cassette and IRES-lacZ was used to disrupt
exon 5. The construct was electroporated into 129-derived R1-45 embryonic
stem (ES) cells. Correctly targeted ES cells were co-cultured with C57BL/6
morulae. The resulting male chimeric animals were crossed to C57BL/6 mice.
The mice were crossed to transgenic CMV-Cre (on a mixed C57BL/6J) to
remove the PGK-neo selection cassette. Mice were then backcrossed to
C57BL/6 for 5 generations (while selecting against the Cre-expressing
transgene) prior to arrival at The Jackson Laboratory. These mice, homozygous
for the targeted mutation, are viable, fertile, normal in size and do not display any
gross physical or behavioral abnormalities. No AP3B1 mRNA is detected by
Northern blot analysis of spleen and kidney tissue, and beta-3A immunoreactivity
is absent in monocytes from homozygous mice. In kidney, no sigma-3 protein is
detected, and mu-3 and delta-3 subunit proteins levels are greatly reduced.
Homozygotes have a diluted coat color (light gray), which is lighter than the coats
of homozygotes carrying the allelic AP3B1 null (Ap3B1
pe
) spontaneous mutation.
Cultured melanocytes from homozygous AP3B1 null mice have very few pigment
granules. Lysosomal-associated membrane proteins and tyrosinase are
mislocalized in cultured fibroblasts and melanocytes. Vesicular zinc in embryonic
fibroblasts from homozygotes is either greatly reduced or absent. Homozygotes
also exhibit prolonged bleeding times. (Jackson Laboratory, Maine, USA)

42
Reverse transcriptase PCR of adaptor protein β1-4 subunits in ameloblast-
like (LS8) cells:
Cells were cultured in Dulbecco’s Modified Eagle’s Media and incubated
at 37°C with 5% CO2. At confluency, cells were collected and RNA extracted
using RNAqueous (Ambion, Austin TX) according to the manufactures protocol.
cDNA was synthesized using iScript cDNA synthesis kit (Bio-Rad, Hercules
California). Specific primers (AP3B1-F:GTCTACCACAAGCCTCCCAA AP3B1 -
R:CCCCAAAGTTGCTATCTCCA (304bp); AP2B1-F
:CAATCAGCAGAACGCTGTGT AP2B1- R:TCAGGGTTATCGGAGTCCTG
(398bp) ; AP3B1-F :CCCACTTGGGAAAACAAAGA AP3B1-
R:TCCATGGAACTGGACTCCTC (375bp) ; AP4B1-F
:GCTGGGACTTCGACAAGAAC AP4B1-R:TATGCAGTAATGCAGCAGCC
(343bp) were used in a standard PCR reaction (Titanium Taq Polymerase (cat
#639209, Clonetech, Mountain View, CA) to amplify the different beta subunits of
all four adaptor protein complexes. PCR product was run in a 1% agarose gel
and visualized under UV light.

Genotyping:
Tail clippings from AP3B1 null and C57BL/6J mice were digested in 10µl
proteinase K diluted into DirectPCR buffer (cat #102, Viagen, Los Angeles, CA)
according to the manufacturers protocol. Clippings were incubated for 45 minutes

43
at 85°C. 2µl of recovered DNA in solution was run in a standard PCR reaction
(Titanium Taq Polymerase (cat #639209, Clonetech, Mountain View, CA) using
the following primers: forward 5'- TCAACTAATCCGTGCAAGTG -3' ,
reverse 5'- TATCCCTGGGAGATCTGCTC -3' (product is 324 base pairs
encoding a portion of exon 5 found in the AP3B1 gene) and forward 5'-
GGGTTGTTACTCGCTCACA -3' , reverse 5'- AAAGCGAGTGGCAACATGG
-3' (product is 290 base pairs encoding a portion of the LacZ gene from the
insert). PCR product was run in a 1% agarose gel and visualized under UV light.

Scanning electron microscopy:
Six week old AP3B1 null (n=3) and C57BL/6J (n=3) mice were euthanized
with CO2 gas, and fixed overnight in 70% ethanol. Mandibular incisors were
extracted and fractured into three pieces perpendicular to the labial surface
enamel at the same region for all samples and mounted for viewing under
scanning electron microscopy. Teeth were examined at 600, 1000 and 5000
magnification.

Micro-computerized tomography:
Three AP3B1 null mice (average age 386 days, average weight 29.5g),
and three C57BL/6J mice (average age 377 days, average weight 29.0g) were

44
euthanized with CO2 gas. Hemi-mandibles were dissected and fixed in 10%
neutral buffered formalin for two days before immobilization in 2% agarose for
Micro-Computerized Tomography analysis.
Specimens were scanned at the University of Southern California
Molecular Imaging Center using the Inveon µCT scanner and the Inveon
Acquisition Workplace software (Siemens Medical Solutions USA Molecular
Imaging, Knoxville, TN). Scan settings were: 80 kilovolts, 250 microamps, 750
projections with 7000 milliseconds/projection, and 360 degrees covered using
1280x1280x1536 pixels for a field of view of 11.95x11.95x14.34mm^3 in high-
magnification mode. Images were reconstructed from raw data to produce 9.34
micron voxel size images using the Cobra EXXIM software (EXXIM Computing,
Livermore, CA). Images were loaded into the Amira 5.2 software (Visage
Imaging, La Jolla, CA) for viewing and analysis. Datasets were re-sliced to view
90-degree cross sections on the outer edge of the incisor at the extrusion site
from the mandible. Thickness and volume measurements were calculated from
the single 90-degree cross section of each tooth using a segmentation editor for
the incisor enamel, dentin, and pulp layers.

Statistical analysis of tissue area using micro-computerized tomography:
To calculate the area of pulp, dentin and enamel, the volumes from µCT
images were divided by the thickness of each imaged section (9.33µm). The

45
resultant calculated areas were subjected to F-testing which showed equal
variance when comparing same tissues (pulp, dentin, and enamel) from the two
strains of mice. Student T-test showed a significant difference for enamel areas
between the two groups (p=.03). Dentin and pulp differences were not
statistically significant.

Linear enamel thickness:
Linear enamel thickness was calculated by measuring the distance from
the enamel surface to the dentin-enamel junction in both AP3B1 null (n=3) and
C57BL/6J (n=3) mice. These measurements were taken from µCT images at the
point in which the labial surface of the mandibular incisor exited the alveolar crest
of the mandible. Each image examined was taken at 90° perpendicular to the
labial surface of enamel. Student T test showed high significance for the
difference in linear enamel thickness for the two groups of mice (p=.001).

Light microscopy imaging of AP3B1 null and C57BL/6J mice:
One year old mice were euthanized using CO2 gas and fixed in 10%
neutral buffered formalin for two days. Incisors and molars were imaged in situ
and after extraction at 10x magnification using light microscopy.

46
Characterization of the enamel surface of mouse teeth by FTIR
spectroscopy:

Teeth from one year old mice, C57BL/6J (n=3; 1 incisor and 2 molars) and
AP3B1 null (n=3; 1 incisor and 2 molars) were extracted and fixed in 10% neutral
buffered formalin. Teeth were pestled in a ceramic mortar. The resulting powder
was carefully fixed on double sided tape (Scotch, 3M) and placed on a DRIFT
sample holder. A scan of powder-free tape was used as C57BL/6J. Pure
hydroxyapatite (HA) powder (SigmaAldrich, St. Louis, USA) was scanned to
measure the absorbance of the pure mineral phase, since enamel mainly
consists of HA crystals. FTIR spectroscopy (Spectrum 400, PerkinElmer, Inc.
Waltham, MA, USA) was used in DRIFT mode with the following parameters:
Wave number range: 4000 cm
-1
- 450 cm
-1

Number of scans: 8
Resolution: 4 cm
-1
The spectra measured were analyzed for typical absorbances characteristic
for protein groups and mineral groups using Spectrum 6.3.2.0151 (PerkinElmer,
Inc., Waltham, MA, USA).
The Ca content (cCa) was measured for each sample by Atomic absorbance
spectroscopy (AAS; AAnalyst 400, PerkinElmer, Inc. Waltham, MA, USA), and
peak areas were normalized to cCa [ppm].

47
Study of the periodontal ligament (PL) in AP3B1 nulls and C57BL/6J
animals:
One year old AP3B1 null and C57BL/6J mice were euthanized with CO2
gas. Hemi-mandibles were dissected, fixed for 3 days in 10% neutral buffered
formalin and than demineralized in 4.5% EDTA for about three weeks. The hemi-
mandibles were embedded in paraffin and sectioned into 6μm thick slices,
mounted on slides and stained with hemotoxylin and eosin. Sections were
visualized and photographed using light microscopy at 4x magnification.
Also, the periodontal ligament was visualized using µCT. Hemi-mandibles were
sagitally sectioned through the central groove of the first and second molar, and
periodontal ligaments were examined in the resultant images.

RESULTS

RT-PCR of β1-4 in ameloblast-like cells:
To verify that all four adaptor proteins (AP1-AP4) are present in
ameloblasts, a RT-PCR reaction was run using primers for the large β subunit
associated with each specific adaptor protein complex (Fig. 10). All four adaptor
proteins are present in ameloblasts.

48

Figure 10. ALL
ADAPTOR PROTEIN
SUBTYPES ARE
PRESENT IN
AMELOBLAST-LIKE
LS8 CELLS: RT-PCR
shows that all adaptor proteins (based on amplification of specific β subunits) are
present in ameloblast-like LS8 cells. Each amplicon is the correct size based on
the specifically designed primer pairs.

Genotyping of animals:
To verify AP-3 is present in the control animals (C57BL/6J) and absent in
the AP3B1 null animals, tail clippings were digested, DNA isolated, and amplified
using specific primers (Fig. 11). For control animals, primers were designed to
amplify a portion of exon 5 (324bp) of AP-3 and for AP3B1 null animals, a portion
of the LacZ gene in the inserted sequence (290bp). Results gave size bands as
described in the Jackson Laboratory protocol for animal genotyping.
Figure 11.
CONFIRMATION OF
GENOTYPES:
Digestion and PCR of
tail clippings from
C57BL/6J and AP3B1 null animals shows amplification of the AP3B1 gene
(324bp) in C57BL/6J and LacZ (290bp) in the AP3B1 mice. Both sets of primers
pairs (AP3B1 and LacZ) were used in each PCR reaction. This data confirms the
genetic strain of the animals used.

49
Age and weight comparison:
To ensure that differences in age or weight of the animals were not
statistically significant, each animal was weighed and age calculated (Fig. 12).
Using age and weight matched animals ensured that any change found in the
comparison of dentitions was due to the mutation in AP-3.

Figure 12. WEIGHT AND AGE COMPARISON OF ANIMALS USED IN THE
GENOTYPIC AND PHENOTYPIC ANALYSIS: Three animals of AP3B1 null (3
females) and three C57BL/6J (2 female, 1 male; controls) were included in the
analysis. No statistical differences were noted in the age (panel A) or weight
measurement (panel B) for experimental vs. controls.

Examination of the dentitions in AP3B1 null and C57BL/6J animals:
Both AP3B1 null and C57BL/6J mice appeared healthy when sacrificed
(mutant animals are easily identified based on their “pearl” coat color; Fig. 13A),
however the mandibular incisors of the AP3B1 null mice appeared discolored
when compared to non-mutant C57BL/6J control animals (Fig. 13B). The most
obvious explanation of the discoloration of the incisor teeth would be surface

50
enamel irregularities or porous enamel (hypomineralized), but in both cases
indicating a higher susceptibility of staining. In addition, the incisor and first molar
teeth of AP3B1 null mice appeared larger than their non-mutant C57BL/6J
counterparts (pearl teeth exhibit megadonture; Fig. 13C and D), however, a
larger sample size is need to confirm these findings.

FIGURE 13. DENTITION OF 12-MONTH OLD AP3B1 NULL ANIMALS
COMPARED TO AGE-MATCHED C57BL/6J CONTROL ANIMALS. Panel A
shows the AP3B1 null coat color vs. the C57BL/6J animal. Panel B; mandibular
incisor teeth of the AP3B1 null mice is a yellow color as compared to C57BL/6J.
Both incisor teeth (panel C) and first molar teeth (panel D) are noticeably larger
in the AP3B1 null mice when compared to C57BL/6J, however, a larger sample
size is needed to confirm these observations.

Enamel thickness comparisons between AP3B1 null and C57BL/6J mice:
MicroCT was used to measure enamel, dentin and pulp surface area from
a section taken perpendicular to the enamel surface in a region that was fully
mature, just after eruption(Fig. 14A-D). Data showed that pearl mice enamel was

51
~170µm thick at its widest position, and this compared to ~140µm in wild type
enamel; this data was highly (p value <0.002) significant. Statistically significant
differences were noted in the cross-sectional area of enamel (p value <.05), but
not dentin or pulpal tissues in this same region (Fig. 14F). The prismatic
arrangement of the enamel in pearl mice appeared relatively normal when
compared to wild-type control animals, although the difference in enamel
thickness was also clearly apparent using SEM (Fig. 15).

FIGURE 14. µCT MEASUREMENTS FOR MATURE ENAMEL THICKNESS
AND ENAMEL, DENTIN AND PULPAL AREA IN A CROSS SECTION
PERPENDICULAR TO THE ENAMEL SURFACE JUST AFTER ERUPTION,
BUT FULLY MATURE REGION: Mandibular incisor teeth were examined by
µCT through a section as indicated in panels A and B. Images typical of those
shown in panel C (AP3B1 null) and panel D (C57BL/6J) were generated for each
animal and analyzed. Measurements for enamel thickness (panel E), and
enamel, dentin and pulp surface area (panel F) were calculated. An n of three
AP3B1 null and n of three C56BL/6J animals were included in the analysis. Using
the student T-test, significance is noted for enamel thickness with a p value of
<0.002 (**), and for enamel area with a p value of <0.05 (*).

52

Figure 15. ENAMEL INCISOR MORPHOLOGY OF AP3B1 NULL AND
C57BL/6J MICE IN THE MATURE REGION, FROM 6-WEEK ANIMALS
FRACTURED PERPENDICULAR TO THE LABIAL SURFACE: Panels A-C,
and D-F show increasing magnification of the enamel in AP3B1 null (A-C) and
C57BL/6J animals (D-F). Dentin-enamel junction (DEJ) is identified with a broken
line (panels A and D). While the enamel morphology appears similar in both
AP3B1 null and C57BL/6J mice, there is a clear difference in the enamel
thickness, illustrated by double-headed arrows and measured perpendicular to
the tooth surface, with AP3B1 null mice showing ~ 30% increase in enamel
thickness confirming the µCT data in figure 4. Scale bars (top right corner) are
50µm (panels A and D), 20µm (panels B and E) and 5µm (panels C and F).

Periodontal ligament analysis:

An H & E staining and µCT analysis was also carried out on 12-13
month-old animals to assess the health of the periodontal ligament (PL) (Fig. 16).
A widened PL is noted in 12-month old mutant animals when compared to wild-
type controls. The widened PL is noted around the entire root surface of the
tooth. Periodontal disease is also clearly apparent with a loss of alveolar bone
height (Fig. 16, panels A vs. E; noted by double-headed bell-bars), and also the
large numbers of inflammatory cells evident (as determined by the basophilic

53
staining) in the periodontal ligament and gingival of AP3B1 mutant animals, and
the absence of inflammatory cells in the gingival of wild-type animals (Fig. 16,
panel B compared to F respectively).

Figure 16: STUDY OF THE PERIODONTAL LIGAMENT (PL) IN AP3B1 NULL
(PANELS A-D) AND C57BL/6J (PANELS F-H) ANIMALS:
Two AP3B1 null and two C57BL/6J animals were studied using H&E staining,
and presented data is typical of that seen in animal groupings. A widened
periodontal ligament is apparent in the AP3B1 null animals compared to
C57BL/6J controls. (Panels B and F; identified by black line tracings outlining the
cemental surface and the alveolar bone, and *, indicating PD width in 1
st
molar
furcation regions). The points of gingival attachment to the tooth surfaces in the
interdental region are indicated by arrows (panels B and F). Interdental alveolar
coronal bone height also identified (panels B and F; identified by double-headed
bell-bars). Increased periodontal ligament vasculature also noted in AP3B1 null
mice (panel B, **). µCT sections (panels C and G) presented from AP3B1 null
and C57BL/6J animals (panels D and H respectively) also indicates a widened
periodontal ligament in AP3B1 null animals (indicated by *). Scale bar panels A
and E 500µm; panels B and F 200µm.

54
Examination of the molars in AP3B1 nulls and C57BL/6J mice:

Crown dimensions for the maxillary first and second molars were
calculated for the ~ 13 month old animals (n=3 for AP3B1 null; and n=3 for
C57BL/6J) after their careful dissections (Fig. 17). While no differences could be
determined for the crown mesiodistal dimensions of the second molars, a
statistically significant difference in size was noted for the mesiodistal diameters
of the first molar teeth of AP3B1 null and C57BL/6J animals (1.414mm and
1.333mm respectively), suggesting pearl mice have an ~ 6.1% larger teeth.

55

Figure 17. ANALYSIS OF THE MESIODISTAL DIMENSIONS OF THE
CROWNS OF FIRST AND SECOND MAXILLARY MOLARS OF 12-MONTH
OLD AP3B1 NULL AND C57BL/6J ANIMALS: Panels A-D illustrate how
measurements were made from either the buccal view (panels A and B) or the
occlusal view (panels C and D). The maxillary right first and second molar teeth
of AP3B1 null mice (n=3) and C57BL/6J (n=3) were dissected and photographed
from the buccal and occlusal surfaces as shown. Crown dimensions were
assessed at the maximal width as shown by the double-headed arrows from the
distal (D) surface to the mesial (M) surface. The measurements taken from the
occlusal images were consistently larger. From the buccal view the width of the
crown was calculated at 1.381 (+/- 0.027) mm for the AP3B1 null mice, and 1.310
(+/- 0.017) mm for the C57BL/6J mice; the difference being statistically significant
with a p value of <0.05 (*). From the occlusal view the width of the crown was
calculated at 1.425 (+/- 0.043) mm for the AP3B1 null mice, and 1.333 (+/- 0.016)
mm for the C57BL/6J mice; the difference being statistically significant with a p
value of <0.05 (*). Standard deviations are also noted in the calculations (+/-). No
significant differences were noted in the crown dimensions of the second molar
teeth of AP3B1 null and C57BL/6J.

56
Quantification of mineral and peptide content in teeth from AP3B1 null and
C57BL/6J mice:
Teeth from both AP3B1 null and C57BL/6J mice were examined using
Fourier transform infrared spectroscopy (FTIR) to determine if there were any
changes in the amounts of mineral or protein content between the two groups of
mice (Fig. 18). The results show that there is a significant decrease (p=0.03) in
the tooth protein content of AP3B1 null mice (35.4%). This data is measured from
whole teeth so it is difficult to determine the relative contributions from the
enamel, dentin or pulp.

Figure 18. COMPARISON OF THE PERCENTAGE OF PEAK AREAS FOR
THE ABSORBANCES OF FUNCTIONAL GROUPS TYPICAL FOR HYDROXYL
APATITE AND PROTEINS: This data shows absorbances of mineral functional
groups and absorbances of protein functional groups. No significant differences
in the absorbances of each functional group were found. However, overall
absorbance for proteins was about 35.4% lower for AP3B1 null teeth than for
teeth of from C57BL/6J mice (p=0.003).

57
Discussion

As previously discussed, enamel matrix protein degredative products are
effectively endocytosed into LAMP positive vesicles, but because there are
variations between in-vitro vs. in-vivo experiments, we chose to examine teeth
from an animal model in which there is a disruption in the normal AP-3
endocytotic pathway to validate our hypothesis that the endocytosis of enamel
matrix proteins plays a role in amelogenesis,.
All LAMPs contain at the C-terminus, a tyrosine sorting sequence, which is
specifically recognized by AP-3. AP-3 binds to this sequence and directs LAMP
coated vesicles to the late endosome/lysosome. In humans, HPS is a disease in
which mutations in AP-3 causes a miss-sorting of LAMPs as well as lysosomal
related organelle complexes. Patients with HPS have occulocutaneous albinism,
prolonged bleeding times, and immunodefiency due to defects in forming
melanosomes, platelet dense granules, and functioning neutrophils(Dell'Angelica
et al., 1999; Di Pietro and Dell'Angelica, 2005; Hermansky and Pudlak, 1959). A
thorough search of the literature on HPS describes only one case of
cosanguineous offspring in which the “most prominent finding was severe dental
decay and aggressive periodontitis”(Jung et al., 2006a), but no other studies
have examined the teeth from HPS patients.
The animal model we used in our study was created as a model for HPS
type II (specifically involving AP3B1 gene mutations), and has an inserted

58
sequence in exon 5 of the AP3B1 gene resulting in an AP3B1 null
phenotype(WM Canfield, ; Yang et al., 2000). This insertion into the AP3B1 gene
destabilizes the whole AP3 complex (Chen et al., 2006).
Normally, (Fig. 19A) the bulk of LAMPs after glycosylation in the Golgi
apparatus, traffic to sorting endosomes where they are bound by the µ subunit of
the AP-3 complexes and sorted to the late endosome/lysosome (Dell'Angelica et
al., 1999). A secondary trafficking pathway results in small quantities of LAMPs
transiting to the plasma membrane via recycling endosomes.
In HPS type II, mutations in the AP3B1 subunit results in the majority of
LAMPs being shunted through recycling endosomes to the plasma membrane,
bypassing normal AP-3 directed trafficking to the lysosome. In addition, LAMPs
normally found on early endosomal membranes (because of plasma membrane
contributions to the early endosomal membrane), are re-recycled back to the
plasma membrane instead trafficking to the lysosome (Fig. 19B). The miss-
sorting of LAMPs by AP-3 mutations may result in tooth changes because of
delayed timing and/or removal of EMP’s from the matrix.

59

Figure 19. INTRACELLULAR TRAFFICKING OF LAMPS: Following LAMPs
synthesis, they are sorted and packaged in Golgi bodies, and sorted to
endosomes. From the endosome LAMPs are either directly transported to
lysosomes (major pathway), or transported to the plasma membrane (minor
pathway)(Panel A). Mutations to any of the AP-3 subunits, impacting on AP-3
activity (red X) results in the accumulation of LAMPs in the plasma membrane
through recycling pathways (Panel B). The schematic shows how processed
enamel matrix proteins (EMPs), through a direct protein-protein interaction with
the LAMPs, could result in their movement into lysosomes for a more complete
proteolysis and a block in AP-3 activity would result in the accumulation of EMPs
in the extracellular matrix.

In this study we wanted to test the hypothesis that disruptions in the
trafficking of LAMPs results in enamel changes. Initial photographic images of
the teeth in AP3B1 null mice revealed a yellowish hue to the incisors when
compared to the C57BL/6J mice (Fig.13B). This staining suggested either
hypomineralized enamel which is allowing for stain to penetrate or a roughened
enamel surface which is capturing stain. AP3B1 null teeth were also noticeably
larger than C57BL/6J animals when compared side by side (Fig. 13C-D),
suggesting that altered trafficking has an impact on tooth size.

60
Crown dimensions for the maxillary first and second molars were
calculated for the ~ 13 month old animals (n=3 for AP3B1 null; and n=3 for
C57BL/6J) after careful dissections. While no differences could be determined for
the crown mesiodistal dimensions of the second molars, a statistically significant
difference in size was noted for the mesiodistal diameters of the first molar teeth
of AP3B1 null and C57BL/6J animals (1.414mm and 1.333mm respectively),
suggesting AP3B1 null mice have an ~ 6.1% larger teeth (Fig. 17). SEM of
AP3B1 null mice incisors was used to examine the overall structure of the
enamel which revealed that the thickness of the enamel layer from the dentin-
enamel junction to the labial surface of 6 week mouse incisors is thicker in
AP3B1 nulls than C57BL/6J animals. SEM showed that although the enamel
layer is bigger, prismatic architecture appears to have a normal morphology. This
finding suggests that the phenotype seen in teeth of AP3B1 null mice is not
caused by disruptions of normal enamel architecture.
Micro-CT offered a more precise way to measure enamel thickness using
incisors from AP3B1 null (n=3) and C57BL/6J (n=3) adult mice. Slices of the
incisor were taken at 90° to the labial surface of the tooth just past the point of
eruption from the mandibular alveolar crest in a fully mature region. Comparing
like tissues (enamel, dentin, and pulp) of AP3B1 null to C57BL/6J mice, there
were no significant differences between the groups for pulp and dentin, but
similar to findings from SEM, there is a significant increase in enamel area of the
AP3B1 nulls when compared to C57BL/6J animals (p<.05). Linear enamel
thickness measurements examined in the same images, showed a highly

61
significant increase (p<.002) for enamel thickness in AP3B1 null mice (avg. 170
µm) compared to C57BL/6J mice (avg. 145 µm), or ~21% thicker.
To rule out any doubt that the changes in enamel could be due to
variations in animal weight or age of the animals used in the study, a comparison
showed that there is no statistical significance between the two groups of mice
for weight or age. This supports the finding that the changes in enamel thickness
are because of altered trafficking.
Collaboration was undertaken with Dr. Petter Lyngstadaas’s lab to
quantify relative protein and mineral contents of the two strains of mice using
FTIR and the conclusion drawn from their lab is that:

A relative quantitative comparison showed that M(utants) had in average
35.4% lower peak areas of absorbance at wavenumbers typical for
proteins. This difference was significant (p=0.003). The results indicated
that the WT samples contained a higher amount of peptides in the teeth.
The values for the single protein absorbances at the respective
wavelengths are not significantly different due to low number of specimen,
but a tendency for lower protein absorbance was seen” (see appendix 1
for the full report).

It is difficult to interpret this data, as whole teeth were ground into powder
and used for the study, making it impossible to determine the contributions of
each tooth specific tissue (i.e. dentin, pulp and enamel) to the result. Needless
to say, the data supports the idea that aberrations in normal trafficking result in
tooth changes.
Examining the histology of the periodontal region in the molars of AP3B1
and C57BL/6J mice, revealed that AP3B1 null mice have advanced periodontal

62
disease with loss of intra-crestal bone and infiltration of inflammatory cells,
evidenced by the large amount of basophilic staining. Noted was
hypervascularization in the gingival tissues of AP3B1 nulls. µCT analysis of the
periodontium revealed an increased periodontal ligament space, a sign of
periodontal disease. All of these findings are indicative of periodontal disease,
which correlates with the high incidence of periodontal disease seen in HPS
patients because of impaired neutrophil function.(Feliciano et al., 2006; Jung et
al., 2006a)
These significant data from our experiments strongly suggest that miss-
routing of LAMPs due to AP3B1 mutations results in enamel changes. A potential
weakness in this study is the small number of animals used, as well as the fact
that there may have been a genetic drift in the AP3B1 nulls from a pure
C57BL/6J strain (over 20+ years of inbreeding), which further complicates
phenotypic analysis.
Initially, our aim was to set up a breeding program to compare AP3B1 null
mice, C57BL/6J, and heterozygotic litter-mates. Unfortunately, Jackson
Laboratories was in the process of cryopreserving the AP3B1 null mouse line at
the same time we started our studies. We obtained all animals available at the
time, but the AP3B1 null mice have a significantly reduced breeding time and
produce very small litters, and all attempts at cross-breeding failed, most
probably because of the advanced age of the animals received.
Our findings nonetheless support our hypothesis that endocytosis is an
important part of amelogenesis because when normal trafficking of key

63
components of the endocytotic machinery are absent, changes in a normal tooth
phenotype is seen. The mechanism is unknown how AP3B1 alterations result in
changes in teeth, but it is clear that alteration of normal trafficking of LAMP
proteins by mutations in AP3B1 result in tooth changes.

64
Chapter 4.

An Amelogenin Minigene to Study Alternative Splicing

PREAMBLE

Although not directly related to endocytosis of enamel matrix protein
derivatives, initially when starting my PhD studies, I was interested in examining
alternatively spliced products of amelogenin and the following is the text from a
paper on using an amelogenin minigene to study alternative splicing (Shapiro et
al., 2006)

INTRODUCTION

For the many Genome Projects, not only is raw chromosomal DNA
sequence being collected, but also individual gene products are being
characterized. Characterizing gene products is partly done by identifying features
such as known and potential RNA exon-intron splice junctions and by
determining each potential protein’s open reading frame. The study of expressed
sequence tags (ESTs; the sequencing of complementary DNA, or cDNA, to

65
individual messenger RNAs, or mRNA) can be used to confirm known or
predicted exon-intron splice junctions, to identify cryptic splice junctions, and to
tabulate the many alternative spliced isoforms produced from each individual
gene. By extensive in silico analysis of available ESTs, it is estimated that about
60-70% of human and murine genes have multiple alternative spliced isoforms
(Brinkman, 2004; Kalnina et al., 2005). This mechanism of gene expression
ultimately contributes to proteome complexity.
The enamel matrix protein amelogenin is one clear example of a gene
subjected to alternate splicing that results in multiple isoforms (Simmer et al.,
1994; Veis, 2003). In the mouse model, two of the better-studied isoforms are
referred to as M180 (including exons 1, 2, 3, 5, 6 and 7) and the leucine-rich
amelogenin peptide or LRAP (including exons 1, 2, 3, 5, part of exon 6, and 7).
Amelogenin M180 codes for a secreted protein of molecular weight 20.3kDa,
while LRAP codes for a secreted protein of molecular weight 6.7kDa. All
available data suggests that M180 is an extracellular structural protein of enamel
(Paine et al., 2000b; Smith, 1998). M180, and related enamel matrix proteins,
may also act as osteopromoters (Boyan et al., 2000; Kawana et al., 2001). LRAP
has recently been studied as a growth factor, or signaling molecule, with
osteopromotive capabilities (Boyan et al., 2000; Kawana et al., 2001). It is also
believed that LRAP is involved with the process of cementogenesis (Boabaid et
al., 2004; Viswanathan et al., 2003).
While normal splicing events give rise to a number of mRNA isoforms,
each coding for a unique protein with perhaps a particular function, aberrant RNA

66
splicing appears to be a common feature of many cancers (Brinkman, 2004;
Kalnina et al., 2005; Venables, 2004). One example of this is the integral
membrane glycoprotein CD44, and that the expression of certain alternate
spliced variants correlates to an invasive and metastatic phenotype in many
cancers (Herrlich et al., 1998; Ponta et al., 2003).
The generation of a “minigene” allows for a cell culture model system to
study in vivo, tissue-specific splicing events (Stoss et al., 1999). Minigenes are
generated by the sub-cloning of a gene of interest into a plasmid that then allows
for the promiscuous expression of that gene’s mRNA in eukaryotic cells. For
genes with multiple exons, this then allows for the identification of alternative
spliced RNA transcripts, and thus minigenes can be used as a model system to
study RNA splicing events. For example, individual splicing factors can be up-
regulated or down-regulated in an attempt to decipher which splicing factors
dictate splice site selection. Our primary goal is to study amelogenin gene
expression and protein function, and this encompasses ultimately determining
why so many amelogenin splice variants exist. This paper describes the
generation of a functional mouse amelogenin minigene, and reports for the first
time two novel amelogenin mRNA transcripts generated using this in vivo
system.

67
MATERIALS AND METHODS

Mouse amelogenin (mAmel) minigene construction:
Oligonucleotide PCR primers PA25203 (forward; 5’-
GCCGGC AACCATCAAGAA ATG GGG ACC TGG ATT TT) and PA25204
(reverse; 5’- CTCGAGATC
CTGAAAAAATATAGCAAAAGGAACAATTAACATAATTTACAAACATTCC) were
used to amplify 5.73kb of mouse genomic DNA (male BALB/c mice; purchased
from BD Biosciences Clontech, Palo Alto, CA; catalogue # 636402). Included in
the primers were Nae I and Xho I restriction enzyme sites (underlined) to allow
for the intended sub-cloning step into the expression plasmid pcDNA4/TO/myc-
His
TM
A (Invitrogen; Carlsbad, CA) (Figure 1). Arrow indicates exon 2 start site,
while bold in the forward primer indicated the amelogenin translation start site
(ATG), and bold in the reverse primer codes for the terminal amelogenin
aspartate codon. This PCR product was subsequently sub-cloned into a TA
cloning vector (pCR

2.1; Invitrogen), sequenced partially from both directions
(Applied Biosystems ABI 377 XL DNA sequencing instrument; USC/Norris
Comprehensive Cancer Center Microchemical Core Facility) to ensure the
correct genomic DNA gene was amplified, released from pCR

2.1 using Nae I
and Xho I restriction enzymes, and directionally sub-cloned into the vector
pcDNA4/TO/myc-His
TM
A at the Eco RV / Xho I multicloning site. The resulting
plasmid is referred to as the mouse amelogenin (mAmel) minigene (Fig. 20). Of

68
note in the mAmel minigene is the CMV promoter driving expression of the
minigene, and the inclusion of c-myc and polyhistidine tags following the terminal
(amelogenin-related) aspartate immediately prior to the stop codon.

Figure 20. FEATURES OF THE mAMEL MINIGENE: Arrow-head indicates the
relative position of the exon 2 / intron 2 boundary. Relative positions of the PCR
primers used to amplify and sub-clone mouse genomic amelogenin DNA, and
amplify its related mRNA transcripts, are indicated. Figure is not to scale.

Minigene transfection, RT-PCR and DNA sequencing:
Mouse ameloblast-like (LS8) cells (Dhamija and Krebsbach, 2001; Zhou
and Snead, 2000), mouse embryo NIH/3T3 cells (American Type Culture
Collection or ATCC catalogue # CRL-1658) or human osteoblast hFOB 1.19 cells
(ATCC catalogue # CRL-11372) were transfected with the mAmel minigene
using Lipofectamine2000 (Invitrogen). After 24 hours, total RNA was isolated
from each cell line, and cDNA to the mRNA was generated using standard
methodologies (Sambrook and Russell, 2001). The resulting cDNAs generated
from the mAmel minigene were subjected to PCR amplification with primer sets
specific to the minigene. These two primers were PA25203, and PA28956

69
(reverse; 5’- CAACTAGAAGGCACAGTCGAGGCTGATCA) (Fig.17). Primer
PA28956 complements a transcribed region within the mAmel minigene that lies
3’ to the mouse genomic sequence. That is, reverse primer PA28956 is
contained fully within plasmid pcDNA4/TO/myc-His
TM
A (Figure 1). Standard
methodologies were used for PCR (Sambrook and Russell, 2001). PCR products
were then separated by size electrophoretically and individual DNA products
were isolated and purified (Sambrook and Russell, 2001). After DNA gel
purification, each product was sub-cloned into pCR

2.1 and transformed into
competent bacteria (Escherichia coli strain DH5- ). Bacteria were then plated on
ampicillin, 5-bromo-4-chloro-3-indolyl- -D-galactosidase (X-Gal) and
isopropylthio- -D-galactoside (IPTG) selective plates (Sambrook and Russell,
2001), and grown at 37˚C overnight. Insert-containing plasmid DNA was
subjected to restriction mapping using the restriction enzyme Eco RI. Insert DNA
of interest was then sequenced to identify and confirm the mAmel minigene
splice products.

Nested PCR:
Nested PCR was used to demonstrate cell-specific mAmel spliced
variants. The external primers used were PA25203 and PA28956, and the
internal primers used were PA32983A (forward; 5’-
GTTTGCCTGCCTCCTGGGAGCAGC) and PA32983B (reverse; 5’-
CCTCTTCTGAGATGAGTTTTTGTTCGAAGGG) (Fig. 17). This inner primer set

70
is specific to the minigene with PA32983A being fully contained within exon 2
and PA32983B being fully contained within plasmid pcDNA4/TO/myc-His
TM
A
(Figure 1). After 10 cycles with the external primer set, an aliquot of product was
then subject to PCR (30 cycles) with the internal primer set. Again, standard
methodologies were used for PCR (Sambrook and Russell, 2001). In addition to
LS8, NIH/3T3 and hFOB 1.19 cells, the human malignant melanoma cell line A-
375 (ATCC catalogue # CRL-1619) was also included. Each cell line was
subjected to exactly the same RT and PCR reaction conditions as all
experimental procedures were done concurrently.

RESULTS

Two novel mouse amelogenin isoforms identified using the mouse
amelogenin minigene:
We have sub-cloned and sequenced a small number of PCR-generated
alternate spliced cDNAs to the mouse amelogenin minigene. Each of these
products was produced in either mouse ameloblast-like (LS8) cells, mouse
embryo NIH/3T3 cells or human osteoblast hFOB 1.19 cells (Fig. 21). To date we
have identified five different amelogenin transcripts, and these products are listed
(Table 1). Three of these transcripts have been described previously: these being
M180, M156 and M59. The two remaining and “novel” transcripts are identified
as recombinant M18 (rM18) which excludes exons 3, 4 and 6 (mAmel[-3,-4,-6]),

71
and recombinant M34 (rM34) which excludes exons 4 and 6 (mAmel[-4,-6]) (data
not shown). ClustalW formatted alignment (Karlin and Altschul, 1990) of these 5
isoforms is presented (Fig. 22).

Size in
bp
LS8 hFOB
1.19
NIH/3T
3
A-375
M180 737
(617)
√,s √ √,s
M156 665
(545)
√,s
M59
(LRAP)
374
(254)
√,s √,s √,s √
rM34 299
(179)
√,s √,s √,s √
rM18 251
(131)
√ √,s √

Table 2. TWO NOVEL MOUSE AMELOGENIN ISOFORMS IDENTIFIED USING
THE MOUSE AMELOGENIN MINIGENE: The various mouse amelogenin
isoforms produced from either LS8, hFOB 1.19, NIH/3T3 or A-375 cells. The
primers PA25203 and PA28956 were used to give the product sizes as listed.
For selected samples produced from the PCR using primers PA25203 and
PA28956, DNA sequencing (indicated by an “s”) confirmed the identities of these
products. Nested PCR using PA25203 and PA28956 as external primers, and
PA32983A and PA32983B as internal primers, gave the product sizes shown in
brackets. A tick indicates that the product was visually evident by reverse
transcription (RT) and nested PCR followed by gel electrophoresis. To date we
have identified five different amelogenin transcripts, three of these transcripts
have been described previously: these being M180, M156 and M59. The two
remaining and “novel” transcripts are identified as recombinant M18 (rM18) which
excludes exons 3, 4 and 6 (mAmel[-3,-4,-6]), and recombinant M34 (rM34) which
excludes exons 4 and 6 (mAmel[-4,-6]) (data not shown).

72

Figure 21. AMINO-ACID SEQUENCE ALIGNMENT FOR M180, M156, M59
(LRAP), rM34 AND rM18:ClustalW formatted alignment (Karlin and Altschul,
1990) of these 5 isoforms is presented

Differential splicing patterns are observed for the amelogenin minigene in
different cell lines
Using the RT and nested PCR approach, we show a cell-specific
amelogenin isoform profile with the cell lines investigated (Figure 3). A summary
of this data are included (Table 1).

73

Figure 22. CELL-SPECIFIC
ALTERNATIVE SPLICED
AMELOGENIN GENE PRODUCT
PROFILES: Following
transfection, reverse transcription
(RT) and nested PCR was used
to demonstrate alternative spliced
amelogenin minigene products in
LS8 (lane 7), hFOB 1.19 (lane 8),
NIH/3T3 (lane 9) and A-375 cells
(lane 10). Lanes 1 and 11 are 100
bp DNA ladder (New England
Biolabs, Inc., MA). Lanes 2-6
each are positive controls
generated from identical PCR
conditions using the originally
sub-cloned, mAmel minigene
spliced products as template
DNA. These are: M180 (lane 2;
617 bp), M156 (lane 3; 545 bp),
M59 (lane 4; 254 bp), rM34 (lane
5; 179 bp) and rM18 (lane 6; 131
bp). Panel B is an enlarged image
of panel A.

Discussion

A widely accepted illustration and nomenclature describing the various
amelogenin alternate spliced transcripts has been presented previously (Simmer
and Snead, 1995; Veis, 2003), and will be used here. Of note is that amelogenin
splice junctions for exons 1, 2, 3, 4, 5 and 7 are predictable based on DNA
sequence, the so-called intronic dinucleotides GT and AG at the 5’ and 3’ splice

74
sites respectively (Veis, 2003), however exon 6 contains three cryptic splice
junctions, and these sites divide exon 6 into four regions that have been defined
as exons 6A, 6B, 6C and 6D (Simmer and Snead, 1995; Veis, 2003). In addition,
each spliced RNA isoform encode a unique protein of an easily definable amino-
acid number and composition. Thus amelogenin isoforms are frequently referred
to by the predicted number of amino-acids contained in the secreted protein; for
example M180 contains 180 amino-acids, M194 includes exon 4 and contains
194 amino-acids, and LRAP contains 59 amino-acids and is often referred to as
M59 (Veis, 2003), and so on.
The use of a mouse amelogenin minigene in this study has resulted in the
discovery of two novel mRNAs for mouse amelogenin, and we refer to these as
rM18 and rM34. Transcription of this minigene is under the control of the CMV
promoter that allows for robust, non-discriminate gene expression in a large
number of non-related cell lines. We have demonstrated that different cell types
produce unique amelogenin mRNA profiles using this same amelogenin
minigene. While seemingly obvious, this data are suggestive that spliceosome
composition may, in part, account for phenotypic differences between cell types.
At the level of detection used in this study, we were not able to
demonstrate that the mouse amelogenin minigene generated the full range of
amelogenin isoforms that have been described previously (Simmer et al., 1994;
Veis, 2003). While it is possible that all of the possible alternatively-spliced
amelogenin isoforms are produced under the conditions used, it may be
necessary to extend the number of PCR cycles before these “relatively minor”

75
products are evident by gel-electrophoresis. There is also no data describing the
relative prevalence of the majority of these amelogenin alternatively-spliced
isoforms, as most have been identified by sequencing RT-PCR DNA products.
The consequence of identifying novel isoforms of amelogenin suggests
that an even more complicated understanding of amelogenin proteomics, as it
relates to whole organism development, will evolve in time. However, with the
exception of M180 and LRAP (M59), no data has been offered to describe
physiological functions to any of the other amelogenin isoforms. In addition, the
generation of rM18 and rM34 isoforms may be a characteristic of the mouse
amelogenin minigene only, and may not be produced under normal physiologic
circumstances in the developing animals in vivo. We believe that this is perhaps
unlikely, as the molecular biology methodologies used to amplify and isolate
cDNAs from mRNAs generally has resulted in the exclusion of smaller PCR-
generated products. This discussion, however, remains an academic one until
more definitive data are forthcoming.
With the generation of a functional mouse amelogenin minigene we now
have an in vivo model system to look at cell-specific RNA splicing events. With
this model system we hope to better understand the molecular mechanism that
dictates exon-intron splice site selection.

76
Conclusions

Amelogenesis is a very complex and regulated process in which the end
result is the formation of a unique mineralized tissue that can withstand the
forces of a lifetime of mastication. Mature dental enamel is the hardest tissue in
the body, comprised of greater than 95% hydroxyapatite, yet immature enamel is
mostly made up of proteins, mainly amelogenins, enamelins and ameloblastin
(Bartlett et al., 2006; Fukumoto et al., 2004; Hu et al., 2000; Hu et al., 2001; Lee
et al., 2003; Margolis et al., 2006; Paine et al., 2001). As amelogenesis
progresses through a maturation stage, specialized enzymes are secreted from
ameloblasts to digest enamel matrix proteins, MMP20 at early maturation and
KLK4 at late maturation (Bartlett et al., 1996; Hu et al., 2002). Although these
specific enzymes efficiently digest enamel matrix proteins into smaller units, no
report to date has fully explained how these smaller organic units are removed
from the matrix.
Ameloblasts, being the only cells adjacent to the enamel matrix with
elongated Tomes’ processes are intimately attached to the basement membrane
of the developing enamel matrix and likely play a role the endocytosis of secreted
proteins. Early data from yeast two hybrid studies showed that one of the major
enamel proteins, amelogenin, interacted with LAMPs. LAMPs are necessary for
the structural integrity of the lysosomal membrane; helping to keep enzymes
separated from the cell cytoplasm. But LAMPs are also found on the plasma
membranes of ameloblasts and the membranes of endocytotic vesicles which led

77
to the hypothesis that enamel matrix proteins are removed from the developing
enamel through endocytic processes which involves the LAMPs.
Initially, unsure that cells had the ability to endocytose enamel matrix
proteins, I designed an experimental strategy to treat a non-amelogenin
producing cell with a recombinant, fluorescently tagged amelogenin. Not only
was the fluorescent protein rapidly endocytosed, it localized to the perinuclear
region of the cell, where late endosomes and lysosomes are located in addition
to the LAMPs. I next examined whether LAMPs co-localized with endocytosed
enamel matrix proteins and found that when various cell types were transiently
transfected with either LAMP1-GFP or CD63-GFP and treated with red
fluorescent tagged enamel matrix proteins, enamel matrix proteins were
packaged into LAMP1 and CD63 positive vesicles. As non ameloblast-like cells
were used in these initial experiments, I verified that ameloblast-like LS8 cells
also endocytosed exogenous fluorescent enamel matrix protein (data not
shown). Examination of endocytosis in live cells using time lapse confocal
microscopy showed the rapid uptake of enamel matrix proteins into a LAMP1
coated vesicle and contraction of the vesicle to concentrate the enamel proteins.
From these experiments, I concluded that there are pathways for the endocytosis
of enamel matrix proteins which are not ameloblast specific, and many different
cell types have the ability to endocytose enamel matrix proteins. Also, once
enamel proteins are endocytosed, they become concentrated in LAMP positive
vesicles.

78
I thought that because cells were able to effectively endocytose and
concentrate enamel matrix proteins into LAMP coated vesicles, there would be a
measurable transcriptional response to the treatment of cells with a high
concentration of exogenously added enamel matrix proteins. I found that my
hypothesis was incorrect, cells do not to any significant measurable degree up-
regulate LAMPs, or AP-3 sigma. The conclusion is that cells are able to
effectively endocytose enamel proteins when added at high concentrations to the
media, without the need for additional transcription of LAMPs or AP-3.
I hypothesized that if the endocytosis of enamel matrix proteins through
LAMPs is an important step in amelogenesis, than aberrations in normal
trafficking of LAMPs would result in tooth/enamel changes. I decided to examine
an animal model of HPS, in which mutations in the AP3B1 gene result in
abnormal trafficking of LAMPs. My findings from comparing AP3B1 null mice to
the C57BL/6J background strain show that there are significant size increases in
the incisor enamel layer, as well as significant overall size increases in the first
molars of the AP3B1 null mice, although I did not see any changes in the overall
enamel prismatic architecture of the teeth. AP3B1 nulls also had increased
staining of teeth, possibly due to a hyperplastic, hypomineralized enamel, as well
as advanced periodontal disease. Using FTIR analysis, the overall protein
content was found to be decreased in the AP3B1 nulls as well (see appendix 1).
These findings support the hypothesis that the endocytosis of enamel
matrix protein derivatives into LAMP coated vesicles is an important step in

79
amelogenesis, because mutations impacting on AP-3 activity, which disrupts the
normal trafficking of LAMPs, results in abnormal teeth.
Amelogenesis is a very complex process resulting in a very specialized
tissue, and although endocytosis is important in the removal of enamel matrix
proteins, other pathways, enzymes and mechanisms most probably play an
important role in the removal of protein from mature hardened enamel. New and
improved techniques will help to further unravel the intricacies of amelogenesis.

80
References

Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., and Watson, J.D. (1994).
Molecular Biology of the Cell, Third edn (New York, Garland Publishing, Inc.).
Andrejewski, N., Punnonen, E.L., Guhde, G., Tanaka, Y., Lullmann-Rauch, R.,
Hartmann, D., von Figura, K., and Saftig, P. (1999). Normal lysosomal
morphology and function in LAMP-1-deficient mice. J Biol Chem 274, 12692-
12701.
Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith,
J.A., and Struhl, K. (1990). Current Protocols in Molecular Biology. (New York,
Wiley).
Bartlett, J.D., Ganss, B., Goldberg, M., Moradian-Oldak, J., Paine, M.L., Snead,
M.L., Wen, X., White, S.N., and Zhou, Y.L. (2006). Protein-protein interactions of
the developing enamel matrix. Curr Topics Dev Biol 74, 57-115.
Bartlett, J.D., and Simmer, J.P. (1999). Proteinases in developing dental enamel.
Critical Reviews in Oral Biology and Medicine 10, 425-441.
Bartlett, J.D., Simmer, J.P., Xue, J., Margolis, H.C., and Moreno, E.C. (1996).
Molecular cloning and mRNA tissue distribution of a novel matrix
metalloproteinase isolated from porcine enamel organ. Gene 183, 123-128.
Beniash, E., Metzler, R.A., Lam, R.S.K., and Gilbert, P.U.P.A. (2009). Transient
amorphous calcium phosphate in forming enamel. Journal of Structural Biology
166, 133.
Berditchevski, F. (2001). Complexes of tetraspanins with integrins: more than
meets the eye. J Cell Sci 114, 4143-4151.
Boabaid, F., Gibson, C.W., Kuehl, M.A., Berry, J.E., Snead, M.L., Nociti, F.H.J.,
Katchburian, E., and Somerman, M.J. (2004). Leucine-rich amelogenin peptide: a
candidate signaling molecule during cementogenesis. J Periodontol 75, 1126-
1136.
Bonifacino, J.S., and Traub, L.M. (2003). Signals for sorting of transmembrane
proteins to endosomes and lysosomes. Annu Rev Biochem 72, 395-447.
Boyan, B.D., Weesner, T.C., Lohmann, C.H., Andreacchio, D., Carnes, D.L.,
Dean, D.D., Cochran, D.L., and Schwartz, Z. (2000). Porcine fetal enamel matrix
derivative enhances bone formation induced by demineralized freeze dried bone
allograft in vivo. J Periodontol 71, 1278-1286.

81
Brinkman, B.M. (2004). Splice variants as cancer biomarkers. Clin Biochem 37,
584-594.
Caterina, J.J., Skobe, Z., Shi, J., Ding, Y., Simmer, J.P., Birkedal-Hansen, H.,
and Bartlett, J.D. (2002). Enamelysin (MMP-20) deficient mice display an
amelogenesis imperfecta phenotype. J Biol Chem 277, 49598-49604.
Chen, H.H., Wang, Y.C., and Fann, M.J. (2006). Identification and
characterization of the CDK12/cyclin L1 complex involved in alternative splicing
regulation. Mol Cell Biol 26, 2736-2745.
Chen, L.S., Couwenhoven, D., Hsu, D., Luo, W., and Snead, M.L. (1992).
Maintainance of amelogenin gene expression by transformed epithelial cells of
mouse enamel organ. Archs oral Biol 37, 771-778.
Cook, N.R., Row, P.E., and Davidson, H.W. (2004). Lysosome associated
membrane protein 1 (Lamp1) traffics directly from the TGN to early endosomes.
Traffic 5, 685-699.
Dell'Angelica, E.C., Klumperman, J., Stoorvogel, W., and Bonifacino, J.S. (1998).
Association of the AP-3 adaptor complex with clathrin. Science 280, 431-434.
Dell'Angelica EC, O.H., Ooi CE, Rabinovich E, Roche KW, Bonifacino JS. (1997).
AP-3: an adaptor-like protein complex with ubiquitous expression. Embo J 16,
917-928.
Dell'Angelica, E.C., Shotelersuk, V., Aguilar, R.C., Gahl, W.A., and Bonifacino,
J.S. (1999). Altered Trafficking of Lysosomal Proteins in Hermansky-Pudlak
Syndrome Due to Mutations in the [beta]3A Subunit of the AP-3 Adaptor.
Molecular Cell 3, 11.
Deutsch, D., Palmon, A., Fisher, L.W., Kolodny, N., Termine, J.D., and Young,
M.F. (1991). Sequencing of bovine enamelin ("tuftelin") a novel acidic enamel
protein. J Biol Chem 266, 16021-16028.
Dhamija, S., and Krebsbach, P.H. (2001). Role of Cbfa1 in ameloblastin gene
transcription. J Biol Chem 276, 35159-35164.
Dhamija, S., Liu, Y., Yamada, Y., Snead, M.L., and Krebsbach, P.H. (1999).
Cloning and characterization of the murine ameloblastin promoter. J Biol Chem
274, 20738-20743.
Di Pietro, S.M., and Dell'Angelica, E.C. (2005). The cell biology of Hermansky-
Pudlak syndrome: recent advances. Traffic 6, 525-533.

82
Di Pietro, S.M., Falcon-Perez, J.M., Tenza, D., Setty, S.R., Marks, M.S., Raposo,
G., and Dell'Angelica, E.C. (2006). BLOC-1 interacts with BLOC-2 and the AP-3
complex to facilitate protein trafficking on endosomes. Mol Biol Cell 17, 4027-
4038.
Duffield, A., Kamsteeg, E.J., Brown, A.N., Pagel, P., and Caplan, M.J. (2003).
The tetraspanin CD63 enhances the internalization of the H,K-ATPase beta-
subunit. Proc Natl Acad Sci U S A 100, 15560-15565.
Eskelinen, E.L., Illert, A.L., Tanaka, Y., Schwarzmann, G., Blanz, J., Von Figura,
K., and Saftig, P. (2002). Role of LAMP-2 in lysosome biogenesis and
autophagy. Mol Biol Cell 13, 3355-3368.
Eskelinen, E.L., Schmidt, C.K., Neu, S., Willenborg, M., Fuertes, G., Salvador,
N., Tanaka, Y., Lullmann-Rauch, R., Hartmann, D., Heeren, J., et al. (2004).
Disturbed cholesterol traffic but normal proteolytic function in LAMP-1/LAMP-2
double-deficient fibroblasts. Mol Biol Cell 15, 3132-3145.
Eskelinen, E.L., Tanaka, Y., and Saftig, P. (2003). At the acidic edge: emerging
functions for lysosomal membrane proteins. Trends Cell Biol 13, 137-145.
Feliciano, N.Z., Rivera, E., Agrait, E., and Rodriguez, K. (2006). Hermansky-
Pudlak syndrome: dental management considerations. J Dent Child (Chic) 73,
51-56.
Feng, L., Seymour, A.B., Jiang, S., To, A., Peden, A.A., Novak, E.K., Zhen, L.,
Rusiniak, M.E., Eicher, E.M., Robinson, M.S., et al. (1999). The beta3A subunit
gene (Ap3b1) of the AP-3 adaptor complex is altered in the mouse
hypopigmentation mutant pearl, a model for Hermansky-Pudlak syndrome and
night blindness. Hum Mol Genet 8, 323-330.
Fong, C.D., Slaby, I., and Hammarstrom, L. (1996). Amelin, an enamel related
protein, transcribed in cells of the epithelial root sheath. J Bone Min Res 11, 892-
898.
Franklin, D.L., Severs, N.J., and Katchburian, E. (1991). Development of the
distal end and Tomes' processes of ameloblasts observed by freeze-fracture and
ultrathin section electron microscopy. J Anat 174, 103-114.
Fukumoto, S., Kiba, T., Hall, B., Iehara, N., Nakamura, T., Longenecker, G.,
Krebsbach, P.H., Nanci, A., Kulkarni, A.B., and Yamada, Y. (2004). Ameloblastin
is a cell adhesion molecule required for maintaining the differentiation state of
ameloblasts. J Cell Biol 167, 973-983.

83
Greene, E.F., Tauch, S., Webb, E., and Amarasiriwardena, D. (2004). Application
of diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) for the
identification of potential diagenesis and crystallinity changes in teeth.
Microchemical Journal 76, 141.
Hermansky, F., and Pudlak, P. (1959). Albinism Associated with Hemorrhagic
Diathesis and Unusual Pigmented Reticular Cells in the Bone Marrow: Report of
Two Cases with Histochemical Studies. Blood 14, 162-169.
Herrlich, P., Sleeman, J., Wainwright, D., Konig, H., Sherman, L., Hilberg, F., and
Ponta, H. (1998). How tumor cells make use of CD44. Cell Adhes Communic 6,
141-147.
Holcombe, R.F., Baethge, B.A., Stewart, R.M., Betzing, K., Hall, V.C., Fukuda,
M., and Wolf, R.E. (1993). Cell surface expression of lysosome-associated
membrane proteins (LAMPs) in scleroderma: relationship of lamp2 to disease
duration, anti-Sc170 antibodies, serum interleukin-8, and soluble interleukin-2
receptor levels. Clin Immunol Immunopathol 67, 31-39.
Hotta H, T.N., Homma M. (1989). Transcriptional enhancement of the human
gene encoding for a melanoma-associated antigen (ME491) in association with
malignant transformation. Cancer Science 80, 1186-1191.
Hu, C.-C., Fukae, M., Uchida, T., Qian, Q., Zhang, C.H., Ryu, O.H., Tanabe, T.,
Yamakoshi, Y., Murakami, C., Dohi, N., et al. (1997). Cloning and
characterization of porcine enamelin mRNAs. J Dent Res 76, 1720-1729.
Hu, C.C., Hart, T.C., Dupont, B.R., Chen, J.J., Sun, X., Qian, Q., Zhang, C.H.,
Jiang, H., Mattern, V.L., Wright, J.T., et al. (2000). Cloning human enamelin
cDNA, chromosomal localization, and analysis of expression during tooth
development. J Dent Res 79, 912-919.
Hu, J.C., Sun, X., Zhang, C., Liu, S., Bartlett, J.D., and Simmer, J.P. (2002).
Enamelysin and kallikrein-4 mRNA expression in developing mouse molars. Eur
J Oral Sci 110, 307-315.
Hu, J.C., Zhang, C.H., Yang, Y., Karrman-Mardh, C., Forsman-Semb, K., and
Simmer, J.P. (2001). Cloning and characterization of the mouse and human
enamelin genes. J Dent Res 80, 898-902.
Iwasaki, K., Bajenova, E., Somogyi-Ganss, E., Miller, M., Nguyen, V.,
Nourkeyhani, H., Gao, Y., Wendel, M., and Ganss, B. (2005). Amelotin--a Novel
Secreted, Ameloblast-specific Protein. J Dent Res 84, 1127-1132.
Izquierdo, N.J. (2008). Hermansky-Pudlak syndrome.
[http://emedicine.medscape.com/article/1200277-overview].

84
Johannes, L., and Lamaze, C. (2002). Clathrin-dependent or not: is it still the
question? Traffic 3, 443-451.
Jung, J., Bohn, G., Allroth, A., Boztug, K., Brandes, G., Sandrock, I., Schaffer,
A.A., Rathinam, C., Kollner, I., Beger, C., et al. (2006a). Identification of a
homozygous deletion in the AP3B1 gene causing Hermansky-Pudlak syndrome,
type 2. Blood 108, 362-369.
Jung, K.K., Liu, X.W., Chirco, R., Fridman, R., and Kim, H.R. (2006b).
Identification of CD63 as a tissue inhibitor of metalloproteinase-1 interacting cell
surface protein. Embo J 25, 3934-3942.
Kalnina, Z., Zayakin, P., Silina, K., and Line, A. (2005). Alterations of pre-mRNA
splicing in cancer. Genes Chromosomes Cancer 42, 342-357.
Kannan, K., Stewart, R.M., Bounds, W., Carlsson, S.R., Fukuda, M., Betzing,
K.W., and Holcombe, R.F. (1996). h-LAMP2 (CD107b) are activation-dependent
cell surface glycoproteins in human peripheral blood mononuclear cells which
mediate cell adhesion to vascular endothelium. Cell Immunol 171, 10-19.
Kantheti, P., Qiao, X., Diaz, M.E., Peden, A.A., Meyer, G.E., Carskadon, S.L.,
Kapfhamer, D., Sufalko, D., Robinson, M.S., Noebels, J.L., et al. (1998). Mutation
in AP-3 delta in the mocha mouse links endosomal transport to storage
deficiency in platelets, melanosomes, and synaptic vesicles. Neuron 21, 111-
122.
Karlin, S., and Altschul, S.F. (1990). Methods for assessing statistical
significance of molecular sequence features by using general scoring schemes.
Proc Natl Acad Sci USA 87, 2264-2268.
Kawana, F., Sawae, Y., Sahara, T., Tanaka, S., Debari, K., Shimizu, M., and
Sasaki, T. (2001). Porcine enamel matrix derivative enhances trabecular bone
regeneration during wound healing of injured rat femur. Anat Rec 264, 438-446.
King, A.C., and Cuatrecasas, P. (1981). Long-term epidermal growth factor-
receptor internalization and processing in quiescent human fibroblasts. J
Supramol Struct Cell Biochem 17, 377-387.
King, A.C., Willis, R.A., and Cuatrecasas, P. (1980). Accumulation of epidermal
growth factor within cells does not depend on receptor recycling. Biochem
Biophys Res Commun 97, 840-845.
Lacruz, R.S., and Bromage, T.G. (2006). Appositional enamel growth in molars
of South African fossil hominids. J Anat 209, 13-20.

85
Lee, S.K., Kim, S.M., Lee, Y.J., Yamada, K.M., Yamada, Y., and Chi, J.G.
(2003). The structure of the rat ameloblastin gene and its expression in
amelogenesis. Mol Cells 15, 216-225.
Lee, S.K., Krebsbach, P., Matsuki, Y., Nanci, A., Yamada, K.M., and Yamada, Y.
(1996). Ameloblastin expression in rat incisors and human tooth germs. Int J Dev
Biol 40, 1141-1150.
Luo, W., Wen, X., Wang, H.J., MacDougall, M., Snead, M.L., and Paine, M.L.
(2004). In vivo overexpression of tuftelin in the enamel organic matrix. Cells
Tissues Organs 177, 212-220.
Mantegazza, A.R., Barrio, M.M., Moutel, S., Bover, L., Weck, M., Brossart, P.,
Teillaud, J.L., and Mordoh, J. (2004). CD63 tetraspanin slows down cell
migration and translocates to the endosomal-lysosomal-MIICs route after
extracellular stimuli in human immature dendritic cells. Blood 104, 1183-1190.
Margolis, H.C., Beniash, E., and Fowler, C.E. (2006). Role of macromolecular
assembly of enamel matrix proteins in enamel formation. J Dent Res 85, 775-
793.
Maycock, J., Wood, S.R., Brookes, S.J., Shore, R.C., Robinson, C., and
Kirkham, J. (2002). Characterization of a porcine amelogenin preparation,
EMDOGAIN, a biological treatment for periodontal disease. Connect Tissue Res
4392-3), 472-476.
McNiven, M.A. (2006). Big gulps: specialized membrane domains for rapid
receptor-mediated endocytosis. Trends Cell Biol 16, 487-492.
Moffatt, P., Smith, C.E., St-Arnaud, R., Simmons, D., Wright, J.T., and Nanci, A.
(2006). Cloning of rat amelotin and localization of the protein to the basal lamina
of maturation stage ameloblasts and junctional epithelium. Biochem J 399, 37-
46.
Moradian-Oldak, J., Paine, M.L., Lei, Y.P., Fincham, A.G., and Snead, M.L.
(2000). Self-assembly properties of recombinant engineered amelogenin proteins
analyzed by dynamic light scattering and atomic force microscopy. Journal of
Structural Biology 131, 27-37.
Nakamura, M., Bringas, P., Nanci, A., Zeichner-David, M., Ashdown, B., and
Slavkin, H.C. (1994). Translocation of enamel proteins from inner enamel
epithelia to odontoblasts during mouse tooth development. The Anatomical
Record 238, 383-396.

86
Nakatsu, F., and Ohno, H. (2003). Adaptor protein complexes as the key
regulators of protein sorting in the post-Golgi network. Cell Struct Funct 28, 419-
429.
Nanci, A., Bendayan, M., and Slavkin, H.C. (1985). Enamel protein biosynthesis
and secretion in mouse incisor secretory ameloblasts as revealed by high-
resolution immunocytochemistry. J Histochem Cytochem 33, 1153-1160.
Nanci, A., Fortin, M., and Ghitescu, L. (1996). Endocytotic functions of
ameloblasts and odontoblasts: Immunohistochemical and tracer studies on the
uptake of plasma proteins. Anat Rec 245, 219-234.
Nanci, A., Slavkin, H.C., and Smith, C.E. (1987). Immunocytochemical and
radioautographic evidence for secretion and intracellular degradation of enamel
proteins by ameloblasts in rat incisors. Anat Rec 217, 107-123.
Orth, J.D., Krueger, E.W., Weller, S.G., and McNiven, M.A. (2006). A novel
endocytic mechanism of epidermal growth factor receptor sequestration and
internalization. Cancer Res 66, 3603-3610.
Paine, C.T., Paine, M.L., Luo, W., Okamoto, C.T., Lyngstadaas, S.P., and
Snead, M.L. (2000a). A tuftelin-interacting protein (TIP39) localizes to the apical
secretory pole of mouse ameloblasts. J Biol Chem 275, 22284-22292.
Paine, M.L., White, S.N., Luo, W., Fong, H., Sarikaya, M., and Snead, M.L.
(2001). Regulated gene expression dictates enamel structure and tooth function.
Matrix Biol 20, 273-292.
Paine, M.L., Zhu, D.H., Luo, W., Bringas, P.J., Goldberg, M., White, S.N., Lei,
Y.P., Sarikaya, M., Fong, H.K., and Snead, M.L. (2000b). Enamel
biomineralization defects result from alterations to amelogenin self-assembly.
Journal of Structural Biology 132, 191-200.
Peden, A.A., Oorschot, V., Hesser, B.A., Austin, C.D., Scheller, R.H., and
Klumperman, J. (2004). Localization of the AP-3 adaptor complex defines a novel
endosomal exit site for lysosomal membrane proteins. J Cell Biol 164, 1065-
1076.
Peden, A.A., Rudge, R.E., Lui, W.W., and Robinson, M.S. (2002). Assembly and
function of AP-3 complexes in cells expressing mutant subunits. J Cell Biol 156,
327-336.
Pilar, G., Cristina, P.a., Victor Hugo, L., and Pedro, B. (2005). Characterization of
enamel with variable caries risk. Archives of oral biology 50, 843.

87
Ponta, H., Sherman, L., and Herrlich, P.A. (2003). CD44: from adhesion
molecules to signalling regulators. Nat Rev Mol Cell Biol 4, 33-45.
Rahiotis, C., Vougiouklakis, G., and Eliades, G. (2008). Characterization of oral
films formed in the presence of a CPP-ACP agent: An in situ study. Journal of
Dentistry 36, 272.
Rajesh S. Mathur, S.P.M.R.C.Y. (2000). Up-Regulation of Epidermal Growth
Factor-Receptors (EGF-R) by Nicotine in Cervical Cancer Cell Lines: This Effect
May Be Mediated by EGF. American Journal of Reproductive Immunology 44,
114-120.
Reith, E.J., and Cotty, V.F. (1967a). The absorptive activity of ameloblasts during
the maturation of enamel. Anat Rec 157, 577-587.
Reith, E.J., and Cotty, V.F. (1967b). The absorptive activity of ameloblasts during
the maturation of enamel. Anat Rec 157, 577-588.
Reseland, J.E., Reppe, S., Larsen, A.M., Berner, H.S., Reinholt, F.P., Gautvik,
K.M., Slaby, I., and Lyngstadaas, S.P. (2006). The effect of enamel matrix
derivative (EMD) on gene expression in osteoblasts. Eur J Oral Sci 114
(Supplement 1), 205-211.
Robinson, C., Brookes, S.J., Shore, R.C., and Kirkham, J. (1998). The
developing enamel matrix: nature and function. Eur J Oral Sci 106 Suppl. 1, 282-
291.
Robinson, M.S. (2004). Adaptable adaptors for coated vesicles. Trends Cell Biol
14, 167-174.
Robinson, M.S., and Bonifacino, J.S. (2001). Adaptor-related proteins. Curr Opin
Cell Biol 13, 444-453.
Rous, B.A., Reaves, B.J., Ihrke, G., Briggs, J.A., Gray, S.R., Stephens, D.J.,
Banting, G., and Luzio, J.P. (2002). Role of adaptor complex AP-3 in targeting
wild-type and mutated CD63 to lysosomes. Mol Biol Cell 13, 1071-1082.
Sambrook, J., and Russell, D.W. (2001). Molecular Cloning: A Laboratory
Manual (New York, Cold Spring Harbor Laboratory Press).
Shapiro, J.L., Wang, H., Wen, X., Tannukit, S., and Paine, M.L. (2006). An
amelogenin minigene to study alternative splicing. DNA Cell Biol 25, 1-5.
Shi, G., Faundez, V., Roos, J., Dell'Angelica, E.C., and Kelly, R.B. (1998).
Neuroendocrine synaptic vesicles are formed in vitro by both clathrin-dependent
and clathrin-independent pathways. J Cell Biol 143, 947-955.

88
Silverstein, R.L., and Febbraio, M. (1992). Identification of lysosome-associated
membrane protein-2 as an activation-dependent platelet surface glycoprotein.
Blood 80, 1470-1475.
Simmer, J.P., and Fincham, A.G. (1995). Molecular mechanisms of dental
enamel formation. Crit Rev Oral Biol Med 6, 84-108.
Simmer, J.P., Fukae, M., Tanabe, T., Yamakoshi, Y., Uchida, T., Xue, J.,
Margolis, H.C., Shimizu, M., DeHart, B.C., Hu, C.-C., et al. (1998). Purification,
characterization and cloning of enamel matrix serine proteinase 1. J Dent Res
77, 377-386.
Simmer, J.P., and Snead, M.L. (1995). Molecular biology of the amelogenin
gene. Dental Enamel: Formation to Destruction, 59-84.
Simmer, P.J., Lau, E.C., Hu, C.C., Moradian-Oldak, J., Sarte, P.E., Slavkin, H.C.,
and Fincham, A.G. (1994). Alternative splicing of the amelogenin primary RNA
transcript in mice. Calcif Tissue Intl 55, 302-310.
Simpson, F., Bright, N.A., West, M.A., Newman, L.S., Darnell, R.B., and
Robinson, M.S. (1996). A novel adaptor-related protein complex. J Cell Biol 133,
749-760.
Simpson, F., Peden, A.A., Christopoulou, L., and Robinson, M.S. (1997).
Characterization of the Adaptor-related Protein Complex, AP-3. J Cell Biol 137,
835-845.
Smith, C.E. (1979). Ameloblasts: secretory and resorptive functions. J Dent Res
58(Spec. Issue B), 695-707.
Smith, C.E. (1998). Cellular and chemical events during enamel maturation. Crit
Rev Oral Biol Med 9, 128-161.
Smith, C.E., Dahan, S., Fazel, A., W., L., and Nanci, A. (1992). Correlated
biochemical and radioautographic studies of protein turnover in developing rat
incisor enamel following pulse-chase labeling with L-[35S]- and L-[methyl-3H]-
methionine. Anat Rec 232, 1-14.
Smith, C.E., and Nanci, A. (1996). Protein dynamics of amelogenesis. Anat Rec
245, 186-207.
Snead, M.L., Lau, E.C., Zeichner-David, M., Fincham, A.G., Woo, S.L., and
Slavkin, H.C. (1985). DNA sequence for cloned cDNA for murine amelogenin
reveal the amino acid sequence for enamel-specific protein. Biochemical and
Biophysical Research Communications 129, 812-818.

89
Snead, M.L., Paine, M.L., Chen, L.S., Yoshida, B., Luo, W., Zhu, D.-H., Lei, Y.-
P., Liu, Y.-H., and Maxson, R.E.J. (1996). The murine amelogenin promoter:
Developmentally regulated expression in transgenic animals. Connect Tissue
Res 35, 41-47.
Snead, M.L., Zeichner-David, M., Chandra, T., Robson, K.J., Woo, S.L., and
Slavkin, H.C. (1983). Construction and identification of mouse amelogenin cDNA
clones. Proc Natl Acad Sci USA 80, 7254-7258.
Socrates, G. (2001). Infrared and Raman Characteristic Group Frequencies, Vol
3rd Edition (West Sussex England, John Wiley and Sons, LTD.).
Stipp, C.S., Kolesnikova, T.V., and Hemler, M.E. (2003). Functional domains in
tetraspanin proteins. Trends Biochem Sci 28, 106-112.
Stoss, O., Stoilov, P., Hartmann, A.M., Nayler, O., and Stamm, S. (1999). The in
vivo minigene approach to analyze tissue-specific splicing. Brain Res Brain Res
Protoc 4, 383-394.
Termine, J.D., Belcourt, A.B., Christner, P.J., Conn, K.M., and Nylen, M.U.
(1980). Properties of dissociatively extracted fetal tooth matrix proteins. J Biol
Chem 255, 9760-9768.
Tompkins, K., George, A., and Veis, A. (2006). Characterization of a mouse
amelogenin [A-4]/M59 cell surface receptor. Bone 38, 172-180.
Veis, A. (2003). Amelogenin gene splice products: potential signaling molecules.
Cell Mol Life Sci 60, 38-55.
Venables, J.P. (2004). Aberrant and alternative splicing in cancer. Cancer Res
64, 7647-7654.
Verdelis, K., Lukashova, L., Wright, J.T., Mendelsohn, R., Peterson, M.G.E.,
Doty, S., and Boskey, A.L. (2007). Maturational changes in dentin mineral
properties. Bone 40, 1399.
Viswanathan, H.L., Berry, J.E., Foster, B.L., Gibson, C.W., Li, Y., Kulkarni, A.B.,
Snead, M.L., and Somerman, M.J. (2003). Amelogenin: a potential regulator of
cementum-associated genes. J Periodontol 74, 1423-1431.
Wang, H.J., Tannukit, S., Zhu, D.H., Snead, M.L., and Paine, M.L. (2005).
Enamel matrix protein interactions. J Bone Mineral Res 20, 1032-1040.

90
WM Canfield, K.J., RD Ye, W Gregory, and S Kornfeld Localization of the signal
for rapid internalization of the bovine-cation independent mannose 6
phosphate/insulin like growth factor II receptor to amino acids 24-29 of the
cytoplasmic tail. 266, 5682-5688.
Woessner, J.F., Jr. (1998). Role of matrix proteases in processing enamel
proteins. Connect Tissue Res 39, 69-73; discussion 141-149.
Xu, L., Harada, H., and Taniguchi, A. (2006a). The Exon 6ABC Region of
Amelogenin mRNA Contribute to Increased Levels of Amelogenin mRNA through
Amelogenin Protein-enhanced mRNA Stabilization. J Biol Chem 281, 32439-
32444.
Xu, L., Harada, H., Yokohama-Tamaki, T., Matsumoto, S., Tanaka, J., and
Taniguchi, A. (2006b). Reuptake of extracellular amelogenin by dental epithelial
cells results in increased levels of amelogenin mRNA through enhanced mRNA
stabilization. J Biol Chem 281, 2257-2262.
Xu, Y., Zhou, Y.L., Ann, D.K., MacDougald, O.A., Shum, L., and Snead, M.L.
(2006c). Transcription factor sumoylation and factor YY1 serve to modulate
mouse amelogenin gene expression. Eur J Oral Sci 114 Suppl 1, 169-177;
discussion 201-162, 381.
Xu, Y., Zhou, Y.L., Luo, W., Zhu, Q.S., Levy, D., MacDougald, O.A., and Snead,
M.L. (2006d). NF-Y and CCAAT/enhancer-binding protein alpha synergistically
activate the mouse amelogenin gene. J Biol Chem 281, 16090-16098.
Yang, W., Li, C., Ward, D.M., Kaplan, J., and Mansour, S.L. (2000). Defective
organellar membrane protein trafficking in Ap3b1-deficient cells. J Cell Sci 113,
4077-4086.
Yokoyama, Y., Ishiguro, R., Maeda, H., Mukaiyama, M., Kameyama, K., and
Hiramatsu, K. (2003). Quantitative analysis of protein adsorption on a planar
surface by Fourier transform infrared spectroscopy: lysozyme adsorbed on
hydrophobic silicon-containing polymer. Journal of Colloid and Interface Science
268, 23.
Yunta, M., and Lazo, P.A. (2003). Tetraspanin proteins as organisers of
membrane microdomains and signalling complexes. Cell Signal 15, 559-564.
Zhou, Y.L., Lei, Y., and Snead, M.L. (2000). Functional antagonism between
Msx2 and CCAAT/enhancer-binding protein alpha in regulating the mouse
amelogenin gene expression is mediated by protein-protein interaction. J Biol
Chem 275, 29066-29075.

91
Zhou, Y.L., and Snead, M.L. (2000). Identification of CCAAT/enhancer-binding
protein alpha as a transactivator of the mouse amelogenin gene. J Biol Chem
275, 12273-12280.
Zhu, D., Paine, M.L., Luo, W., Bringas, P., Jr., and Snead, M.L. (2006). Altering
biomineralization by protein design. J Biol Chem 281, 21173-21182.
Zou, Y., Wang, H., Shapiro, J.L., Okamoto, C.T., Brookes, S.J., Lyngstadaas,
S.P., Snead, M.L., and Paine, M.L. (2007). Determination of protein regions
responsible for interactions of amelogenin with CD63 and LAMP1. Biochem J.

92
Appendix

Characterization of the enamel surface of mouse teeth by FTIR
spectroscopy (as reported from Dr. Petter Lyngstadaas’s lab)
Materials and Methods
2 groups of mouse teeth were analysed.
1. WT – n=3; 1 incisor and 2 molars
2. M – n=3; 1 incisor and 2 molars
The specimen had been received on Feb 10, 2009. They were wrapped in paper
and stored in microcentrifuge tubes at room temperature until analysis. The
samples were analysed with FTIR spectroscopy. To receive a spectrum of good
quality, the teeth were pestled in a ceramic mortar. The resulting powder was
carefully fixed on double sided tape (Scotch, 3M) and placed on the DRIFT
sample holder. A scan of the tape was used as C57BL/6J. A scan of pure
hydroxyapatite (HA) powder (SigmaAldrich, St. Louis, USA) was used to illustrate
the absorbance of the pure mineral phase, since enamel mainly consists of HA
crystals.
The powder samples were analysed by FTIR spectroscopy (Spectrum 400,
PerkinElmer, Inc. Waltham, MA, USA) in DRIFT mode. The parameters used for
each measurement were:
Wavenumber range: 4000 cm
-1
- 450 cm
-1

Number of scans: 8
Resolution: 4 cm
-1
The spectra measured were analysed for typical absorbances characteristic for
protein groups and mineral groups. In the obtained spectra, peak areas were
measured for absorbances typical for mineral phase or for proteins, using
Spectrum 6.3.2.0151 (PerkinElmer, Inc., Waltham, MA, USA).
The Ca content (cCa) was measured for each sample by Atomic absorbance
spectroscopy (AAS; AAnalyst 400, PerkinElmer, Inc. Waltham, MA, USA), in
order to determine the amount of powder in each sample analysed in the FTIR.
The quality of the FTIR spectra depends on the amount of sample used for the
measurement; thus the resulting peak areas of characteristic absorbances are
dependent on the amount of powder (Greene et al., 2004). One method to
normalize data would be to calculate the ratio of carbonate absorbance (CO
3
2-
,
~1430 cm
-1
) and phosphate absorbance (PO
4
, ~1034 cm
-1
) in order to normalize
against this ratio (Greene et al., 2004). The peak area for the PO
4
absorbance
could not be determined in the spectra for the tooth samples, due to overlay with

93
other absorbances. The peak areas were therefore normalized against cCa
[ppm].
Results
The following tables give an indication of which peaks were to be expected in the
FTIR spectra for a mixture of enamel (mainly mineral phase) and dentin
(minerals+proteins).
Tooth enamel contains HA, with characteristic absorbances at the following
wavenumbers (Table 1):

Group Wavenumber
[cm
-1
]
Source
OH stretch 3500 (Rahiotis et al., 2008)
CO
3
2-
1400-1500 (Pilar et al., 2005)
CO
3
2-
1419
(Greene et al., 2004)
F 1097
P-O 1030-1090 (Pilar et al., 2005)
PO
4
3-
1034 (Greene et al., 2004)
P-O 954
(Pilar et al., 2005)
CO
3
2-
840-890
PO
4
scissor
O-P-O scissor
600 (Pilar et al., 2005)

The most characteristic absorbances of proteins in the FTIR are (Table 2):
Group Wavenumber
[cm
-1
]
Source
NH stretch 3300
(Yokoyama et al., 2003)
(Socrates, 2001)
Overtone
amide II
3100
C=N 2240
C=N 2120
Amide I 1655
Amide II 1565
Amide III 1300
sym CNC
stretch
800-900
Amide IV 725

94

Changes of peak areas were compared for the absorbances printed in bold
letters.

1. Absorbances typical for teeth

Figure 1: Spectrum of powder of WT incisor, and powder of M incisor. The HA
spectrum is given to clarify the difference of absorbances of the mineral phase
and of proteins.
2. Correlation between cCa and peak areas
Correlation between the peak areas of the mineral phase was performed by
Pearson Product Moment Correlation in SigmaStat 3.5. Results are given in
Table 3:
860cm-1 1430cm-1 cCa
600cm-1

0.63
0.18
0.489
0.325
0.114
0.83
Correlation Coefficient
P Value
860cm-1

0.945
0.00445
0.573
0.234
Correlation Coefficient
P Value
1430cm-1

0.411
0.418
Correlation Coefficient
P Value
Table 3: Product moment correlation study for non-normalized peak areas, n=6
for all groups.

95
It became clear from this correlation study, that cCa and calculated peak areas,
except the absorbances at 600 cm
-1
are well correlated. The areas of different
absorbances of one sample are well correlated. We thus reasoned that sample
loading influences the peak area (as pointed out by Greene EF et al. 2004) in
addition to protein content. cCa was in average slightly lower for WT teeth,
however the difference in cCA for the 2 groups was not significant.
3. Peak areas
Peak areas were divided by cCa in ppm. The averages ± stdev in % relative to
WT are given in Figure 2. The graph is divided into 2 areas; absorbances of
mineral functional groups and absorbances of protein functional groups. No
significant differences in the absorbances were found. However, the overall
absorbance for proteins was about 35.4% lower for AP3B1 null teeth than for
teeth of the C57BL/6J mice (p=0.003).

Figure 2: Comparison of the percentage of peak areas for the absorbances of
functional groups typical for hydroxyl apatite and proteins (Average ± Stdev).

96

Discussion
Weaknesses of the study:
Dentin was present in the samples, which lowers the accuracy of measurements
of protein in enamel. Possibly, by using only the incisors which contain a higher
ratio enamel/dentin, the measurements could become more precise.
The number of specimen was very low with n=3 per group, which resulted in a
low statistical impact.
The amount of Ca in each sample could depend on the enamel structure which
could be influenced in the AP3B1 null mice. The method used assumed that all
the samples contained the same ratio of mineral to non-mineral substance.
The amount of tooth powder on each sample was not measured directly.
The quality of the spectra could be influenced by the crystal structure of HA in the
enamel.
On some tooth samples bone remained around the roots, however, the teeth
where bone was observed did not have an outstanding cCa or absorbance of
mineral or protein groups.
Alternatives
Use of incisors with a higher ratio enamel/dentin could give more precise results.
A larger number of specimens could be analysed to achieve statistical relevance.
The use of the FTIR microscope could be considered. Other studies analysed
teeth sections of 10 µm thickness (Beniash et al., 2009) or as decalcified
sections of 2 µm thickness (Verdelis et al., 2007) in transmission mode.
Comments on the report from 09-26-03
In the report on amelotin transgenic mice (08-25-11), we concluded that the
group termed Tg had a higher protein absorbance at certain wavelengths (3221,
3109, 1646, 1580, 1300, 845, and 727 cm
-1
). In the former report, the results for
those specimens were included in the calculations and increased the significance
of the results (i.e. higher protein content in the WT group). Reason for this was
that the absorbance areas of mineral and protein groups in amelotin transgenic
mice were negatively correlated to the Ca content; thus they increased the
significance of the current study with AP-3 mice as concluded in the report from
09-26-03.

97

Conclusion
A higher absorbance signal at wavelengths typical for protein absorbances
was detected for the wild type sample. A relative quantitative comparison
showed that M had in average 35.4% lower peak areas of absorbance at
wavenumbers typical for proteins. This difference was significant (p=0.003).
The results indicated that the WT samples contained a higher amount of
peptides in the teeth. The values for the single protein absorbances at the
respective wavelengths are not significantly different due to low number of
specimen, but a tendency for lower protein absorbance was seen.

Asset Metadata

Creator Shapiro, Jason Lee (author)

Core Title Integrative analysis of gene expression and phenotype data

Contributor Electronically uploaded by the author (provenance)

School School of Dentistry

Degree Doctor of Philosophy

Degree Program Craniofacial Biology

Publication Date 07/29/2009

Defense Date 06/11/2009

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

Tag adaptor protein 3,amelogenesis,enamel,endocytosis,LAMP proteins,OAI-PMH Harvest

Language English

Advisor Paine, Michael L. (committee chair), Chen, Casey (committee member), Okamoto, Curtis Toshio (committee member), Shuler, Charles F. (committee member), Snead, Malcolm L. (committee member)

Creator Email jasonsha@usc.edu,shapiro_jason@hotmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2417

Unique identifier UC1142299

Identifier etd-Shapiro-2609 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-406502 (legacy record id),usctheses-m2417 (legacy record id)

Legacy Identifier etd-Shapiro-2609.pdf

Dmrecord 406502

Document Type Dissertation

Rights Shapiro, Jason Lee

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 uscdl@usc.edu

Abstract (if available)

Abstract Dental enamel is the hardest mineralized tissue in the body and initially forms from a protein matrix. During enamel formation, ameloblasts secrete unique proteins into the extracellular space to direct mineralization of enamel crystallites. This specialized organic matrix contains the proteins amelogenin, ameloblastin, and enamelin. Amelogenin is the major enamel matrix protein secreted, comprising ~95% of the total matrix, yet after maturation and completion of amelogenesis, only ~1% of the initial organic matrix remains in mineralized enamel. Ameloblasts, being the only cells in close proximity to the organic matrix, most likely are responsible for the removal of proteinaceous matrix debris for proper enamel maturation. The rapid removal of enamel matrix proteins is poorly understood in the literature, principally because clathrin (the prototypical pathway for endocytosis) is expressed at very low levels in ameloblasts. Most of the historical record of endocytosis is focused on clathrin mediated endocytosis whereby molecules are endocytosed into clathrin coated vesicles, pinched off from the plasma membrane into the cytoplasm, and fused with early endosomes. From the early endosome, these molecules are targeted to the late endosome and eventually to the lysosome where they are digested by a myriad of proteolytic enzymes.