INSIGHT INTO THE ROLE OF PERIODONTAL LIGAMENT ASSOCIATED

PROTEIN-1/ASPORIN IN THE MAINTENANCE OF THE PERIODONTAL LIGAMENT USING A RAT ANKYLOSIS

MODEL

Doctor of Clinical Dentistry (Orthodontics) Thesis

Wayne Chen

Orthodontic Unit

School of Dentistry Faculty of Health Science The University of Adelaide

South Australia AUSTRALIA

2012

1

Table of Contents Table of Contents 1

Figures and Tables 3

Glossary of Abbreviated Terms 5

Statement 7

Acknowledgements 8

Summary 9

Section 1 11

Literature review 12

Comparative Dental Anatomy 12

The Periodontium 16

Cementum 16

Bone 17

Bone metabolism 18

Cellular constituents 26

Periodontal ligament 27

Maintenance of the periodontal ligament 29

PLAP-1 32

Pulp 34

Ankylosis 35

Aetiology of ankylosis 37

Diagnosis of ankylosis 39

Management of ankylosis 41

Experimentally induced ankylosis 44

Immunohistochemistry 47

References 49

Section 2 60

Statement of Purpose 61

Article 1 63

Abstract 63

Introduction 64

2

Aims 66

Materials & Methods 66

Results 68

Discussion 75

Conclusion 78

Acknowledgements. 78

References 79

Article 2 82

Abstract 82

Introduction 83

Aims 85

Materials & Methods 85

Results 86

Discussion 90

Conclusion 92

Acknowledgements 92

References 93

Concluding remarks 94

Appendices 95

3

Figures & Tables

Literature Review

Figure 1: Classification of tooth attachment 12

Figure 2: Dental structure of a rodent 16

Figure 3: Arrangement of periodontal fibre groups 28

Figure 4: Chemical structure of PLAP-1/asporin 33

Figure 5: 3D structure of PLAP-1/asporin 33

Figure 6: Bitewing radiographs of ankylosis 36

Figure 7: Combination between a pair of ankylosed teeth 38

Figure 8: Periotest 39

Figure 9: Periapical radiograph of an ankylosed 21 40

Paper 1

Figure 1: Grid system used to standardise the region of analysis between sections 68

Figure 2a: Rat articular cartilage from femur used as positive control 69

Figure 2b: Negative control of the rat articular cartilage 69

Figure 2c: The experimental tissue stained with PLAP-1 69

Figure 2d: The experimental tissue used for negative control 69

Figure 3a: PLAP-1 staining of chondrocytes 70

Figure 3b: PLAP-1 staining of blood vessels 70

Figure 3c: PLAP-1 staining of periodontal ligament 70

Figure 3d: PLAP-1 staining of periodontal ligament regions 70

Figure 3e: PLAP-1 staining of gingival epithelial tissues 70

Figure 4a: Experimental section with ankylosis 71

Figure 4b: Experimental section without ankylosis 71

Figure 5a: PLAP-1 staining at cementum third of PDL of experimental side 72

Figure 5b: PLAP-1 staining at cementum third of PDL of control side 72

Figure 6a: PLAP-1 staining near root apical region 72

Figure 6b: PLAP-1 staining near cementum third of PDL with ankylosis 72

4

Table 1: Statistical data between control and traumatised sides with ankylosis at various intensities 73

Table 2: Comparison in PLAP-1 intensity between control and traumatised sides with no ankylosis 74

Paper 2

Figure 1a: PLAP-1 staining of the pulp chamber on the control side 87

Figure 1b: PLAP-1 staining of the pulp chamber on the control side (20x magnification) 87

Figure 1c: Negative control of pulp staining displaying the lack of PLAP-1 staining (10 x magnification) 87

Figure 1d: Negative control of pulp staining displaying the lack of PLAP-1 staining (20 x magnification) 87

Figure 2a: PLAP-1 staining of pulp on control side 88

Figure 2b: PLAP-1 staining of pulp on experimental side without ankylosis 88

Table 1: Comparison of PLAP-1 staining intensities within the pulp adjacent to the dentine in sections with ankylosis 89

Table 2: Comparison of PLAP-1 staining intensities within the central pulpal section in sections with ankylosis 89

Figure 3a: PLAP-1 staining of central pulpal region in experimental side (10x magnification) 89

Figure 3b: PLAP-1 staining of central pulpal region in experimental side (20 x magnification) 89

Figure 4: Tertiary dentine and cellular inclusions on experimental side 90

5

Glossary of Abbreviated Items General

ABC avidin-biotin complex

BMP bone morphogenetic protein

B-SA biotin-streptavidin

Cbfa-1 Core binding factor a1

EDTA ethylenediaminetetraacetic acid

EGF epithelial growth factor

FGF fibroblast growth factor

HEBP 1-hydroxyethylidene-1, 1-bisphosphonate

IGF insulin like growth factor

IL interleukin

IMVS Institute of Medical & Veterinary Science

KV kilovolts

LRR leucine rich repeats

LTB4 leukotriene B4

mRNA messenger ribonucleic acid

PAP peroxidase anti-peroxidase

PBS phosphate buffered saline

PDGF platelet derived growth factor

PDL periodontal ligament

PG prostaglandin

PLAP-1 periodontal ligament associated protein-1

PTH parathyroid hormone

PTHrP parathyroid hormone related protein

Runx2 Runt-related transcription factor-2

TGF transforming growth factor

TNF tumour necrosis factor

TNFR tumour necrosis factor receptor

6

Measure of Length

mm millimetre

µm micrometre

Measure of Volume

ml millilitre

Measure of Weight

mg milligram

g gram

kg kilogram

mw molecular weight

7

SIGNED STATEMENT

8

ACKNOWLEDGEMENTS I wish to express my sincere appreciation to the following people for their support in

the completion of this thesis:

Professor W. J. Sampson, P.R. Begg Chair in Orthodontics, The University of

Adelaide, for his readily available expert advice and guidance. His dedication and

enthusiasm for dental research is truly an inspiration.

Dr C. W. Dreyer, Associate Professor in Orthodontics, The University of Adelaide,

for his readily available expert opinion and support.

Dr Kencana Dharmapatni, Hanson Institute, Adelaide, for her expert advice regarding

immunohistochemistry and draft preparation. Her generosity and enthusiasm was

simply astounding.

Tom Sullivan, Division of Population Oral Health, The University of Adelaide, for his

expert statistical help.

Ms Marjorie Quinne & Sandie Hughes, for their assistance with sectioning tissues.

Thankyou.

Last but not least, I must thank my beautiful wife Imelda, my little princess Bella and

my little prince Sebastian. Your tolerance, support and love are simply amazing and

it has made all the hard work worthwhile.

9

SUMMARY The cells of the periodontal ligament have been shown to be osteogenic but under

normal conditions, the PDL space remains patent without the occurrence of ankylosis.

Periodontal Ligament Associated Protein-1 (PLAP-1)/Asporin is a recently

discovered protein that has been suggested to play a significant role in suppressing the

osteogenic tendency of the periodontal ligament and maintaining the fibrous

ligamentous nature of the periodontal ligament. Furthermore, PLAP-1/Asporin has

also been shown to be associated with the differentiation and mineralisation of dental

pulp stem cells.

In this study, the expression of PLAP-1 was investigated using a reversible ankylosis

model induced by hypothermal insult. In paper 1, the principal aim was to determine

the normal distribution of PLAP-1 reactivity in a normal rat maxilla and to analyse the

pattern of PLAP-1 reactivity in association with the formation of ankylosis. In

addition, another study (paper 2) was performed with the aim of investigating the

distribution pattern of PLAP-1 within a normal rat molar pulp as well as its changes

following freezing trauma.

The results from the first paper showed that PLAP-1 was expressed in the PDL, dental

pulp, blood vessel walls and the nasal cartilage. Not all sections obtained ankylosis.

Sections which did not obtain ankylosis demonstrated no significant PLAP-1

expression differences between control and experimental sides. Sections that did

obtain ankylosis yielded a tendency towards increased PLAP-1 reactivity especially

near the cementum. However, it was difficult to deduce whether the relationship of

PLAP-1 to the ankylotic union was associated with bone formation or resorptive

activities.

The results from paper two showed that PLAP-1/Asporin was expressed exclusively

within the pulp under normal conditions and appeared to be associated with the

odontoblastic and cell rich zone. Following trauma, PLAP-1/Asporin expression

10

decreased marginally (not statistically significant) alongside the dentine but increased

significantly in the central pulpal region along with disruption and breakdown of the

cellular structures.

From the results derived, it can be concluded that PLAP-1/Asporin is indeed

expressed in several tissue/cell types and regions including the dental pulp and is not

exclusively associated with the periodontal ligament. In addition, PLAP-1 appears to

have a direct association with ankylosis although it is uncertain whether PLAP-1 aids

in bone mineralisation or resorption. The second null hypothesis was also rejected

although the change in expression of PLAP-1 within the pulp is more morphological

than physiological. Results from the study also suggest that PLAP-1/Asporin does

not appear to play a direct role in the formation of the tertiary dentine.

Further research is required to elucidate the true role of PLAP-1 within the

periodontal ligament as well as the pulp. Additional investigations are also required

to gain further insight into the maintenance of the periodontal ligament.

11

SECTION 1

12

LITERATURE REVIEW

Comparative Dental Anatomy (tooth attachments) In the animal kingdom, there are many different types of periodontal attachment of

teeth. They have been classified according to the area of attachment (eg. crestal,

marginal or socketed) and the mode of attachment (i.e. ankylosis, fibrous or

combined).

Classification of tooth attachment

The classification of tooth attachment has undergone some debate (for full review, see

Gaengler & Metzler1). However, for the purpose of this literature review, only the

classical classification of acrodonty, pleurodonty and thecodonty shall be used (Figure

1).

Figure 1

Classification of tooth attachment

a) Pleurodont = when the tooth is joined to the inner margin of the jaw bone

b) Acrodont = when the tooth is attached at the crest of the jaw bone

c) Thecodont = where the root of the tooth lies in a socket within the jaw bone

(from Peyer2)

Acrodonty is the term used when the attachment from teeth to bone is at the crest of

the jaws.

A NOTE:

This figure/table/image has been removed to comply with copyright regulations. It is included in the print copy of the thesis held by the University of Adelaide Library.

13

In pleurodonty, the tooth is joined to the inner margin of the jaw bone. The teeth can

either be homodont (all teeth similar in shape), haplodont (molar crowns without

ridges or tubercles), or polyphyodont (having many sets of teeth throughout life).

The third main class of thecodonty is where the root of the tooth lies in a socket

within the jaw proper. All placental mammalian teeth, including humans, are

socketed. In thecodonty, a fibrous ligament is the preferred mode of attachment.

Modes of Attachment Despite the various positional relationship of tooth to bone, a common feature is that

the linkages from teeth to jaw are all collagenous with the difference being the degree

of mineralisation. There are three main types of attachment:

1) Ankylosis

2) Fibrous attachment

3) Socketed attachment

In ankylosis, there is a direct mineralised union between the tooth and the supporting

jaw. This form of attachment is found in many bony fishes and nearly all living

reptiles. Ankylosis can also be subclassified to acrodont or pleurodont ankylosis.

There is also protothecodont ankylosis where the tooth is fused to the jaw at the base

of a groove formed by the labial and lingual flange. In naturally occurring ankylosis,

the tooth is not directly fused to the jaw bone but rather via a more spongy structure

which is then connected to the jaw proper. Tomes3 in 1904 named the structure the

‘bone of attachment’.

In animals with fibrous attachment of their teeth, the collagen is only partially

mineralised hence giving some degree of movement to each individual tooth. Fibrous

attachment can be either direct or indirect. In direct attachment, the collagen fibres

run directly from the base of tooth to the jaw bone. In indirect attachment, fibres from

the tooth form a mineralised structure called the pedicel, which is in turn ankylosed to

the jaw.

14

All thecodont dentitions have a fibrous periodontal ligament as their support

mechanism. Amongst living creatures, this type of attachment is only found in

mammals and crocodiles4. There are several distinctive features in mammals

including developing multiple roots in molars, constricted root apices and the

formation de novo of sockets for succedaneous teeth5.

Dentition of Fishes In fishes, there is an almost infinite diversity of dentition and their supporting

apparatus. In early fishes, there were no true teeth but rather dermal denticles that

were part of their body armour. Many of the teeth were attached via ankylosis.

However, through evolution, fibrous attachment is now commonly found (e.g.

elasmobranches such as sharks). All sharks and rays have a specialised form of

fibrous attachment known as the hinge mechanism. Despite fibrous attachment being

evolutionarily advanced compared to ankylosis, many contemporary fishes still have

ankylosis as their primary mode of attachment of teeth (e.g. mackerel)3.

Dentition of Amphibians For amphibians, a common arrangement is a double row of teeth arranged in

concentric lines in the maxilla, between which a single row of teeth upon the lower

jaw passes when the mouth is closed. The outer of the two rows of teeth in the upper

jaw is situated on the premaxilla and maxilla and extends backwards3.

There are, of course, variations to this generalisation. For example, the frog has no

teeth in the lower jaw. When the edentulous mandible is in the closed position, it

passes to the lingual side of the maxillary teeth and fits snugly against the palatal side

of the teeth – especially as frogs have no lips and the teeth have rounded surfaces3.

Dentition of Reptiles In the reptiles, the dominant form of tooth attachment is ankylosis and the relationship

of teeth to jaw can either be acrodont, pleurodont or protothecodont5.

15

One notable exception is the order crocodilian. The teeth are not attached to the jaw

via ankylosis and it is the only example of thecodonty in living reptiles5. However, it

does differ from mammalian gomphosis in a few key areas – firstly, the sockets are

persistent and secondly, the roots are cylindrical with wide apices. There are also

microscopic differences which are beyond the scope of this project.

Dentition of Mammals All mammals have tapered roots with constricted apices as all or part of their dentition.

The posterior teeth are also multi-rooted. These are unique mammalian features. The

periodontal attachment has also undergone significant evolution. In particular, the

mammalian gomphosis allows movement of teeth relative to the alveolar base without

accompanying weakening of the support provided. The ability of the periodontal

ligament and alveolar bone to remodel without weakening tooth support helps

maintain occlusion and allows for the limited tooth replacement. In addition,

compared to other animals, mammals have specialised epithelial attachment around

the tooth (i.e. junctional epithelium) that provides a better resistance to bacterial insult.

Overall, mammal teeth have four classes: incisors, canines, premolars and molars.

Most mammals are diphyodont with teeth restricted to 2 rows – one in the maxilla and

one in the mandible2.

Rodents are commonly used in animal studies. They are characterised as

simplicidentales because they have only one pair of incisors2. These large incisors are

designed for gnawing and are fascinating due to their ability of continual growth

throughout life. Generally, enamel is only evident on the labial side but occasionally

it is also present on the mesial and distal side. Usually the canine and all premolars

are missing with the exception of the upper 3rd premolar which can be present in some

cases. There are usually three molars in each quadrant2. Figure 2 presents a lateral

view of the general dental structures of a rodent.

16

Figure 2

Dental structure of a rodent showing one incisor and three molars in each jaw.

(from Hillson6)

The Periodontium The Periodontium can be defined as those tissues that surround and support the teeth.

They include cementum, periodontal ligament, alveolar bone and that part of the

gingiva that faces the tooth7.

Cementum Cementum is a mineralised, avascular connective tissue that covers the surface of the

root. Cementum is approximately 45-50% hydroxyapatite and 50% collagen and non-

collagenous matrix proteins. Type I collagen comprises 90% of the organic

component in cementum. Other constituents include type III collagen, type XII

collagen and possibly type V and XIV collagen. Non-collagenous proteins include:

alkaline phosphatase, bone sialoprotein, fibronectin, osteocalcin, osteonectin,

osteopontin, proteoglycans, proteolipids, vitronectin and several growth factors8.

Cementum can be subdivided into four types: acellular, cellular, mixed and acellular

afibrillar. Acellular cementum is the most extensively encountered type. It is formed

A NOTE:

This figure/table/image has been removed to comply with copyright regulations. It is included in the print copy of the thesis held by the University of Adelaide Library.

17

immediately as root formation commences (under influence of the Hertwig’s

epithelial root sheath) and typically covers the cervical two-thirds of the root dentine.

Once the tooth is in occlusion, the rate of deposition of cementum quickens leading to

some cementoblasts being trapped in the mineralising tissue7. This is termed cellular

cementum. Mixed cementum is composed of alternating layers of acellular and

cellular cementum and is most commonly found in the apical portion and furcation

areas. Less is known about the acellular afibrillar cementum which lacks collagen

and appears to play no role in supporting the teeth. This cementum is found near the

cemento-enamel junction8.

Alveolar Bone The alveolar bone forms the sockets in which the teeth sit and constitutes part of the

periodontium. The bone that lines the alveolus to which Sharpey’s fibres insert is

termed the bundle bone. Bone is similar to cementum in that it is a mineralised

connective tissue. However, its mineral content is much higher. Bone comprises

roughly 28% type I collagen, 5% noncollagenous matrix proteins and 67%

hydroxyapatite8.

In an adult, alveolar bone has a dense outer layer termed the compact bone with a

central medullary cavity. The bone marrow, which sits in the medullary cavity, is

interspersed by a network of bone trabeculae – the spongy bone9. Bone can be

formed in three ways: endochondral ossification, intramembranous ossification and

sutural ossification.

Endochondral Ossification

Endochondral ossification is the process by which bone formation occurs with the aid

of a cartilage precursor. It occurs in weight bearing bones as the cartilage allows bone

formation while also offering some function. Initially, the mesenchymal stem cells

migrate and condense forming the outline of the eventual bone to be formed at that

site. These mesenchymal cells differentiate into chondroblasts which produce an

extracellular matrix predominantly of type II and X collagen as well as chondroitin

sulphate proteoglycans10. A periosteal bony collar is then formed in long bones. The

chondrocytes then undergo progressive hypertrophy with mineralisation of the

intercolumnar cartilage matrix11. Blood vessels then invade the cartilage bringing

18

with them osteoblast precursors. At the same time, cartilage begins to be resorbed by

multinucleated cells. Due to the actions of these newly derived osteoblasts,

ossification centres begin to develop, both in the epiphyseal cartilage and plate. Once

bone stops growing in length, the epiphyseal plate disappears – first the lower then the

upper. The marrow cavity becomes continuous and the blood vessels throughout the

diaphysis, metaphyses and epiphyses intercommunicate10.

Intramembranous Ossification

Intramembranous ossification is the process by which bone develops directly on to a

soft connective tissue membrane without the need for a cartilage anlage. Initially, at a

pre-determined site, the mesenchyme cells (or ectomesenchyme, if in the craniofacial

region) proliferate and undergo condensation11. The local vascularity also becomes

increased and the osteoblasts differentiate to produce bone matrix de novo. Once

begun, intramembranous bone formation proceeds at an extremely fast pace. This

rapid formation does not allow for the complete remodelling of the resident

extracellular matrix, resulting in a bone matrix comprised of new bone intermingled

with old matrix. This highly cross-hatched and irregular bone is termed the coarse-

fibred woven bone. The various centres of ossification begin from being tiny bony

spicules proceeding to thin bony plates and ultimately fuse with each other to form a

single bone. Over time, the immature bone undergoes a slow transition that will

eventually turn into mature lamellar bone8. Alveolar bone is formed through

intramembranous ossification.

Sutural Bone Growth

Sutural bone growth describes the bone formation that occurs between sutures. In the

beginning of this process, the outer layer splits thus exposing the two layers of

periosteum (the outer and fibrous layer and the osteogenic layer). The fibrous layer

joins with its opposing corresponding layer while the osteogenic layer runs down

through the suture along with its opposing counterpart8. The combined osteogenic

layer of the suture is termed cambium. When the brain grows, successive waves of

new bone cells differentiate from cambium allowing for the deposition of new bone

matrix8. The two cambial layers are separated so that independent growth can occur.

Bone Metabolism

19

Although outwardly inert, bone is a dynamic tissue involved in high metabolic

activities. Bone is constantly under repair with 20% of the cancellous bone surface

undergoing remodelling at any one time12. In a growing child, the end result of bone

metabolism is the growth of the skeleton (bone formation exceeds bone resorption)

whereas in the elderly, bone resorption often exceeds bone formation. In a healthy

adult, however, there exists a balance between bone resorption and bone formation.

The main purpose of remodelling is to help repair microdamages in the bone matrix

and prevent the accumulation of old bones as well as act as a reservoir for minerals

which aid in mineral homeostasis13.

Bone metabolism is a complex process which is tightly regulated. There are many

biochemical compounds involved and they can be grouped into either systemic

regulators or local regulators.

Systemic Regulators Parathyroid Hormone

Parathyroid hormone (PTH) is secreted by the parathyroid gland and is vital in the

regulation of calcium homeostasis. PTH has direct actions on both bone and kidney

while it effects its actions on the intestinal tract indirectly. The actions of PTH are

mediated via a G-protein coupled receptor system in the cells of target tissues14. PTH

is known to have both anabolic and catabolic actions on the skeleton.

PTH has been shown to act directly on osteoblasts via the PTH receptor (PTH and

PTHrP interaction) with stromal cells requiring physical contact with haematopoietic

precursors in order to effect PTH’s catabolic activity15. This interaction results in the

subsequent release of factors (which shall be discussed below) that stimulate

osteoclast activity/function and numbers. PTH also enhances collagenase synthesis,

decreases type I collagen synthesis and decreases alkaline phosphatase in osteoblasts14.

However, the precise mechanism by which PTH stimulates the osteoclast in order to

activate bone resorption remains unclear.

PTH also has anabolic actions which are mediated directly through osteoblasts

themselves. They are able to increase osteoblast numbers possibly through the

20

regulation of cell attachment via the regulation of E-cadherins16. PTH also stimulates

the differentiation of osteoblasts as well as being shown to be mitogenic for bone cells

in vivo17.

Overall, PTH is recognised more for its catabolic actions and its role in calcium

homeostasis as not only does it promote resorption but it also increases the kidney

reabsorption of calcium and stimulates the changes of vitamin D to its active form.

Parathyroid hormone-related peptide

Parathyroid hormone-related peptide (PTHrP) is a known regulator for osteogenic cell

differentiation and/or function. Similar to PTH, PTHrP also regulates both bone

formation and bone resorption. Their effects are effected via binding to cells of

osteoblast phenotype14.

Vitamin D

Vitamin D in its active form (1α,25 dihydroxyvitamin D3) plays a major role in bone

metabolism and mineral homeostasis. Vitamin D has been shown to affect both the

osteoblasts as well as the osteoclasts.

Vitamin D has a biphasic effect on osteoblasts in that it either stimulates or inhibits

the normal developmental pathway or gene expression profiles depending on its

presence during either the differentiation or proliferation stage18. Vitamin D at the

proliferation stage inhibits osteoblast differentiation with decreased proliferation,

decreased collagen synthesis and alkaline phosphatase activity. If the osteoclasts are

already differentiated, then Vitamin D up-regulates osteoblast associated genes such

as osteopontin or osteocalcin hence increasing mineralisation13,18. Therefore, vitamin

D exerts a dual effect on bone remodelling. Furthermore, Vitamin D can affect the

activity of local factors involved in bone metabolism but the results vary depending

on the stage of cell differentiation or experimental design18.

Calcitonin

Calcitonin is a hormone that can oppose the catabolic effects of parathyroid hormone.

It leads to the loss of ruffled border of the osteoclast as well as minimising the

secretion of proteolytic enzymes thereby reducing bone resorption. Furthermore, it

21

increases the loss of calcium from the kidney leading to a drop in the serum calcium

level13.

Glucocorticoids

Glucocorticoids can exert both stimulatory and inhibitory effects on bone cells. These

hormones promote the differentiation of osteoblasts from mesenchymal stem cells

although they also decrease osteoblastic activity13. This was backed up by Weinstein

et al.19 who showed exogenous glucocorticoid to reduce both osteoblastogenesis and

osteoclastogenesis on ex vivo bone marrow cultures. This leads to reduced bone

turnover and bone formation with the clinical manifestation of decreased bone density

and cancellous bone area.

Thyroid Hormone

Similar to the other hormones, the thyroid hormones can also have both catabolic and

anabolic effects on the skeleton. Most of its biological effects, however, result in

increased bone turnover (through its effects on osteoblasts). For example, in

hyperthyroidism, there is pathologically increased bone resorption20.

Sex Hormones

There are 3 main sex hormones involved in bone remodelling – oestrogen,

progesterone and androgens.

Oestrogen

Oestrogen is the main female sex hormone and it has both direct and indirect

effects on bone metabolism. It has been proposed to have a direct inhibiting

effect on the osteoclast as oestrogen receptors have been found on

osteoclasts21. It also stimulates the secretion and activation of TGF-β by

osteoclasts22 and indeed osteoblasts which inhibit osteoclastic activity through

either their autocrine or paracrine actions23. Oestrogen has also been proposed

to promote apoptosis of osteoclasts; therefore, if oestrogen levels decrease,

such as in post-menopausal women, the osteoclasts do not undergo

programmed cell death and hence they live longer ultimately leading to

increased bone resorption. The indirect effects of oestrogen on bone

metabolism are due to its postulated role in the regulation of

22

osteoblast/marrow mononuclear cell production of cytokines which

themselves have a proven role in bone turnover (e.g. Il-1, Il-6, TNF-α)24. The

role oestrogen plays in terms of bone formation is uncertain. Even though

oestrogen has been shown to stimulate type I collagen synthesis in both

fibroblasts and osteoblasts, clinical results have been inconclusive23. Similarly,

oestrogen effects on the proliferation and differentiation of osteoblasts still

remains unclear. The increased levels of oestrogen at puberty are associated

with an overall increase in bone mass23.

Progesterone

Progesterone is another steroid hormone that has been implicated in bone

metabolism. Studies have shown progesterone to increase proliferation and

differentiation of human osteoblast cells and also the increase of insulin

growth factor-II production by these cells25. However, similar to oestrogen,

clinical results of progesterone trials are variable and its role in bone

metabolism is still unclear.

Androgens

Androgens can act directly on osteoclasts leading to a decrease in bone

resorption. This is clearly demonstrated when after orchiectomy, the subject

experiences increased bone resorption and rapid bone loss. Androgens are

also able to regulate the bone-resorbing factors secreted by osteoclasts. For

example, testosterone can decrease PG-E2 production in calvarial organ

cultures exposed to IL-1. The effect of androgens on osteoblast proliferation

and differentiation is still unclear23. Similar to oestrogen, androgens have a

major effect on bone metabolism during puberty particularly, in males.

In addition to their effects on bone cells, sex steroids can affect extraskeletal calcium

homeostasis by their effects on intestinal calcium absorption (increased reabsorption)

renal calcium handling (e.g. Reifenstein & Albright26 showed both oestrogen and

testosterone to decrease the urinary and fecal calcium levels (as well as phosphorous

excretions); also, their effects on parathyroid hormone levels are noted (e.g. oestrogen

may directly regulate PTH secretion).

23

Local Regulators The local factors that contribute and play a role in bone metabolism or bone

remodelling consist of growth factors, cytokines and arachidonic acid metabolites

such as prostaglandins or leukotrienes.

The main growth factors which have been implicated in bone metabolism are insulin-

like growth factor, transforming growth factor-beta family, fibroblast growth factors,

and platelet-derived growth factors.

Insulin-like Growth Factor

Insulin-like growth factor (IGF) has the ability to both induce bone resorption as well

as promote bone growth. The catabolic actions were demonstrated when IGF induced

the formation of osteoclasts in bone marrow cultures27 whereas their anabolic abilities

were proven when IGF promoted the proliferation of osteoblasts in vitro as well as the

increased synthesis of bone matrix and expression of the collagen type I gene28.

Transforming Growth Factor-Beta

Similar to IGF, transforming growth factor-beta (TGF-β) has been implicated in both

the formation and dissolution of bone. TGF has been shown to have variable effects

on bone resorption in that the effects appear to be dose and experiment dependent.

Chenu et al.29 showed TGF inhibited the formation of osteoclast-like cells in vitro by

preferentially differentiating to granulocytes rather than osteoclasts. However, if the

concentration of TGF was low (10-100pg/ml), then the formation of osteoclast-like

cells increased30. The evidence of TGF’s effect on bone formation is more persuasive

with reports of exogenous TGF (local and systemic) inducing bone formation. The

mechanism by which TGF promotes bone formation is attributed to its chemotactic

ability for osteoblasts as well as proliferation31, although the proliferation effect is

disputed by some32.

Bone Morphogenetic Proteins

The bone morphogenetic proteins (BMPs) are part of the transforming growth factor-

beta superfamily. BMPs have been shown to be essential in the bone formation

process33. They have the ability to promote the osteoblast phenotype with increased

type I collagen synthesis, increased alkaline phosphatase activity and also increased

24

osteocalcin expression. BMPs are also able to guide the differentiation path of

mesenchymal stem cells into osteoblasts34. Furthermore, BMP-2 has also been proven

to be chemotactic for osteoblasts35.

Fibroblast Growth Factors

Fibroblast growth factors have been shown to have a dose dependent effect on

collagen synthesis (high concentration decreases collagen synthesis while low

concentration increases collagen production). Both FGF-1 and FGF-2 have been

shown to stimulate bone resorption in marrow cultures31. However, contradictory

findings by Canalis et al.36 demonstrated that FGF-2 is capable of inducing the

proliferation of osteoblasts and stromal cells in vitro which was backed up by in vivo

studies37,38 which also showed increased numbers of osteoblasts as well as new bone

formation after systemic administration.

Platelet Derived Growth Factors

The platelet derived growth factors are chemotactic for osteoblasts as well as having a

positive effect on their proliferation. However, PDGF does have a time dependent

effect on bone formation. Continuous exposure to PDGF leads to decreased

mineralisation via the inhibition of osteoblast function whereas intermittent exposure

of PDGF to the osteoblast precursors appears to increase mineralisation; possibly

through the increase in osteoblast numbers albeit without inhibition of their

functions39. PDGF also has an effect on bone resorption because PDGF-AB and

PDGF-BB have been shown to promote bone resorption31.

Numerous growth factors have been added to the growing list of local regulators of

bone turnover; these include, epidermal growth factor, transforming growth factor –

alpha and vascular endothelial growth factor.

Prostaglandin/Leukotrienes

Both prostaglandins and leukotrienes have been proven to affect the balance between

bone formation and resorption. Prostaglandins can stimulate both bone resorption and

formation with most of their effects largely mediated via the protein kinase A40. The

exact mechanism is likely to be linked to prostaglandin’s ability to regulate the

replication and differentiation of precursor cells. For example, prostaglandin

25

increases osteoblast precursors as well as having a mitogenic effect on them40. In

comparison to prostaglandins, leukotrienes have minimal data in relation to the role

they play in bone metabolism. However, it has been shown that LTB4 can stimulate

bone resorption while it is postulated that other leukotrienes may have similar

catabolic effects on the skeleton but are less potent40.

Cytokines

Many cytokines are involved in the process of bone metabolism but the most

extensively studied and those with possibly the most important roles are interleukin-1,

interleukin-6 and tumour necrosis factor.

Interleukin-1 (IL-1) is produced in bone with resident macrophages being the most

likely source although osteoblasts and osteoclasts may also produce this cytokine40.

IL-1 is an extremely potent stimulator of bone resorption in vitro and also has potent

in vivo effects41. Furthermore, it has also been shown to exhibit an ability to inhibit

bone formation in vitro42.

Tumour necrosis factor- alpha and beta both promote bone resorption. They have also

been shown to inhibit the synthesis of collagen in bone43. In addition, it has been

suggested that TNF is able to modulate the effects of oestrogen on bone although this

hypothesis could not be substantiated in follow-up studies40.

Similar to interleukin-1, interleukin-6 (IL-6) stimulates bone resorption although its

effect is less extensive. The major mechanism by which IL-6 contributes to bone

resorption is via the regulation of osteoclast precursor cell differentiation. In fact, the

ability of IL-1 and TNF to stimulate osteoclast-like cell development in marrow

cultures has been suggested to be due to the production of IL-640.

Bone metabolism or even bone remodelling is a vast and complicated topic on which

volumes of textbooks have been written. The above summary is a brief account of the

major players currently identified in bone metabolism.

26

Bone and Its Cellular Constituents

Bone is a dynamic tissue which is constantly remodelled. In fact, in a growing child

the entire skeleton will be “brand new” in 12 months. The maintenance of bone is

through the functions of osteoblasts (bone forming cells), osteocytes and osteoclasts

(bone resorbing cells).

Osteoblasts Osteoblasts are the cells responsible for the generation of new bone. They are derived

from mesenchymal stem cells and can range from 15-80µm7. When active,

osteoblasts are plump and cuboidal in shape. When viewed under the light microscope,

they exhibit extensive and well-developed protein synthesis organelles. The

osteoblasts form a layer on the surface of the bone and it has been postulated that they

function in controlling the influx of ions8 although this theory is unproven.

Furthermore, whether the bone-lining cells are actually osteoblasts is debatable too10.

Osteoblasts secrete the organic matrix which then becomes mineralised. As they lay

down the matrix, some osteoblasts become trapped and these cells become known as

osteocytes.

Osteocytes

The osteocytes are mononuclear cells that occupy the osteocytic lacuna within the

bone matrix. They have many fine processes extending through fine canals

(canaliculi) to maintain contact with osteoblasts, bone-lining cells and adjacent

osteocytes. Their primary functions are to sense the conditions of the

microenvironment and maintain the architecture of bone as well as collection of

nutrients8.

Osteoclasts

Osteoclasts are multinucleated giant cells. They generally encase 10-20 and

sometimes up to 100 nuclei per osteoclast44. Hence they are also much larger than

osteocytes or osteoblasts although there are variations within species. For example,

the rodent osteoclasts are generally smaller than human osteoclasts44. Osteoclasts are

derived from mononuclear precursor cells that are of haematopoietic origin10,44-49.

Their primary function is to resorb bone and mineralised cartilage.

27

Periodontal Ligament The periodontal ligament is a specialised connective tissue that joins the cementum

covering the root of the tooth to the bundle bone of the alveolar process via Sharpey’s

fibres. The ligament width ranges from 0.15 to 0.38mm8. Broadly, there are two

main components to the periodontal ligament: cellular and non-cellular constituents.

Cellular Components of the PDL

The periodontal ligament contains numerous types of cells including:

Fibroblasts

Fibroblasts are the principal cell of the periodontal ligament region and

comprise approximately 20% of the cellular component in sheep and

up to 55% in rodents50. They are large cells with an extensive

organelle network for the synthesis of protein. They are aligned along

the fibre bundles and have extensive processes that wrap around the

bundles.

Macrophages

The periodontal ligament contains some resident macrophages to help

protect the tissues from invading antigens.

Epithelial cell rests of Malassez

These epithelial cells are the remnants of Hertwig’s epithelial root

sheath. They form clusters of epithelial cells which form a network

within the periodontal ligament and surround the root of the teeth

although more commonly in the apical region8. Whether or not the

cell rests of Malassez have a function, or are mere remnants, is still

unclear. However, some studies have implicated the cell rests as

having a pivotal role in the maintenance of the periodontal ligament

space51-53.

Undifferentiated mesenchymal cells

These small spherical cells tend to be associated with the blood vessels

within the periodontal ligament and are thought to be the precursor

cells for osteoblasts, cementoblasts and fibroblasts7.

Cementoblasts

Cementoblasts are the cells responsible for laying down the cementum

and they are phenotypically similar to osteoblasts8. Indeed, they are so

28

similar that it is debatable whether the cementoblasts are a completely

different cell type or are positional osteoblasts (for recent review, see

Bosshardt, 2005)54.

Osteoblasts and Osteoclasts

Technically speaking, both the osteoblasts and osteoclasts are located

within the periodontal ligament. However, as they are functionally

associated with bone, they will be discussed along with the mineralised

bone tissue.

Non-Cellular Component of the PDL

The non-cellular component of the periodontal ligament contains an extracellular

compartment of collagenous fibres (with some oxytalan fibres) and a noncollagenous

extracellular matrix. The extracellular collagen fibres are arranged in bundles and can

be organised into groups as shown in Figure 3. The amorphous background is largely

composed of ground substance (~70% is water) together with glycosaminoglycans,

glycoproteins and glycolipids8.

Figure 3

Arrangement of major bundles of periodontal ligament fibres

(Adapted from Tennant7)

A NOTE:

This figure/table/image has been removed to comply with copyright regulations. It is included in the print copy of the thesis held by the University of Adelaide Library.

29

There is, of course, an extensive network of both blood vessels and nerve

fibres/endings permeating throughout the periodontal ligament.

Maintenance of the PDL

In a healthy periodontium, the periodontal ligament functions to connect the teeth to

the jaws, sustain the masticatory load, provides sensory information and also prevents

the bony union of the roots of the teeth to the alveolar bone. Numerous studies have

demonstrated that the death or removal of periodontal ligament cells leads to the loss

of the periodontal membrane eventually resulting in extensive root resorption and

widespread ankylosis50,51,55. This knowledge combined with the fact that cells of the

periodontal ligament have been shown to be osteogenic in vitro56 leads to the

conclusion that certain factors/cells within the periodontal ligament must normally

suppress the osteogenic tendency and maintain the fibrous ligamentous nature of the

periodontal ligament.

The maintenance of the periodontal membrane is still a debatable topic. The current

opinion is that a plethora of factors and cells may indeed combine to maintain the

patency of the periodontal ligament space. One cell type with such possible function

is the epithelial cell rests of Malassez. According to Spouge57 and Wesselink &

Beertsen58, the rests of Malassez were originally thought to have no physiological

function. However, Ten Cate59 demonstrated that these clusters of epithelial cells

possessed an active metabolic potential. Possibly the most influential argument was

the accidental findings from Loe and Waerhaug60 in 1961 where they were

experimenting with replantation of teeth. They noted that in replanted teeth where the

periodontal ligament re-established itself, the epithelial cell rests of Malassez were

always present. By contrast, in samples where the periodontal ligament did not

reattach, epithelial cells could not be seen. The authors hypothesised that the

epithelial remnants of Malassez could possibly play a role in the maintenance of the

periodontal ligament. In a review, Spouge53 commented that although no special

functions have yet been shown for these epithelial remnants, the mere presence of

these epithelial cells could act as ankylosis inhibitor. This is based on the concept that

epithelium seems to be incompatible with bone as nowhere throughout the body is

30

epithelium in contact with bone. Bone and cementum are, however, very similar and

show a marked tendency to fuse at times. The fact that the roots of the teeth are in

extremely close contact to the bone, and yet ankylosis remains an uncommon finding,

would suggest that something, possibly the epithelial cell rests, play a role in the

maintenance of the periodontal membrane53. Their views are supported by Lindskog

et al.52 who replanted teeth with experimental cavities on root surfaces with and

without placement of enamel organ epithelium. It was discovered that bone was not

found in cavities with epithelium as compared to the control cavities where the bone

demonstrated in-growth. A more recent study by Fujiyama et al.51 demonstrated that

when the nerve supply to the PDL is reduced, the distribution of the cell rests of

Malassez also decreases along with a corresponding decrease in the width of the

periodontal space. Dentoalveolar ankylosis was also a common finding after

denervation. The authors concluded that not only are the cell rests of Malassez likely

to be involved in the maintenance of the periodontal ligament but the sensory

innervation of the ligament may also play an indirect role. This is backed up by

results from Fong et al.61 who, in their heterotopic transplantation model,

demonstrated the likelihood of epithelial cells within the periodontal ligament to

prevent ankylosis.

Aside from the epithelial cell rests of Malassez, it was speculated by Melcher62 that,

in healthy individuals, fibroblasts within the periodontal ligament are able to block

osteogenesis within the periodontium by releasing locally acting regulators such as

cytokines and growth factors. This was later validated with an in vitro experiment

conducted by Melcher and Cheong63 and later supported by Lekic & McCulloch64. It

has since been shown that the prostaglandins (including B2, D2, E2, F2 alpha and I2)

secreted by the fibroblasts are capable of inhibiting the mineralisation of PDL in

vitro65,66. Furthermore, glycosaminoglycans have also been implicated in having a

role in PDL maintenance as Kirkham et al.67 demonstrated that the removal of

glycosaminoglycans via enzyme digestion permitted the formation of mineralised

crystals in sheep PDL. However, the exact mechanism by which fibroblasts and their

products inhibit PDL mineralisation is yet to be fully understood68.

Nevertheless, recent molecular investigations have yielded further clues and identified

several molecular factors that may have a role in PDL maintenance. One such factor

31

is the RGD-CAP. RGD-CAP is the name given to a collagen-associated protein

which contains the RGD (arginine-glycine-aspartic acid) sequence. It is also known

as βig-h3. In 2002, Ohno et al.69 showed that RGD-CAP was present in human PDL

cells. Furthermore, when PDL cells were co-cultured with known osteogenic

stimulants such as dexamethasone or 1α,25-dihydroxyvitamin D3, the expression of

RGD-CAP mRNA gradually declined. In addition, exogenous RGD-CAP was able to

suppress alkaline phosphatase activity and also inhibit bone nodule formation in vitro.

Hence, RGD-CAP was considered to have a role in the maintenance of PDL

homeostasis by regulating mineralisation69.

In a similar experiment, but testing the role of epidermal growth factor (EGF)

receptors, Li et al.70 found that as the cultured PDL cells were stimulated to mineralise,

the expression of EGF receptor also decreased. The authors concluded that epidermal

growth factor receptors have a negative regulatory function on human periodontal

ligament mineralisation. Further details of the experiment and its findings would have

been desirable to further back their conclusion but, unfortunately, detailed information

was not available in an English format.

Another possible candidate for the maintenance of periodontal ligament space is

S100A4. S100A4 is a member of the S100 calcium binding protein family71. It is

expressed by several cell types such as odontoblasts and osteoblasts and also other

regions such as liver and bone marrow72 but the expression is highest within the

periodontal ligament. Duarte et al.71 demonstrated that S100A4 is secreted by PDL

cells and is able to inhibit mineralisation in a rat osteogenic cell culture. The authors

postulated that the calcium binding property of S100A4 allows it to act as an inhibitor

of the formation of hydroxyapatite crystals although later studies suggested that the

anti-osteogenic property of S100A4 is likely to be through the suppression of

osteoblastic genes (such as genes for osteopontin, osteocalcin and transcription factors

like Runx2/Cbfa-1) in the PDL cells73. Later investigations also suggest that the role

of S100A4 in bone physiology is to act as a negative regulator of matrix

mineralization possibly by modulating the process of osteoblast differentiation73.

Recently, the homeobox protein Msx2 has also been nominated as a possible

candidate for PDL maintenance. The expression of Msx2 in periodontal ligament and

32

also tendon cells was higher than in osteoblasts with reduction of the Msx2 protein

expression positively associated with osteoblast differentiation and mineralisation in

vitro74. Conversely, increased Msx2 inhibited osteoblastic differentiation and hence

also mineralisation. It is thought that Msx2 achieves this effect via suppressing the

activity of Runx2/Osf2 with TLE1 (a human homolog of Drosophilia Groucho

protein) as a co-suppressor. As Msx2 has been shown to be downregulated in patients

with ossification of the posterior longitudinal ligament, it was suggested that Msx2

may have a role in the inhibition of mineralisation of all tendon and ligaments

including the periodontal ligament74.

Another protein suggested in having a role in PDL homeostasis is a basic helix loop

helix protein called Twist. It has been shown by Komaki et al.56 that the expression of

Twist within the PDL is intense and located along the alveolar bone surface. It was

shown that in cell cultures where osteoblast-related genes were stimulated to increase,

the expression of Twist decreases correspondingly. Conversely, knock-out of Twist

proteins leads to an increase in osteoblast-related proteins such as osteopontin and

bone sialoprotein. The exact mechanism is unknown although it could be attributed to

the ability of Twist to interact with the DNA-binding domain of Runx-2 thus

inhibiting its function as shown by Bialek et al.75. Hence, Twist may contribute to the

maintenance of the PDL by acting as a negative regulator of osteoblastic

differentiation.

PLAP-1/Asporin

Another newly discovered factor associated with the periodontal ligament is the

periodontal ligament associated protein-1 (PLAP-1). PLAP-1 (also known as asporin)

was simultaneously discovered by separate research groups76-78 in 2001. PLAP-1 is a

member of the leucine-rich repeat proteoglycan family and is similar to biglycan and

decorin (human asporin is 50% identical and 70% similar to decorin and biglycan)76.

However, as PLAP-1 does not contain glycosaminoglycan attachment sites77 and

contains a unique sequence of aspartate residues, it is not considered a true

proteoglycan79. The normal structure of PLAP-1/asporin contains a putative

propeptide, 4 aminoterminal cysteines, 10 leucine rich repeats and 2 C-terminal

cysteins (Figure 4 & 5).

33

Figure 4

The chemical structure of PLAP-1/asporin

(Adapted from Yamada et al.80)

Figure 5

The 3-D structure of PLAP-1/asporin

(green = Glu-194; purple=Arg-170; gray=carbon; blue=nitrogen; white=hydrogen;

red=oxygen; yellow ribbon=β-sheet; red ribbon=secondary helix structure)

(Adapted from Tomoeda et al.81)

Asporin mRNA can be found in many areas including the uterus, heart, liver and also

the extra cellular matrix in cartilage77,82. However, within the maxilla (of rat), unlike

the other molecular candidates such as S100A4 and Msx-2, PLAP-1 is highly location

specific with the periodontal ligament being the only region to express the protein

within the maxilla83. Furthermore, in situ hybridisation showed that PLAP-1 was

highly expressed in the dental follicle during dental formation indicating a central role

for PLAP-1 in the development of the periodontal tissues. However, conflicting data

34

have been demonstrated by Lee et al84 who found PLAP-1/Asporin at the globular

calcific region in the junction of predentine and denine.

The function of PLAP-1/asporin is still unclear although in their research on

osteoarthritis and asporin, Nakajima and co-workers85 found asporin acts as a negative

regulator of TGF-beta (TGF-beta 1 regulates proliferation, differentiation as well as

the matrix production of chondrocytes and their progenitor cells) in cartilage, thereby

affecting chrondrogenesis and ‘playing a critical role in etiology and pathogenesis of

osteoarthritis’. PLAP-1/asporin has also been found to affect the activities of other

growth factors (eg. TGF-β). In their investigation, Yamada et al.83 showed that

PLAP-1 regulates periodontal ligament cell cytodifferentiation and also mineralisation

through its negative feedback interaction with bone morphogenetic protein-2 (BMP-2).

A follow on study revealed that the mechanism by which PLAP-1 exerts the described

biological effect is through the leucine-rich repeats (LRR) motif. The LRR are a

protein structural motif that is composed of 20-30 amino acid stretches that are

extremely rich in the amino acid leucine. A particular LRR motif that appears to be

highly associated with PLAP-1 is LRR581.

The Pulp The dental pulp is a region of soft connective tissue which lies beneath the dentine

within a tooth. The main cellular constituents are odontoblasts, fibroblasts,

mesenchymal stem cells, macrophages and lymphocytes. The extracellular component

of the pulp is composed of ground substance (eg. glycosaminoglycans) and collagen

fibres (mainly type I & III). Additionally, the pulp contains blood vessels, nerve fibres

as well as lymphatic vessels.8

The primary function is the formation of dentine through the actions of the

odontoblasts. Other functions of the pulp include the provision of nutrients and

moisture as well as neurosensory information such as pain, pressure or temperature

differences. Finally, the dental pulp is also able to provide protection due to the

formation of reparative dentine following a traumatic episode.

35

Ankylosis There are many names or terms used to describe the phenomenon of the fusion of root

of the tooth to the underlying jaw bone. They may include the following terms:

tooth/dentoalveolar ankylosis; infra-occlusion; secondary-retention; reimpaction;

halbretention, reinclusion, replacement resorption and submergence86.

The first sign of a bony union between the root of the tooth and the alveolar bone was

reported in 1922 by Albin Oppenheim87 from Vienna, Austria. However, Dr

Oppenheim’s research at that time was mainly focused on the resorption of deciduous

teeth and the significance of this finding was not realised. In fact, it was more than 10

years later that Frederick Noyes88 first linked the clinical signs of tooth ankylosis to

the histological cause.

As mentioned previously, dentoalveolar ankylosis is the fusion of the cementum or

dentine to the alveolar bone89. It is a condition that occurs mostly in the deciduous

dentition with a prevalence rate of 1.3% to 14.3% of the population in general90,91.

Ankylosis also has a familial tendency92 (ie. occurs more frequently amongst siblings)

as well as a racial component with the condition roughly four times more common in

whites than blacks93. It also occurs mainly (~91.8%) in the deciduous molars region

and is more than twice as frequent in the mandible when compared to the maxilla94.

Early studies on the relationship between ankylosis of primary molars associated with

congenital absence of the succedaneous teeth yielded conflicting results95,96 but later

studies all seemed to support the notion that agenesis of the permanent premolar

predisposes the associated primary molar to be ankylosed97,98.

The fusion of the root to the alveolar bone destroys the eruptive potential of the tooth.

It also prevents normal dentoalveolar development in that region of the jaw. The

effects of ankylosis can vary depending on when the fusion occurs. If the fusion

occurs after adult equilibrium is already achieved, then the consequences are

generally minor. However, due to the fact that vertical facial height increases

throughout life99, the effects of ankylosis will still progress after adolescence, albeit at

a much slower rate. If, however, the ankylosis occurs while the jaw is developing,

then the consequences can be severe. The problems occur when the ankylosed tooth

36

becomes submerged leading to tilting of adjacent teeth, over-eruption of opposing

teeth leading to the loss of space as well as ectopic or failure of eruption of the

succedaneous tooth100 (Figure 6). In addition, as erupting teeth grow bone as they

move, a severely infra-occluded tooth will lead to an arrest of localised jaw

development resulting in a large bony defect which will make restorative options

more difficult.

Figure 6

Sequential bitewing radiograp-hs showing ankylosis of a mandibular first molar as well as the progressive nature of its associated orthodontic prob-lems. Note the loss of space, over-eruption and tipping of neighbouring teeth.

(Radiographs from Kurol101)

Another sequela of ankylosis is the phenomenon of replacement resorption

(sometimes used synonymously with ankylosis). After ankylosis develops, the root of

37

the tooth becomes susceptible to the remodelling activity of the bone. That is, the

root is resorbed and gradually replaced by new bone – hence, the term replacement

resorption.

As replacement resorption progresses, the root gradually becomes thinner and weaker.

When there is severe replacement resorption, the crown of the tooth can fracture.

The rate of replacement resorption varies and is dependent on the skeletal growth rate

of the patient. Therefore, in adults ankylosed teeth have the potential to last for

decades.

The periodontal membrane can be thought of as a double periosteum which covers

both the cementum and the alveolar bone102. Ankylosis is thought to occur when

there is a breakdown locally of the periodontal ligament allowing for in growth of

endosteal progenitor cells from the adjacent bone marrow to repopulate the defect

rather than root-side periodontal ligament progenitor cells103. Under the influence of

local cell signalling mechanisms, the endosteal progenitor cells migrate to the site of

the defect. Despite the ability of these cells to differentiate into various periodontal

ligament cells, the ultimate phenotype they differentiate into depends on the local

regulators64. Unfortunately, it has been shown that after injury, root-side progenitor

cells preferentially differentiate into cells capable of osteogenesis and osteoclasis

thereby favouring ankylosis over periodontal ligament regeneration103.

Aetiology of Ankylosis The aetiology of ankylosis is still unclear although there are three main theories:

mechanical trauma, disturbed local metabolism and genetics. The fact that around

92% of ankylosis was found in the molar region tends to support the theory of

mechanical trauma as the posterior region is by far where the most force is exerted94.

However, there are significant doubts to this theory as most of the ankylosis was

found in the deciduous dentition and only 8% were permanent molars. If occlusal

trauma was the cause then there should be more permanent molars ankylosed as

adults exert a much greater chewing force than a child or adolescent. Furthermore,

there is no evidence linking ankylosis to trapeze artists or other entertainers who

perform weight-lifting tricks with their teeth94. Another possible and arguably more

38

feasible cause of dental ankylosis is a disturbed local metabolism. It is theorised that

the disturbed local metabolism results in a lysis of the periodontal ligament at a

particular point thus exposing both the denuded cementum and bone to one another

leading to the formation of ankylosis. Localised ossification of the periodontal

membrane could also occur leading to the formation of ankylosis94. The third theory

is based on the finding that there is a significantly higher incidence of ankylosis

between siblings90-92,104,105.

Biederman106 found a statistical way to test the three hypotheses. In mouths with two

ankylosed deciduous molars, there are mathematically, six possible combinations.

Figure 7

The various combinations possible between two ankylosed teeth.

(from Biederman106)

1. Cross: upper right-lower left

lower right-upper left

2. Same side upper right-lower right

upper left-lower left

3. Same jaw upper right-upper left

lower right-lower left

If the cause of ankylosis is random (e.g. a defective formation, lysis or ossification,

then the frequency in all three of the categories should roughly match. If occlusal

trauma was the cause then the predominant category should be ‘same side’ as equal

force is produced against both upper and corresponding lower teeth when masticating.

If the cause is disturbed local metabolism, then they are likely to fall into the ‘same

jaw’ category as they are more likely to develop in parallel fashion. The results

overwhelmingly supported the hypothesis that dental ankylosis is formed as a result of

localised metabolic disturbance in the periodontal ligament. However, the result was

A NOTE:

This figure/table/image has been removed to comply with copyright regulations. It is included in the print copy of the thesis held by the University of Adelaide Library.

39

not without question. The authors could not explain the 8 ‘same side’ occurrences

and also they could not explain the site selectivity of dental ankylosis.

In 1963, Dixon107 proposed that it may not be the local disturbance of metabolism but

a disturbance in the interaction between normal root resorption and hard tissue repair

in deciduous molars that leads to the formation of ankylosis.

Diagnosis of Ankylosis The diagnosis of ankylosis has also been an area of uncertainty. Absolute proof of

bony union between the tooth root and bone can be only be confirmed by histological

sections but is obviously clinically not applicable as a diagnostic tool. The literature

so far (backed up by recent report by Crowther et al.108) suggests the use of a

combination of the following indicators when ankylosis is suspected:

• Loss of mobility – Bucco-lingual forces can be used via palpation to detect the

extent of loss of mobility. Some authors suggest normal mobility to be lost

when more than 10% of root surface is ankylosed90 although others have

suggested even a minute microscopic area can induce immobility106. The

mobility can also be tested by electronic instruments such as the Periotest

(Figure 8). However, problems of error readings, unit malfunction and test-re-

test reliability issues have been reported.

Figure 8 Periotest

(from Campbell et al.103)

A NOTE:

This figure/table/image has been removed to comply with copyright regulations. It is included in the print copy of the thesis held by the University of Adelaide Library.

40

• Infra-Occlusion – This is the most definitive clinical indication that ankylosis

has occurred109. A tooth that has ‘submerged’ beneath the occlusal plane

when previously at occlusal height is very likely to be ankylosed.

• Percussion – A high pitched sound can generally be heard as opposed to a dull

sound when more than 20% of the root is ankylosed110.

• Radiograph – Obliteration of the periodontal ligament space can be indicative

of periodontal membrane breakdown and subsequent bony union (Figure 9).

However, it is not an effective diagnostic tool as the ankylotic area could be

hidden by other anatomy and could also be microscopic. In fact, Raghoebar et

al.86 have calculated the efficiency of radiographs in diagnosing dental

ankylosis to be ~21% only, even with an experienced operator. In addition, it

was shown by Andreason111 that ankylosis initially favours the labial and

lingual root surfaces which are basically impossible to diagnose via

conventional radiographs.

• Orthodontic Force – An ankylosed tooth does not respond to orthodontic

forces106,112.

A condition that may mimic dentoalveolar ankylosis early on is primary failure of

eruption. It is a rare condition in which the eruptive mechanism is disrupted leading

to either complete failure of eruption or only partial eruption. Teeth affected by this

Figure 9 Note the lack of periodontal ligament space and replacement resorption on the apical half of the 21 (from Campbell et al.103)

A NOTE:

This figure/table/image has been removed to comply with copyright regulations. It is included in the print copy of the thesis held by the University of Adelaide Library.

41

condition tend to become ankylosed but the failure to erupt is in evidence prior to

ankylosis. The affected teeth also do not respond well to orthodontic traction - at best

only 1 – 2 mm of movement can be achieved108. In fact, application of orthodontic

force often directly leads to the development of ankylosis113.

Also, transient ankylosis has been reported by Andreasen & Skougaard114 and

Hammarstrom et al.115 to disappear within 8 weeks which may explain the

spontaneous re-eruption observed by Raghoebar et al.116 as well as Belanger et al.117.

An uncommon form of ankylosis can be due to the inostosis of enamel. Because the

enamel epithelium protecting the tooth disintegrates (e.g. via infection), enamel can

be resorbed and, subsequently, bone/cementum may be deposited in its place thus

placing a solid fixation on the tooth 118.

Management of Ankylosis The management of dento-alveolar ankylosis can vary a great deal and depends on the

individual circumstances. Currently, the management can be arranged in the

following sub-classifications; deciduous teeth with a permanent successor, deciduous

teeth without a permanent successor and ankylosed permanent teeth.

Deciduous ankylosed tooth with succedaneous tooth

The main aim when dealing with a primary tooth with a permanent successor is to

ensure the normal development and eruption of the permanent tooth. A permanent

successor can resorb the ankylotic area as it erupts into the primary tooth119.

Therefore, if the infra-occlusion is minor, monitoring is indicated. If the tooth does

not spontaneously exfoliate upon the estimated time, an additional allowance of 6

months is acceptable105 before extraction is indicated. Also, if the primary tooth is

significantly below the occlusal plane then immediate removal of the affected tooth is

recommended plus space maintenance90. If there is already tipping of teeth then

orthodontic intervention may be required.

42

Deciduous ankylosed teeth without a permanent successor

Ankylosis is a common finding in deciduous molars without a permanent successor120.

Treatment required depends on the onset of ankylosis with respect to growth of the

individual as well as the date of the diagnosis. As a general rule, the ankylosed tooth

should be carefully monitored until there is a risk of tipping of adjacent teeth or over-

eruption of the opposing tooth. Restorative procedures need to be undertaken if there

is a need to re-establish the mesial and distal contacts as well as the occlusal plane.

Orthodontic movement may be required if there is minor tipping to help facilitate the

restorative procedure if the plan is to retain the tooth for as long as possible. Long

term studies have demonstrated that when left in situ, the primary molars without a

permanent successor can last more than 10-15 years and should be considered an

acceptable semi-permanent solution121,122. Obviously if there is a risk of development

of a large jaw defect (e.g. in early onset cases), then extraction (with options of 1.

orthodontic closure, 2.prosthetic replacement or 3. implant in the future) is likely to be

the best choice. Biederman106 did describe a technique called luxation. It involves

gentle bucco-lingual movement of the ankylosed tooth in an attempt to break the

ankylotic union and hope for re-establishment of the periodontal membrane. This

technique together with immediate application of orthodontic force has proven

successful for Geiger and Bronsky123. However, the outcomes are unpredictable and

there is a real chance for re-ankylosis103. Raghoebar et al.124 have also shown that

most of the ankylotic areas for deciduous molars are in the furcation, which is

unlikely to break by luxation and may quite possibly extend the ankylotic area.

Permanent ankylosed tooth

Similar to deciduous ankylosed teeth without a permanent successor, the treatment

options for an ankylosed permanent tooth can vary greatly. Once again, the choice of

treatment should be decided on after due consideration to the growth of the patient,

the date of diagnosis, the state of infra-occlusion and neighbouring teeth when

diagnosed and the patient’s opinion. If the patient is an adult with minimal future

skeletal growth, then the treatment options consist of monitoring and restoring the

tooth as required. Patients need to be advised that due to the continual vertical facial

growth that exists in all human beings, remake of crowns may be needed. When the

growth of the patient is not yet complete, other procedures may be indicated; luxation,

43

localised ostectomy, decoronation, corticotomy with or without osteogenic distraction

have all been suggested.

Specific Techniques of Management

Luxation (as mentioned earlier) is the earliest described technique that may manage

ankylosis.

Localised ostectomy has also been proposed. It involves identifying the ankylosed

area, raising a periodontal flap and surgically removing the affected mineralised tissue.

This technique only works if the ankylotic union is in the crestal area as ankylosis

elsewhere on the root surface presents access problems100.

Corticotomy is the technique where a whole block of cortical bone and soft tissue is

isolated along with the tooth and is repositioned as desired. This technique does not

correct the ankylosis and it is also limited by the restriction of the mucosa’s ability to

be extended125.

Decoronation is the procedure first described by Malmgrem et al.126 in 1984. It

involves sectioning the crown of the affected tooth (thereby decoronating it) and

leaving the root purposely in situ. The root has been shown to retain the integrity of

the alveolar bone and allow further apposition of bone127-129. This procedure has been

proposed as a surgical technique that allows preservation of the bone volume for the

future and avoids aesthetic disturbances. It presents a viable alternative to aggressive

options in cases where other therapeutic alternatives are not feasible127. However,

decoronation does rely on the phenomenon of replacement resorption to resorb the

remaining root prior to the placement of an implant.

Another surgical technique that has gained interest in recent times is the application of

distraction osteogenesis. It is a variation of osteotomy where a localised osteotomy is

performed and force is delivered to the targeted section of bone in the direction in

which new bone growth is desired. Modification of the classic distraction technique

has also been performed through adjustment of the bone that is sitting on the newly

formed callus which allows for three dimensional movements - thus utilising the

‘floating bone concept’130. The distraction can be done via orthodontic wires, external

44

(most likely tooth-borne) distractors or internal bone-supported screw distractors131,132.

The whole block is moved into the desired position where final adjustments may still

need to take place.

Extraction of the affected tooth is still a viable option. However, the extraction of an

ankylosed tooth requires extra care as it may lead to further trauma resulting in an un-

aesthetic bony ridge defect and influence the chances of delivering an optimal

prosthetic treatment127.

Unfortunately, most of these techniques described appear in single or case reports

with no support from randomised clinical trials and hence cannot be relied upon for

predictable long term successful outcomes90.

Experimentally Produced Ankylosis Efforts have long been made to induce dental ankylosis experimentally in order to

further study the pathogenesis as well as the relationship between ankylosis and its

surrounding biological tissues. In fact, according to Andreason & Skougaard114,

attempts as early as 1928 were being carried out by Feldman to create artificial dental

ankylosis.

As discussed previously, the periodontal ligament and its constituents normally

depress osteogenic actions which ensure the periodontal ligament space is free of

calcified tissue. Sufficient damage to the periodontal ligament limits its anti-

osteogenic ability giving rise to an opportunity for ankylosis to develop. It is,

therefore, not surprising that almost all of the various methods employed to induce

ankylosis in vivo include inducing certain forms of trauma.

There are five main modes of inducing an ankylotic union between the root and

alveolar bone.

• Mechanical/Physical

• Chemical

• Electrical

• Thermal

45

• Heat

• Cold

• Denervation

Mechanical/Physical

Several investigators have attempted to produce ankylosis through inducing

mechanical trauma by using devices such as dental burs to injure the periodontal

ligament tissue as well as the surrounding bone and root114,133,134. Their techniques

yielded inconsistent development of ankylosis possibly due to insufficient trauma.

Rubin et al.134 in 1984 also investigated the possibility to induce ankylosis via

occlusal trauma. This was simulated in the form of a stainless steel crown formed to

purposely be in premature contact. The study did not create any ankylosis which

backs up Biederman’s94 conclusion that masticatory force is unlikely to be a cause of

ankylosis.

In contrast, physical damage in the form of replantation proved to be a reliable

method of inducing ankylosis. This is particularly true if the periodontal ligament

extracellular matrix and part of its cellular component are not kept intact (e.g. increase

extra-oral time). As such, it has been a common methodology in investigating

ankylosis and the healing process of the PDL as a whole55,60,110,115,135. Similar results

were also found in transplantation136 and luxation type trauma134.

Chemical

Early attempts to produce ankylosis via chemical means were documented by Rubin

& Biederman137 in 1961. Phenol was placed on parts of surgically exposed root

surfaces but no ankylosis was noted. Phenol was reused in a later attempt134 and was

similarly unsuccessful. However, in a study by Erausqin & Devoto138, formalin and

formaldehyde were used successfully to produce widespread ankylosis. Furthermore,

trioxymethylene-corticoid and acrylic spherule paste both were found to be capable of

inducing ankylosis although the results were inconsistent. Zinc oxide eugenol was

also tested in the same study but was found to be a poor causative agent for dental

ankylosis.

46

The systemic delivery of 1-hydroxyethylidene-1, 1-bisphosphonate (HEBP) was

reported by Wesselink & Beertsen139 to be capable of inducing ankylosis in the mouse.

Electrical

According to Rubin et al.134, electrical diathermy was carried out by Gottlieb and

Orban in 1930 and was shown to be successful in inducing ankylosis. However, no

other studies have since employed a similar methodology.

Thermal

Positive experimental ankylosis has been reported from the generation of heat through

the root canal systems in both rat and monkey incisors140-142. In 1982, Michaeli et

al.142 generated heat via the direct application of electrocautery needle into the pulp

cavity which seemed to limit the injury primarily to the periodontal ligament. Their

experiment demonstrated fusion of the root to bone within 7 days. In contrast,

Atrizadeh et al.140 and Line et al.141 did not observe ankylosis until 1 month after the

heat application which probably reflected differences in their methodology.

The application of ultra-low temperatures to the periodontal ligament apparatus was

not reported until 1986 when Wesselink et al.143 applied liquid nitrogen and Tah &

Stahl144 utilised cryoprobe on the buccal alveolar plate of rats. Both studies reported

consistent ankylosis formation although widespread alveolar bone (particularly on the

buccal experimental side) necrosis was also noted. Tah et al.145 repeated the protocol

at a later date and had similar success in inducing ankylotic union. A less traumatic

method was developed by Dreyer et al.146 which allowed for the induction of dental

ankylosis while limiting the majority of the insult to the periodontal ligament. This

was achieved via the application of dry ice to the occlusal surface of the tooth crown

hence insulting the periodontal ligament apparatus indirectly through the thermal

conductivity of the enamel and particularly the dentinal tubules. Recently, this

protocol was applied by Shaboodien147 and Di Iulio148. Similar histological findings

were reported although in Di Iulio’s experiment, ankylosis was not a consistent

finding.

47

Denervation

Fujiyama et al.51 denervated rat teeth by transection of the inferior alveolar nerve.

This led to the formation of ankylosis with the authors attributing this finding to the

concurrent decrease in rests of Malassez. This was in contrast to an earlier study by

Berggreen et al.149 which found little association between denervation and the

formation of ankylosis although both the experimental and control teeth underwent

replantation which is known to cause ankylosis and hence would mask any real

difference denervation may have had on the periodontal ligament apparatus

Immunohistochemistry Immunohistochemistry is a method for localising specific antigens (proteins of

interest) in tissues or cells based on antigen-antibody recognition; it seeks to exploit

the specificity provided by the binding of an antibody with its antigen and which can

be detected at a light microscopic level150.

There are two main types of antibodies: monoclonal and polyclonal. Polyclonal

antibodies are a cocktail of various antibodies that may recognise several binding sites

(epitopes) whereas monoclonal only recognises one particular epitope. Hence,

polyclonal antibodies may be more sensitive but monoclonal antibodies are more

specific.

The staining of a cellular epitope provides an insight to the molecular role which the

protein may play in a given condition. Immunohistochemistry has proven to be

invaluable to both diagnosis and research.

There are many techniques in immunohistochemistry and they include:

• one-step direct conjugate

• two-step indirect method

• peroxidase antiperoxidase (PAP)

• avidin-biotin complex/conjugate (ABC)

• biotin-streptavidin (B-SA)

• polymer-based labelling systems

48

Avidin-biotin complex procedure (ABC) This technique utilises the high affinity binding between biotin and avidin. In this

technique, a primary antibody with biotin can first be used to locate the target antigen.

Horseradish peroxidase that is conjugated to avidin can then be added (or a secondary

antibody first, which improves on the specificity and sensitivity of the results)

resulting in the binding between the primary antibody and the horseradish complex

thus localising and manifesting the target antigen/protein for the investigator150.

The ABC method does have its disadvantages in that many tissues contain significant

amounts of endogenous biotin that may result in false positive staining (this

background staining can be eliminated by specific blocking solutions). Secondly, it

has been shown that various batches of biotin and avidin tend to differ in their

affinities and binding power thus affecting the predictability of certain staining

protocols150.

Nevertheless, overall the ABC method is arguably the most commonly used

immunohistochemical method due to its simplicity and reasonably reliable/predictable

outcomes.

49

References

1. Gaengler P, Metzler E. The periodontal differentiation in the phylogeny of teeth--an overview. J Periodontal Res 1992;27:214-225. 2. Peyer B. Comparative odontology. Chicago, USA: The university of chicago press; 1968. 3. Tomes CS. A manual of dental anatomy: human and comparative. London: J & A Churchill; 1923. 4. McIntosh JE, Anderton X, Flores-De-Jacoby L, Carlson DS, Shuler CF, Diekwisch TG. Caiman periodontium as an intermediate between basal vertebrate ankylosis-type attachment and mammalian "true" periodontium. Microsc Res Tech 2002;59:449-459. 5. Shellis RP. Comparative anatomy of tooth attachment. In: Berkovitz BKB, Moxham B, Newman H, editors. The periodontal ligament in health and disease. Great Britain: Pergamon Press; 1982. p. 3-24. 6. Hillson S. Teeth. Cambridge, Melbourne: Cambridge University Press; 2005. 7. Tennant M. Oral Histology Primary examination orientation of the Royal Australasian College of Dental Surgeons. Sydney: Royal Australasian College of Dental Surgeons; 2006. 8. Ten Cate AR. Oral Histology: structure and function. St Louis: Mosby; 1989. 9. Tortora GJ, Grabowski SR. Principles of anatomy and physiology. New York: Harper Collins College; 1993. 10. Meikle MC. Craniofacial development, growth & evolution. Norfolk: Bateson; 2002. 11. Bilezikian JP, Raisz LG, Rodan GA. Principles of bone biology. San Diego: Academic Press; 1996. 12. Hill PA. Bone remodelling. Br J Orthod 1998;25:101-107. 13. Hadjidakis DJ, Androulakis, II. Bone remodeling. Ann N Y Acad Sci 2006;1092:385-396. 14. Fitzpatrick LA, Bilezikian JP. Parathyroid hormone: structure, function, and dynamic actions. In: Seibel MJ, Robins SP, Bilezikian JP, editors. Dynamics of bone and cartilage metabolism. London, San Diego: Academic Press; 1999. p. 187-202. 15. Martin TJ, Udagawa N. Hormonal regulation of osteoclast function. Trends Endocrinol Metab 1998;9:6-12.

50

16. Babich M, Foti LR. E-cadherins identified in osteoblastic cells: effects of parathyroid hormone and extracellular calcium on localization. Life Sci 1994;54:PL201-208. 17. Schluter KD, Hellstern H, Wingender E, Mayer H. The central part of parathyroid hormone stimulates thymidine incorporation of chondrocytes. J Biol Chem 1989;264:11087-11092. 18. Carmeliet G, Verstuyf A, Daci E, Bouillon R. The vitamin D hormone and its nuclear receptor: genomic mechanisms involved in bone biology. In: Seibel MJ, Robins SP, Bilezikian JP, editors. Dynamics of bone and cartilage metabolism. London, San Diego: Academic Press; 1999. p. 217-232. 19. Weinstein RS, Jilka RL, Parfitt AM, Manolagas SC. Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids. Potential mechanisms of their deleterious effects on bone. J Clin Invest 1998;102:274-282. 20. Britto JM, Fenton AJ, Holloway WR, Nicholson GC. Osteoblasts mediate thyroid hormone stimulation of osteoclastic bone resorption. Endocrinology 1994;134:169-176. 21. Oursler MJ, Pederson L, Fitzpatrick L, Riggs BL, Spelsberg T. Human giant cell tumors of the bone (osteoclastomas) are estrogen target cells. Proc Natl Acad Sci U S A 1994;91:5227-5231. 22. Robinson JA, Riggs BL, Spelsberg TC, Oursler MJ. Osteoclasts and transforming growth factor-beta: estrogen-mediated isoform-specific regulation of production. Endocrinology 1996;137:615-621. 23. Khosla S, Spelsberg TC, Lawrence B. Sex steroid effects on bone metabolism. In: Seibel MJ, Robins SP, Bilezikian JP, editors. Dynamics of bone and cartilage metabolism. London, San Diego: Academic Press; 1999. p. 233-246. 24. Manolagas SC, Jilka RL. Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N Engl J Med 1995;332:305-311. 25. Tremollieres FA, Strong DD, Baylink DJ, Mohan S. Progesterone and promegestone stimulate human bone cell proliferation and insulin-like growth factor-2 production. Acta Endocrinol (Copenh) 1992;126:329-337. 26. Reifenstein EC, Albright F. The Metabolic Effects of Steroid Hormones in Osteoporosis. J Clin Invest 1947;26:24-56. 27. Mochizuki H, Hakeda Y, Wakatsuki N, Usui N, Akashi S, Sato T et al. Insulin-like growth factor-I supports formation and activation of osteoclasts. Endocrinology 1992;131:1075-1080.

51

28. Hock JM, Centrella M, Canalis E. Insulin-like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology 1988;122:254-260. 29. Chenu C, Pfeilschifter J, Mundy GR, Roodman GD. Transforming growth factor beta inhibits formation of osteoclast-like cells in long-term human marrow cultures. Proc Natl Acad Sci U S A 1988;85:5683-5687. 30. Shinar DM, Rodan GA. Biphasic effects of transforming growth factor-beta on the production of osteoclast-like cells in mouse bone marrow cultures: the role of prostaglandins in the generation of these cells. Endocrinology 1990;126:3153-3158. 31. Croucher PI, Russell RGG. Growth factors. In: Seibel MJ, Robins SP, Bilezikian JP, editors. Dynamics of bone and cartilage metabolism. London, San Diego: Academic Press; 1999. p. 83-96. 32. Noda M, Rodan GA. Type-beta transforming growth factor inhibits proliferation and expression of alkaline phosphatase in murine osteoblast-like cells. Biochem Biophys Res Commun 1986;140:56-65. 33. Leboy PS. Regulating bone growth and development with bone morphogenetic proteins. Ann N Y Acad Sci 2006;1068:14-18. 34. Wang EA, Israel DI, Kelly S, Luxenberg DP. Bone morphogenetic protein-2 causes commitment and differentiation in C3H10T1/2 and 3T3 cells. Growth Factors 1993;9:57-71. 35. LInd M, Eriksen EF, Bunger C. Bone morphogenetic protein-2 but not bone morphogenetic protein-4 and -6 stimulates chemotactic migration of human osteoblasts, human marrow osteoblasts, and U2-OS cells. Bone 1996;18:53-57. 36. Canalis E, Centrella M, McCarthy T. Effects of basic fibroblast growth factor on bone formation in vitro. J Clin Invest 1988;81:1572-1577. 37. Nagai H, Tsukuda R, Mayahara H. Effects of basic fibroblast growth factor (bFGF) on bone formation in growing rats. Bone 1995;16:367-373. 38. Nakamura T, Hanada K, Tamura M, Shibanushi T, Nigi H, Tagawa M et al. Stimulation of endosteal bone formation by systemic injections of recombinant basic fibroblast growth factor in rats. Endocrinology 1995;136:1276-1284. 39. Hsieh SC, Graves DT. Pulse application of platelet-derived growth factor enhances formation of a mineralizing matrix while continuous application is inhibitory. J Cell Biochem 1998;69:169-180. 40. Lorenzo JA, Raisz LG. Cytokines and prostaglandins. In: Seibel MJ, Robins SP, Bilezikian JP, editors. Dynamics of bone and cartilage metabolism. London, San Diego: Academic Press; 1999. p. 97-110.

52

41. Mundy GR. Local factors regulating osteoclast function. In: Rifkin BR, Gay CV, editors. Biology and physiology of the osteoclast. USA: CRC Press; 1992. p. 171-186. 42. Stashenko P, Dewhirst FE, Rooney ML, Desjardins LA, Heeley JD. Interleukin-1 beta is a potent inhibitor of bone formation in vitro. J Bone Miner Res 1987;2:559-565. 43. Bertolini DR, Nedwin GE, Bringman TS, Smith DD, Mundy GR. Stimulation of bone resorption and inhibition of bone formation in vitro by human tumour necrosis factors. Nature 1986;319:516-518. 44. Pierce AM, Lindskog S, Hammarstrom L. Osteoclasts: structure and function. Electron Microsc Rev 1991;4:1-45. 45. Katagiri T, Takahashi N. Regulatory mechanisms of osteoblast and osteoclast differentiation. Oral Dis 2002;8:147-159. 46. Roodman GD. Advances in bone biology: the osteoclast. Endocr Rev 1996;17:308-332. 47. Roodman GD. Cell biology of the osteoclast. Exp Hematol 1999;27:1229-1241. 48. Vaananen HK, Zhao H, Mulari M, Halleen JM. The cell biology of osteoclast function. J Cell Sci 2000;113 ( Pt 3):377-381. 49. Zaidi M, Pazianas M, Shankar VS, Bax BE, Bax CM, Bevis PJ et al. Osteoclast function and its control. Exp Physiol 1993;78:721-739. 50. Berkovitz BKB, Shore R. Cells of the periodontal ligament. In: Berkovitz BKB, Moxham B, Newman H, editors. The Periodontal Ligament in Health and Disease. London: Mosby-Wolfe; 1995. p. 9-34. 51. Fujiyama K, Yamashiro T, Fukunaga T, Balam TA, Zheng L, Takano-Yamamoto T. Denervation resulting in dento-alveolar ankylosis associated with decreased Malassez epithelium. J Dent Res 2004;83:625-629. 52. Lindskog S, Blomlof L, Hammarstrom L. Evidence for a role of odontogenic epithelium in maintaining the periodontal space. J Clin Periodontol 1988;15:371-373. 53. Spouge JD. A new look at the rests of Malassez. A review of their embryological origin, anatomy, and possible role in periodontal health and disease. J Periodontol 1980;51:437-444. 54. Bosshardt DD. Are cementoblasts a subpopulation of osteoblasts or a unique phenotype? J Dent Res 2005;84:390-406. 55. Andreasen JO, Andreasen FM, Andersson L. Textbook and color atlas of traumatic injuries to the teeth. Oxford, UK: Blackwell; 2007.

53

56. Komaki M, Karakida T, Abe M, Oida S, Mimori K, Iwasaki K et al. Twist negatively regulates osteoblastic differentiation in human periodontal ligament cells. J Cell Biochem 2007;100:303-314. 57. Spouge JD. The rests of Malassez and chronic marginal periodontitis. J Clin Periodontol 1984;11:340-347. 58. Wesselink PR, Beertsen W. The prevalence and distribution of rests of Malassez in the mouse molar and their possible role in repair and maintenance of the periodontal ligament. Arch Oral Biol 1993;38:399-403. 59. Ten Cate AR. The Histochemical Demonstration of Specific Oxidative Enzymes and Glycogen in the Epithelial Cell Rests of Malassez. Arch Oral Biol 1965;10:207-213. 60. Loe H, Waerhaug J. Experimental replantation of teeth in dogs and monkeys. Arch Oral Biol 1961;3:176-184. 61. Fong MM, Darendeliler MA, Hunter N, Shen G. Epithelial cells in PDL are critical in resuming the integral relation between tooth root and supporting bone after trauma--a transplantation experiment. Arch Oral Biol 2007;52:182-188. 62. Melcher AH. Repair of wounds in the periodontium of the rat. Influence of periodontal ligament on osteogenesis. Arch Oral Biol 1970;15:1183-1204. 63. Melcher AH, Cheong T. Fibroblast-like cells depress formation of bone-like tissue in vitro. Journal of Dental Research 1988;67:290. 64. Lekic P, McCulloch CA. Periodontal ligament cell population: the central role of fibroblasts in creating a unique tissue. Anat Rec 1996;245:327-341. 65. Ogiso B, Hughes FJ, Davies JE, McCulloch CA. Fibroblastic regulation of osteoblast function by prostaglandins. Cell Signal 1992;4:627-639. 66. Ogiso B, Hughes FJ, Melcher AH, McCulloch CA. Fibroblasts inhibit mineralised bone nodule formation by rat bone marrow stromal cells in vitro. J Cell Physiol 1991;146:442-450. 67. Kirkham J, Brookes SJ, Shore RC, Bonass WA, Robinson C. The effect of glycosylaminoglycans on the mineralization of sheep periodontal ligament in vitro. Connect Tissue Res 1995;33:23-29. 68. Cho MI, Garant PR. Development and general structure of the periodontium. Periodontol 2000 2000;24:9-27. 69. Ohno S, Doi T, Fujimoto K, Ijuin C, Tanaka N, Tanimoto K et al. RGD-CAP (betaig-h3) exerts a negative regulatory function on mineralization in the human periodontal ligament. J Dent Res 2002;81:822-825.

54

70. Li S, Yang PS, Cao JF, Ge SH, Pan KQ. [Expression of epidermal growth factor receptor in human periodontal ligament cells during their mineralization in vitro]. Hua Xi Kou Qiang Yi Xue Za Zhi 2006;24:11-14. 71. Duarte WR, Iimura T, Takenaga K, Ohya K, Ishikawa I, Kasugai S. Extracellular role of S100A4 calcium-binding protein in the periodontal ligament. Biochem Biophys Res Commun 1999;255:416-420. 72. Duarte WR, Shibata T, Takenaga K, Takahashi E, Kubota K, Ohya K et al. S100A4: a novel negative regulator of mineralization and osteoblast differentiation. J Bone Miner Res 2003;18:493-501. 73. Kato C, Kojima T, Komaki M, Mimori K, Duarte WR, Takenaga K et al. S100A4 inhibition by RNAi up-regulates osteoblast related genes in periodontal ligament cells. Biochem Biophys Res Commun 2005;326:147-153. 74. Yoshizawa T, Takizawa F, Iizawa F, Ishibashi O, Kawashima H, Matsuda A et al. Homeobox protein MSX2 acts as a molecular defense mechanism for preventing ossification in ligament fibroblasts. Mol Cell Biol 2004;24:3460-3472. 75. Bialek P, Kern B, Yang X, Schrock M, Sosic D, Hong N et al. A twist code determines the onset of osteoblast differentiation. Dev Cell 2004;6:423-435. 76. Henry SP, Takanosu M, Boyd TC, Mayne PM, Eberspaecher H, Zhou W et al. Expression pattern and gene characterization of asporin. a newly discovered member of the leucine-rich repeat protein family. J Biol Chem 2001;276:12212-12221. 77. Lorenzo P, Aspberg A, Onnerfjord P, Bayliss MT, Neame PJ, Heinegard D. Identification and characterization of asporin. a novel member of the leucine-rich repeat protein family closely related to decorin and biglycan. J Biol Chem 2001;276:12201-12211. 78. Yamada S, Murakami S, Matoba R, Ozawa Y, Yokokoji T, Nakahira Y et al. Expression profile of active genes in human periodontal ligament and isolation of PLAP-1, a novel SLRP family gene. Gene 2001;275:279-286. 79. Ikegawa S. Expression, regulation and function of asporin, a susceptibility gene in common bone and joint diseases. Curr Med Chem 2008;15:724-728. 80. Yamada S, Kitamura M, Murakami S. PLAP-1: A novel molecule regulating homeostasis of periodontal tissues. Japanese Dental Science Review (2008) 44, 137—144 2008;44:137-144. 81. Tomoeda M, Yamada S, Shirai H, Ozawa Y, Yanagita M, Murakami S. PLAP-1/asporin inhibits activation of BMP receptor via its leucine-rich repeat motif. Biochem Biophys Res Commun 2008;371:191-196. 82. Gruber HE, Ingram JA, Hoelscher GL, Zinchenko N, Hanley EN, Jr., Sun Y. Asporin, a susceptibility gene in osteoarthritis, is expressed at higher levels in the more degenerate human intervertebral disc. Arthritis Res Ther 2009;11:R47.

55

83. Yamada S, Tomoeda M, Ozawa Y, Yoneda S, Terashima Y, Ikezawa K et al. PLAP-1/asporin, a novel negative regulator of periodontal ligament mineralization. J Biol Chem 2007;282:23070-23080. 84. Lee E-H, Park H-J., Jeong, J-H., Kim, Y-J., Cha, D-W., Kwon, D-K., Lee, S-H., Cho, J-Y. The role of Asporin in mineralization of human dental pulp stem cells. J. Cell. Physiol 2011;226:1676-2682. 85. Nakajima M, Kizawa H, Saitoh M, Kou I, Miyazono K, Ikegawa S. Mechanisms for asporin function and regulation in articular cartilage. J Biol Chem 2007;282:32185-32192. 86. Raghoebar GM, Boering G, Vissink A. Clinical, radiographic and histological characteristics of secondary retention of permanent molars. J Dent 1991;19:164-170. 87. Oppenheim A. Histologische Beunde beim Zahnwexhsel. Ztschr. F. Stomatol. 1922;20:159. 88. Noyes F. Submerging deciduous molars. The Angle Orthodontist 1932;2:77-87. 89. Pindborg JJ. Pathology of the Dental Hard Tissues. Copenhagen: Munksgaard; 1970. 90. Ekim SL, Hatibovic-Kofman S. A treatment decision-making model for infraoccluded primary molars. Int J Paediatr Dent 2001;11:340-346. 91. Kurol J. Infraocclusion of primary molars: an epidemiologic and familial study. Community Dent Oral Epidemiol 1981;9:94-102. 92. Via WF, Jr. Submerged Deciduous Molars: Familial Tendencies. J Am Dent Assoc 1964;69:127-129. 93. Krakowiak FJ. Ankylosed primary molars. ASDC J Dent Child 1978;45:288-292. 94. Biederman W. The incidence and etiology of tooth ankylosis. Am. J. Orthodontics 1956;42:921-926. 95. Darling AI, Levers BG. Submerged human deciduous molars and ankylosis. Arch Oral Biol 1973;18:1021-1040. 96. Steigman S, Koyoumdjisky-Kaye E, Matrai Y. Submerged deciduous molars and congenital absence of premolars. J Dent Res 1973;52:842. 97. Brown ID. Some further observations on submerging deciduous molars. Br J Orthod 1981;8:99-107. 98. Kurol J, Thilander B. Infraocclusion of primary molars with aplasia of the permanent successor. A longitudinal study. Angle Orthod 1984;54:283-294.

56

99. Proffit WR, Fields H, Sarver D. Contemporary Orthodontics. Mosby Elsevier; 2007. 100. Phelan MK, Moss RB, Jr., Powell RS, Womble BA. Orthodontic management of ankylosed teeth. J Clin Orthod 1990;24:375-378. 101. Kurol J. Impacted and ankylosed teeth: why, when, and how to intervene. Am J Orthod Dentofacial Orthop 2006;129:S86-90. 102. Biederman W. The problem of the ankylosed tooth. Dent Clin North Am 1968:409-424. 103. Campbell KM, Casas MJ, Kenny DJ. Ankylosis of traumatized permanent incisors: pathogenesis and current approaches to diagnosis and management. J Can Dent Assoc 2005;71:763-768. 104. Brearley LJ, McKibben DH, Jr. Ankylosis of primary molar teeth. I. Prevalence and characteristics. ASDC J Dent Child 1973;40:54-63. 105. Douglass J, Tinanoff N. The etiology, prevalence, and sequelae of infraclusion of primary molars. ASDC J Dent Child 1991;58:481-483. 106. Biederman W. Etiology and treatment of tooth ankylosis. Am. J. Orthodontics 1962;48:670-684. 107. Dixon DA. Observations on submerging deciduous molars. The Dental Practitioner 1963;XIII:303-315. 108. Crowther L, Darendeliler AM, Vickers D, Dineen M, Adams R, Raphael S. Ankylosis Brighter Futures. Sydney: Australian Society of Orthodontist (NSW) and University of Sydney; 2007. 109. Mullally BH, Blakely D, Burden DJ. Ankylosis: an orthodontic problem with a restorative solution. Br Dent J 1995;179:426-429. 110. Andersson L, Blomlof L, Lindskog S, Feiglin B, Hammarstrom L. Tooth ankylosis. Clinical, radiographic and histological assessments. Int J Oral Surg 1984;13:423-431. 111. Andreasen JO. Analysis of pathogenesis and topography of replacement root resorption (ankylosis) after replantation of mature permanent incisors in monkeys. Swed Dent J 1980;4:231-240. 112. Mitchell DL, West JD. Attempted orthodontic movement in the presence of suspected ankylosis. Am J Orthod 1975;68:404-411. 113. Proffit WR, Vig KW. Primary failure of eruption: a possible cause of posterior open-bite. Am J Orthod 1981;80:173-190.

57

114. Andreasen JO, Skougaard MR. Reversibility of surgically induced dental ankylosis in rats. Int J Oral Surg 1972;1:98-102. 115. Hammarstrom L, Blomlof L, Lindskog S. Dynamics of dentoalveolar ankylosis and associated root resorption. Endod Dent Traumatol 1989;5:163-175. 116. Raghoebar GM, van Koldam WA, Boering G. Spontaneous reeruption of a secondarily retained permanent lower molar and an unusual migration of a lower third molar. Am J Orthod Dentofacial Orthop 1990;97:82-84. 117. Belanger GK, Strange M, Sexton JR. Early ankylosis of a primary molar with self correction: case report. Pediatr Dent 1986;8:37-40. 118. Franklin CD. Ankylosis of an unerupted third molar by inostosis of enamel. A case report. Br Dent J 1972;133:346-347. 119. Dixon DA. Observations on submerging deciduous molars. Transactions of the British Society for the Study of Orthodontics 1962;00:101-114. 120. Albers DD. Ankylosis of teeth in the developing dentition. Quintessence Int 1986;17:303-308. 121. Bjerklin K, Bennett J. The long-term survival of lower second primary molars in subjects with agenesis of the premolars. Eur J Orthod 2000;22:245-255. 122. Ith-Hansen K, Kjaer I. Persistence of deciduous molars in subjects with agenesis of the second premolars. Eur J Orthod 2000;22:239-243. 123. Geiger AM, Bronsky MJ. Orthodontic management of ankylosed permanent posterior teeth: a clinical report of three cases. Am J Orthod Dentofacial Orthop 1994;106:543-548. 124. Raghoebar GM, Boering G, Jansen HW, Vissink A. Secondary retention of permanent molars: a histologic study. J Oral Pathol Med 1989;18:427-431. 125. Anholm JM, Crites DA, Hoff R, Rathbun WE. Corticotomy-facilitated orthodontics. Cda J 1986;14:7-11. 126. Malmgren B, Cvek M, Lundberg M, Frykholm A. Surgical treatment of ankylosed and infrapositioned reimplanted incisors in adolescents. Scand J Dent Res 1984;92:391-399. 127. Cohenca N, Stabholz A. Decoronation - a conservative method to treat ankylosed teeth for preservation of alveolar ridge prior to permanent prosthetic reconstruction: literature review and case presentation. Dent Traumatol 2007;23:87-94. 128. Diaz JA, Sandoval HP, Pineda PI, Junod PA. Conservative treatment of an ankylosed tooth after delayed replantation: a case report. Dent Traumatol 2007;23:313-317.

58

129. Filippi A, Pohl Y, von Arx T. Decoronation of an ankylosed tooth for preservation of alveolar bone prior to implant placement. Dent Traumatol 2001;17:93-95. 130. Huck L, Korbmacher H, Niemeyer K, Kahl-Nieke B. Distraction osteogenesis of ankylosed front teeth with subsequent orthodontic fine adjustment. J Orofac Orthop 2006;67:297-307. 131. Alcan T. A miniature tooth-borne distractor for the alignment of ankylosed teeth. Angle Orthod 2006;76:77-83. 132. Kinzinger GS, Janicke S, Riediger D, Diedrich PR. Orthodontic fine adjustment after vertical callus distraction of an ankylosed incisor using the floating bone concept. Am J Orthod Dentofacial Orthop 2003;124:582-590. 133. Parker WS, Frisbe HE, Grant TS. The experimental production of dental ankylosis. Angle Orthod 1964;34:103-107. 134. Rubin PL, Weisman EJ, Bisk F. Experimental tooth ankylosis in the monkey. Angle Orthod 1984;54:67-72. 135. Sherman P, Jr. Intentional replantation of teeth in dogs and monkeys. J Dent Res 1968;47:1066-1071. 136. Morris ML, Moreinis A, Patel R, Prestup A. Factors affecting healing after experimentally delayed tooth transplantation. J Endod 1981;7:80-84. 137. Rubin PL, Biederman W. Attempt to produce tooth ankylosis. J Dent Res 1961;40:744. 138. Erausquin J, Devoto FC. Alveolodental ankylosis induced by root canal treatment in rat molars. Oral Surg Oral Med Oral Pathol 1970;30:105-116. 139. Wesselink PR, Beertsen W. Ankylosis of the mouse molar after systemic administration of 1-hydroxyethylidene-1,1-bisphosphonate (HEBP). J Clin Periodontol 1994;21:465-471. 140. Atrizadeh F, Kennedy J, Zander H. Ankylosis of teeth following thermal injury. J Periodontal Res 1971;6:159-167. 141. Line SE, Polson AM, Zander HA. Relationship between periodontal injury, selective cell repopulation and ankylosis. J Periodontol 1974;45:725-730. 142. Michaeli Y, Pitaru S, Zajicek G. Localized damage to the periodontal ligament and its effect on the eruptive process of the rat incisor. J Periodontal Res 1982;17:300-308. 143. Wesselink PR, Beertsen W, Everts V. Resorption of the mouse incisor after the application of cold to the periodontal attachment apparatus. Calcif Tissue Int 1986;39:11-21.

59

144. Tal H, Stahl SS. Healing following devitalization of sites within the periodontal ligament by ultralow temperatures. J Periodontol 1986;57:735-741. 145. Tal H, Kozlovsky A, Pitaru S. Healing of sites within the dog periodontal ligament after application of cold to the periodontal attachment apparatus. J Clin Periodontol 1991;18:543-547. 146. Dreyer CW, Pierce AM, Lindskog S. Hypothermic insult to the periodontium: a model for the study of aseptic tooth resorption. Endod Dent Traumatol 2000;16:9-15. 147. Shaboodien SI. Traumatically induced dentoalveolar ankylosis in rats. Adelaide: University of Adelaide; 2005. 148. Di Iulio DS. Relationship of epithelial cells and nerve fibres to experimentally induced dentoalveolar ankylosis in the rat. Adelaide: University of Adelaide; 2007. 149. Berggreen E, Sae-Lim V, Bletsa A, Heyeraas KJ. Effect of denervation on healing after tooth replantation in the ferret. Acta Odontol Scand 2001;59:379-385. 150. Dabbs DJ. Diagnostic immunohistochemistry. New York, Edinburgh: Churchill Livingstone; 2006.

60

SECTION 2

61

STATEMENT OF PURPOSE In a healthy periodontium, the periodontal ligament functions to connect the teeth to

the jaws, sustain the masticatory load, provides sensory information and also prevents

bony union of the roots of the teeth to the alveolar bone.

The cells of the periodontal ligament have been shown to be osteogenic but under

normal conditions, the PDL space remains patent without the occurrence of ankylosis.

Periodontal Ligament Associated Protein-1 (PLAP-1)/Asporin are a recently

discovered protein that has been suggested to play a significant role in suppressing the

osteogenic tendency of the periodontal ligament and maintaining the fibrous

ligamentous nature of the periodontal ligament. Furthermore, PLAP-1/Asporin has

also been shown to be associated with the differentiation and mineralisation of dental

pulp stem cells.

Therefore, it is the intention of this project to investigate the relationship between

PLAP-1/Asporin and the periodontal ligament space as well as the dental pulp using

an ankylosis model.

The investigation has been divided into two parts with separate hypothesis and aims

as stated below.

Paper 1

Aims

To confirm the location specificity of PLAP-1 within the maxilla using

immunohistochemistry.

To investigate the expression of PLAP-1 in the periodontal ligament using a rat

ankylosis model.

Hypothesis

There is no change in the expression of PLAP-1 in ankylotic areas due to hypothermal

insult compared to non-ankylotic areas within the periodontal ligament.

62

Paper 2

Aims To determine the normal expression of PLAP-1/Asporin within the pulp chamber

using immunohistochemistry.

To investigate the relationship between PLAP-1/Asporin expression in the coronal

pulp following a hypothermic insult using immunohistochemistry.

Hypothesis

There is no change in the expression of PLAP-1 in traumatised pulp compared to non-

traumatised pulp.

It is intended that both papers be submitted to the journal Archives of Oral Biology

and hence is prepared in the journal’s requested format.

63

PAPER 1 (Prepared for submission in Archives of Oral Biology)

Title: The role of PLAP-1/Asporin in the Maintenance of the Rat Periodontal

Ligament using a Rat Ankylosis Model.

Authors: Chen, W.C., Sampson, W., Dreyer, C., Dharmapatni, K.

Affiliations:

Chen, W.C. – Postgraduate student, Orthodontic Unit, School of Dentistry, Faculty of

Health Sciences, University of Adelaide, South Australia.

Sampson, W. – Professor and P.R. Begg Chair, Orthodontic Unit, School of Dentistry,

Faculty of Health Sciences, University of Adelaide, South Australia.

Dreyer, C. – Associate Professor, Orthodontic Unit, School of Dentistry, Faculty of

Health Sciences, University of Adelaide, South Australia.

Dharmapatni, K. – Research Fellow, Hanson Institute, School of Medicine, Faculty of

Health Sciences, University of Adelaide, South Australia.

Abstract

Background: Periodontal Ligament Associated Protein-1(PLAP-1)/Asporin is a novel

protein suggested to have an important role in the regulation of the periodontal

ligament space. However, the scarce data available demonstrate contrasting results.

Currently, there are no studies which investigate the expression and relationship of

PLAP-1 and periodontal ligament (PDL) space using an ankylosis model.

Aims: The aims of this study are to determine the distribution of PLAP-1 protein

within the maxilla and to explore correlation between PLAP-1 expression and

ankylosis in a transient aseptic ankylosis model.

64

Methods: The maxillary right first molars of 30 male Sprague-Dawley rats were

subjected to a single 20 minute application of dry ice to induce PDL ankylosis.

Groups of five animals were sacrificed after 7, 10, 14, 18, 21 and 28 days of treatment

respectively. The maxillae were dissected out and underwent routine tissue fixation

and processing for immunohistochemical detection of PLAP-1 expression. The

immunostained cells were then analysed semi-quantitatively using a standardised grid

system.

Results: PLAP-1 was expressed in the PDL, dental pulp, blood vessel walls and the

nasal cartilage. Not all sections contained ankylosis. Sections which did not contain

ankylosis demonstrated no significant PLAP-1 expression differences between control

and experimental sides. Sections that did demonstrate ankylosis yielded a tendency

towards increased PLAP-1 reactivity especially near the cementum. However, it was

difficult to deduce whether the relationship of PLAP-1 to the ankylotic union was

associated with bone formation or resorptive activities.

Conclusion: The current investigation suggests that PLAP-1 may be involved in bone

metabolism but future investigations are required to elucidate its true role within the

periodontal ligament.

Key Words: periodontal ligament associated protein-1, asporin, ankylosis,

periodontal ligament, maintenance.

Introduction

In a healthy periodontium, the periodontal ligament functions to connect the teeth to

the jaws, sustain the masticatory load, provide sensory information and also prevent

bony union (ankylosis) of the tooth roots to the alveolar bone.

Interestingly, cells of the periodontal ligament have been shown to be osteogenic in

laboratory studies.1-3 Furthermore, it is common knowledge that the death or removal

of periodontal ligament cells leads to the loss of periodontal membrane eventually

65

resulting in extensive root resorption and widespread ankylosis. Presumably, certain

factors/cells exist within the periodontal ligament which must normally suppress the

osteogenic tendency and maintain the fibrous ligamentous nature of the periodontal

ligament.

The search for the cell(s) and/or factor(s) which are responsible for maintaining the

patency of the periodontal ligament is not new. Early investigations have suggested

several cells/factors which could play a vital role in inhibiting ankylosis and they

include the cell rests of Malassez3-6, glycosaminoglycans7 and fibroblasts8,9 as well as

their associated prostaglandins10,11. However, no agreement has been reached

regarding the exact cell/factor(s) involved in periodontal ligament maintenance.

Recent advances in technology have allowed more detailed investigations of the

periodontal ligament at a molecular level. This has led to the discovery of several

new factors which may play a role in the maintenance of the periodontal ligament

space: RGD-CAP12 (a collagen-associated protein which contains the arginine-

glycine-aspartic acid sequence)12; epidermal growth factor receptors13; S100A414,15

(member of the S100 calcium binding family; Msx216 (homeobox protein); Twist17 (a

basic helix loop helix protein) and PLAP-118,19 (periodontal ligament associated

protein-1).

Of all the possible factors, PLAP-1 (also known as Asporin) demonstrates the most

promise as, unlike the other molecular factors mentioned which are expressed

ubiquitously, PLAP-1 is thought to be highly location specific. In fact, the only

region in the maxilla demonstrating its presence is reported to be the periodontal

ligament space.19 PLAP-1 is a member of the leucine-rich repeat proteoglycan

family. Furthermore, in situ hybridisation has shown that PLAP-1 is highly

expressed in the dental follicle during tooth formation indicating a central role for

PLAP-1 in the development of the periodontal tissues. In addition, Yamada et al.19

have shown that PLAP-1 regulates periodontal ligament cell cytodifferentiation and

also mineralisation through its negative feedback interaction with bone morphogenetic

66

protein-2 (BMP-2). A subesequent study revealed that the mechanism by which

PLAP-1 exerts the described biological effect is through the leucine-rich repeats

(LRR) motif. A particular LRR motif that appears to be highly associated with

PLAP-1 is LRR5.18

A recent investigation20 using a previously established ankylosis protocol21 has

revealed the possibility of spontaneous resolution of the ankylotic area demonstrating

the capacity for the periodontal ligament to regenerate and repair minor damages.

Utilising the same tissue samples will provide a unique opportunity to investigate the

expression and possible role of PLAP-1 in the formation and regulation of ankylosis.

Aims

• To confirm the location specificity of PLAP-1 within the rat maxilla using

immunohistochemistry.

• To investigate the correlation between PLAP-1 expression in the rat

periodontal ligament using an ankylosis model.

Hypothesis

Null Hypothesis

There is no change in the expression of PLAP-1 in ankylotic areas due to hypothermal

insult compared to non-ankylotic areas within the periodontal ligament.

Materials & Methods

The maxillary right first molars of 30, 8-week-old male Sprague-Dawley rats (housed

in the University of Adelaide Animal House) were subjected to a single 20 minute

application of dry ice in order to induce sterile necrosis and ankylosis in the inter-

radicular region. Anaesthesia was achieved via a 1:1 combination of Hypnorm®

67

(Janssen-Cilag Ltd., Buckinghamshire, UK) and Hypnovel® (Roche, Berne,

Switzerland). The contralateral first molar served as a control. Groups of five

animals were sacrificed via cardiac perfusion with 4% paraformaldehyde after 7, 10,

14, 18, 21 and 28 days, respectively. The maxillae were dissected out and underwent

decalcification using 4% EDTA and were then paraffin embedded. Serial 7µm

coronal sections of the furcation area were obtained using a Leitz 1512 microtome and

then placed onto silane coated slides.

The staining for PLAP-1 immunoreactivity was performed via the Labelled

StreptAvidin Biotin method. The primary antibody was an unlabelled rabbit

polyclonal antibody (450-31930, Sapphire Bioscience, NSW, Australia) diluted with

PBS at 1 to 400 of stock concentration and incubated overnight in a wet chamber at

room temperature (RT). The linkage reagent was a biotinylated goat anti-rabbit

secondary antibody conjugated to horseradish peroxidase (K060911, LSAB®2-HRP,

Dako, Australia) and was incubated for 30 minutes, RT. Visualisation of the target

protein (PLAP-1) was through the use of AEC dye (K3469, Dako, Australia).

Sections were then counterstained with haematoxylin and lithium carbonate. Growing

rat femoral head containing cartilage was used as the positive control as it has been

previously reported that chondrocytes express PLAP-1/Asporin.36 Sections incubated

only with a rabbit serum (N169987, Universal negative control, Dako, Australia) were

used as the negative control.

Stained sections were mounted and analysed using an Olympus B071 optical

microscope with the images displayed on an attached personal computer with a 24.0

inch monitor. The software programme Analysis (Olympus Soft Imaging Solutions,

Germany) was used for image processing and viewing. A grid was pre-constructed

according to Shaboodien’s22 method to aid in accuracy and reproducibility (figure 1).

A central vertical line divided the crown into equal halves and was oriented as parallel

as possible to the axis of the tooth. A horizontal line was established by joining and

extrapolating from the cemento-enamel junction of both the buccal and palatal side.

Four additional vertical lines were drawn (2 on either side of the central vertical line)

which helped divide the roots into equal halves (best approximation) and 3 additional

68

horizontal lines were drawn to divide the roots into equal cervical, middle and apical

third (Figure 1). A 4 X magnification was used to superimpose the pre-constructed

grid to allow for standardisation of the area of interest (shown by solid black lines in

Figure 1). Within the PDL itself, a subjective evaluation was made to divide it into

thirds – adjacent to the alveolar bone, middle third and adjacent to the cementum.

Due to the diffuse nature of the stain, a semi-quantitative scoring method was utilised

(0=no staining, 1=mild staining; 2=moderate staining, 3=intense staining) to measure

the staining intensity of PLAP-1.

Figure 1 Superimposition of grid system on sections with roots (The cementum, PDL & alveolar bone within the solid line is the area included for the analysis.)

Fifteen sections 70µms apart per rat were stained and analysed. Wald statistics for

Type 3 GEE (General Estimating Equation) Analysis as well as Post-hoc comparisons

(see Appendices) were made. One section from each rat was randomly selected for

repeat scoring and the intra-observer reliability was calculated with a weighted Kappa

coefficient. All calculations performed by SAS Version 9.2 (SAS Institute Inc., Cary,

NC, USA).

Results

Periodontal ligament associated protein-1 was found within the rat maxilla but not

exclusively expressed within the periodontal ligament (Fig 2). PLAP-1 was found to

69

be expressed in the dental pulp, nasal cartilage, mid-palatal suture, blood vessel walls,

epithelial tissues and the periodontal ligament space.

Figure 2

a) Rat femur articular cartilage used as positive control. Staining of the differentiated chondrocytes was noted but PLAP-1 was not expressed by all the cells. b) Negative control of the rat articular cartilage showing no staining. c) Tissue stained with PLAP-1. d) Negative control showing no dental pulp or PDL staining.

The staining obtained within the cartilaginous area was associated with the cytoplasm

of mainly developing chondrocytes up until the stage when they become hypertrophic.

However, this positive expression did not occur uniformly (Fig 3a). The expressions

within the chondrocytes were more defined as were the expressions detected on the

blood vessel walls (Fig 3b). This is in contrast to the expressions found within the

dental pulp and the periodontal ligament space which appeared diffuse (Fig 3c). The

results of PLAP-1 in relation to dental pulp will be presented in the second article.

a b

PLAP-1 stained chondrocytes

d

Alveolar Bone

dentine

pulp PDL

c

pulp

dentine

PDL

200µm 200µm

500µm 500µm

70

Under normal, un-traumatised circumstances, the periodontal ligament space adjacent

to the alveolar bone tended to exhibit a higher PLAP-1 expression than other regions

of the PDL space (Fig 3d) although this is often associated with the increased number

of blood vessels associated with the alveolar bone surface. The expression shown on

gingival epithelial tissues was widespread and generally most intense at the surface

epithelium and becoming less so towards the basement lamina (Fig 3e).

Figure 3

a) PLAP-1 staining of certain chondrocytes within the articular cartilage. b) PLAP-1 staining of blood vessels. c) PLAP-1 staining pattern within the PDL. d) PLAP-1 staining of the PDL adjacent to the alveolar bone and its association with blood vessels. e) PLAP-1 staining of gingival tissues.

b

e

d

a

c

PLAP-1 stained chondrocytes

Unstained chondrocytes

dentine Alveolar bone

PDL

dentine

dentine Gingival epithelial tissues

PDL

Alveolar bone

Alveolar bone

Blood vessels

PDL

200µm

200µm

100µm

100µm

500µm

71

Not all sections exhibited ankylosis within the furcation region and some could not be

included for analysis due to crucial regions being lost or damaged during tissue

processing. It was found that despite identical staining procedures and environmental

conditions, there were significant variations between animals in the intensity of

PLAP-1 expression. Ankylosis was deemed acquired when there was osseous-like

material intruding in the periodontal ligament space (Fig 4). In agreement with other

studies23,24 which used a similar method, the ankylosis ranged from small islands of

bone-like material to multiple columns/spicules connecting the inter-radicular bone to

the cementum of the root.

Ankylosis was mainly found in days 10, 14 and 18. There were extremely few

ankylotic unions found in days 7 and virtually none by days 21 and 28. The time

elapsed from initial trauma did not seem to have a statistically significant effect on the

expression of PLAP-1 (see Appendices).

Figure 4

a) with ankylosis on experimental side on day 14. b) without ankylosis on control side on day 14

Mixed results were obtained for the slides in which ankylosis was observed (Table 1).

In the cementum third of the periodontal ligament space, the experimental side

yielded statistically significant greater expression of PLAP-1 particularly for higher

intensities (P<0.0001) (Fig 5). However, this relationship was not found to be

consistent in the other PDL regions. Nevertheless, aside from low intensity staining in

the bone third of the PDL, the remaining subgroups (various intensities at each

dentine

dentine

Alveolar bone

Alveolar bone

PDL

PDL

Ankylotic union

pulp

pulp a b 200µm

200µm

72

location) all showed a tendency towards the experimental side demonstrating more

PLAP-1 expression. It was interesting to note that toward the apical regions, many of

the highly stained PLAP-1 cells seemed to be multinucleated cells sitting in resorptive

lacunae (Fig 6a). However, it is difficult to ascertain whether this also applies to the

ankylotic region due to the spicule nature of the bone-like material at the inter-

radicular area. However, stained regions adjacent to the cementum did not appear to

be associated with resorptive lacunae (Fig 6b).

Figure 5

a) shows ankylosis with more intense staining near the cementum in a day 18 experimental rat whereas b) shows very little staining near cementum third in control sides of a day 18 specimen (note also increased staining at the alveolar third compared to cementum third in (b).

Figure 6

a) Shows on a day 14 experimental rat that toward the apical region many multinucleated cells associated with resorptive lacunae are stained positive. b) Whilst PLAP-1 is positive all along the cementum side there were no multinucleated cells and no obvious resorptive lacunae (day 14 experimental rat).

a b PDL

PDL

Alveolar bone

Alveolar bone

Ankylotic union

pulp

dentine

dentine

500µm 100µm

a b

PDL

PDL

Alveolar bone

Alveolar bone

Ankylotic union

pulp

Multinucleated PLAP-1 stained cells in lacunae

Multinucleated PLAP-1 stained cells in lacunae

Increased staining along cementum 1/3rd of PDL

Root dentine

200µm 100µm

73

Table 1: Combined Post-hoc data comparing control and traumatised sides with ankylosis at various intensities of PLAP-1 expression

PDL Region PLAP-1

Intensity Score

Compared

groups

P-Value Odds Ratio

Cementum 1. Predictors of intensity ≥ 1

Control vs. Exp <.0001 0.041

2. Predictors of intensity ≥ 2

Control vs. Exp <.0001 0.060

3. Predictors of intensity ≥ 3

Control vs. Exp <.0001 0.014

Bone 1. Predictors of intensity ≥ 1

Control vs. Exp 0.4950 1.375

2. Predictors of intensity ≥ 2

Control vs. Exp 0.0044 0.510

3. Predictors of intensity ≥ 3

Control vs. Exp 0.0763 0.381

Middle third 1. Predictors of intensity ≥ 1

Control vs. Exp 0.2785 0.619

2. Predictors of intensity ≥ 2

Control vs. Exp <.0001 0.158

3. Predictors of intensity ≥ 3

Control vs. Exp No statistics derived due to lack of numbers

For the sections which did not show definite ankylosis, there were no consistent

differences in PLAP-1 expression between control and experimental sides at each

time point or region (Table 2). This included slides which theoretically should have

had ankylosis previously but may have resolved spontaneously.

Comparisons between repeat measurements show a weighted Kappa coefficient of

0.95 (95% CI 0.92, 0.97), indicating very good agreement.

74

Table 2: Comparison in PLAP-1 intensity between control and traumatised sides with no discernible ankylosis.

PDL region Days Compared

groups P-value Odds ratio

Cementum 7 Control vs Exp 0.1238 1.498

10 Control vs Exp 0.2767 1.412

14 Control vs Exp 0.0597 1.918

18 Control vs Exp 0.4792 0.833

21 Control vs Exp 0.8065 0.907

28 Control vs Exp 0.8390 0.882

Bone 7 Control vs Exp 0.2933 1.687

10 Control vs Exp 0.0856 2.893

14 Control vs Exp 0.0020 4.741

18 Control vs Exp 0.7555 1.122

21 Control vs Exp 0.0011 1.860

28 Control vs Exp 0.0707 1.841

middle 7 Control vs Exp 0.0036 3.185

10 Control vs Exp 0.0369 1.834

14 Control vs Exp 0.2563 2.143

18 Control vs Exp 0.0003 3.274

21 Control vs Exp 0.6198 1.257

28 Control vs Exp 0.4909 1.618

75

Discussion

The use of hypothermal trauma to induce ankylosis has been reported previously.25

Subsequent investigations of ankylosis using the same protocol found the production

of consistent and widespread osseous-like material bridging the periodontal ligament

space connecting the inter-radicular bone to the cementum of the root and it did not

subside at the end of the experimental time frame.22 The lack of consistent ankylosis

in the present study is likely due to subtle technique differences. As the cryotherapy is

delivered by dry ice pellets to a rat molar, it can be envisaged that numerous factors

such as size of dry ice pellets, the amount of pressure and the placement of the ice

pellets would contribute to the level of trauma delivered. This is suggested with many

of the sections containing ankylosis demonstrating a lack of viable osteocytes in the

crestal regions of the inter-radicular bone (presumably due to the hypothermic insult)

whilst this is not observed in sections without ankylosis. Nevertheless, the

spontaneous resolution of ankylotic areas in this study sample provided a unique

opportunity to investigate the maintenance mechanism of the periodontal ligament

space. Ideally, however, a reliable protocol should be established which allows for

consistent ankylosis formation whilst still allowing for spontaneous resolution to

occur; that is transient ankylosis.

Transient ankylosis, although rarely reported, is not new. Andreasen & Skougaard26,

Andersson27 and Blomlof & Lindskog28 all reported transient ankylosis. Whether

ankylosis is transient or permanent seems to depend on the size of the injured

region27. It is thought that with minor regions of insult, the root resorption craters are

repaired with cementum and ankylotic regions are replaced with vital periodontal

membrane. However, when the injury is severe and widespread, cells of the

periodontal ligament are not able to proliferate and replace the ankylotic areas hence

perpetuating the bony union.27 The tipping point between ankylosis and repair is not

well understood.

The finding of PLAP-1 expression in numerous regions of the maxilla is in

contradiction to the findings of Yamada & co-workers19 who found specific

76

expression of PLAP-1 within the PDL or its progenitor, the dental follicle. The

variation in findings may be due to methodological differences such as the use of

different primary antibody. However, a recent study29 also reported findings of

PLAP-1 expression within the dentine.

Lorenzo et al30 found PLAP-1/Asporin to be expressed in smooth muscle cells.

Theoretically, the smooth muscle cells present in the tunica media layer of the blood

vessel walls may lead to a positive reaction to anti-PLAP-1 antibodies. Unfortunately,

the tissues used in this study displayed no major vessels. However, there were

abundant capillaries with consistent and positive PLAP-1 staining which is limited to

its walls suggesting an association with endothelial cells.

The evidence of PLAP-1/Asporin expression in cartilage is somewhat conflicting.

Initial exploration of PLAP-1 at the organ level found it to be absent in mouse

articular cartilage31,32 but expressed in the perichondrium. Nakajima et al33 did show

expression of PLAP-1/Asporin, albeit minimally, in normal human articular cartilage.

However, in osteoarthritic cartilage, the expression of Asporin significantly

increases30,34. At a cellular level, Asporin is classified as an extracellular protein of

the LRR family group of proteins and hence should localise around cells in articular

cartilage.35 Indeed, Nakajima’s work33 suggested a cell surface localisation for

Asporin. However, Gruber et al.36 found positive Asporin staining to occur inside the

chondrocytes of rat articular cartilage and not in the extra-cellular matrix. Similar

results were found in the samples derived both from the spine of sand rat and humans.

They concluded Asporin to have a cytoplasmic localisation and also reported that

positive staining was not detected in the same cells of similar development. Whist

this study did not specifically investigate Asporin and cartilage, the observations of

the articular cartilage used as positive controls and the occasional nasal cartilage

agrees with the work of Gruber et al.36 in that PLAP-1/Asporin is found in cartilage

and located within the cytoplasm of the certain chondrocytes.

77

The periodontal ligament yielded diffuse staining of PLAP-1 which is likely due to

either background ‘noise’, cellularity of the region or its association with

proteoglycans in the extracellular matrix. In particular, fibroblasts which are the

principal cells of the PDL, are known to stain positive to PLAP-1.32 Ideally, the use

of monoclonal antibodies would have minimised background ‘noise’ and improve

specificity. However, at the time of writing, there was no commercially available anti-

PLAP-1 monoclonal antibody indicated for immunohistochemistry which is directed

at rat specimens.

Another possible cause for potential background staining was the thickness of the

sections. In this study, the sections were made at 7um but according to Dabbs37 the

ideal thickness is 3-5um for immunohistochemistry. It is believed that thicker

sections can increase background staining because it is difficult to block off the

endogenous peroxidase entirely thus reducing the signal-to-noise ratio especially

when a polyclonal antibody is being utilised.

The ambiguous results obtained make the results difficult to interpret. The cementum

third of the PDL which, under normal circumstances does not stain significantly with

PLAP-1, seems to have increased reactivity with PLAP-1 when osseous-like material

is formed. It would also appear that, albeit not statistically significant, most of the

other regions show a similar tendency. This would seem to suggest that PLAP-

1/Asporin is not the negative regulator of PDL mineralisation as Yamada and co-

workers19 suggested but rather it promotes osteoblast driven collagen mineralisation

as Kalamajski and colleagues38 found. However, as previously mentioned, ankylosis

is not necessarily permanent and resorptive activities including TRAP positive

multinucleated cells have been associated with the ankylotic bridge. Furthermore,

many of the positively stained areas aside from the ankylotic areas show resorptive

activity with multinucleated cells. However, the spicule nature of the ankylosis

renders it somewhat difficult to determine whether a resorptive bay was present and

indeed many of the positively stained cells alongside the cementum were not

associated with resorption lacunae. In addition, the differing results of Kalamajski38

and Yamada19 may be due to the source of cells used in their experiment; Yamada19

78

used cells derived specifically from PDL whereas Kalamajski38 did not.29 Further

investigations are required to clarify the role and function of PLAP-1 within the

periodontal ligament.

Conclusion

• The current study demonstrated that the periodontal ligament is not the only

region in which the PLAP-1 protein is located. PLAP-1/Asporin can be found

in nasal cartilage, blood vessel walls, periodontal ligament and the dental pulp.

• The data derived are inconclusive, although suggestive, of PLAP-1 being more

associated with mineralisation than maintenance of the periodontal ligament

space.

• The null hypothesis was rejected as significant differences were found,

particularly adjacent to the cementum, between the PLAP-1 expressions of

ankylotic regions compared to non-ankylotic areas.

Acknolwedgements

We thank the Australian Dental Research Foundation and the Australian Society of

Orthodontists Foundation for Research and Education for their generous support in

funding this project. Finally, we would like to thank Tom Sullivan from the

Department of Population Oral Health, University of Adelaide for his expert statistical

help.

Approval of the experimental procedures was provided by the Ethics Committee of

The University of Adelaide under ethics number M-01-2004.

There is no foreseeable conflict of interest

79

References

1. Andreasen JO, Andreasen FM, Andersson L. Textbook and color atlas of traumatic injuries to the teeth. Oxford, UK: Blackwell; 2007.

2. Berkovitz BKB, Shore R. Cells of the periodontal ligament. In: Berkovitz BKB, Moxham B, Newman H, editors. The Periodontal Ligament in Health and Disease. London: Mosby-Wolfe; 1995. p. 9-34.

3. Fujiyama K, Yamashiro T, Fukunaga T, Balam TA, Zheng L, Takano-Yamamoto T. Denervation resulting in dento-alveolar ankylosis associated with decreased Malassez epithelium. J Dent Res 2004;83:625-629.

4. Lindskog S, Blomlof L, Hammarström L. Evidence for a role of odontogenic epithelium in maintaining the periodontal space. J Clin Periodontol 1988;15:371-373.

5. Löe H, Waerhaug J. Experimental replantation of teeth in dogs and monkeys. Arch Oral Biol 1961;3:176-184.

6. Spouge JD. A new look at the rests of Malassez. A review of their embryological origin, anatomy, and possible role in periodontal health and disease. J Periodontol 1980;51:437-444.

7. Kirkham J, Brookes SJ, Shore RC, Bonass WA, Robinson C. The effect of glycosylaminoglycans on the mineralization of sheep periodontal ligament in vitro. Connect Tissue Res 1995;33:23-29.

8. Lekic P, McCulloch CA. Periodontal ligament cell population: the central role of fibroblasts in creating a unique tissue. Anat Rec 1996;245:327-341.

9. Melcher AH, Cheong T. Fibroblast-like cells depress formation of bone-like tissue in vitro. Journal of Dental Research 1988;67:290.

10. Ogiso B, Hughes FJ, Davies JE, McCulloch CA. Fibroblastic regulation of osteoblast function by prostaglandins. Cell Signal 1992;4:627-639.

11. Ogiso B, Hughes FJ, Melcher AH, McCulloch CA. Fibroblasts inhibit mineralised bone nodule formation by rat bone marrow stromal cells in vitro. J Cell Physiol 1991;146:442-450.

12. Ohno S, Doi T, Fujimoto K, Ijuin C, Tanaka N, Tanimoto K et al. RGD-CAP (betaig-h3) exerts a negative regulatory function on mineralization in the human periodontal ligament. J Dent Res 2002;81:822-825.

13. Li S, Yang PS, Cao JF, Ge SH, Pan KQ. [Expression of epidermal growth factor receptor in human periodontal ligament cells during their mineralization in vitro]. Hua Xi Kou Qiang Yi Xue Za Zhi 2006;24:11-14.

80

14. Duarte WR, Iimura T, Takenaga K, Ohya K, Ishikawa I, Kasugai S. Extracellular role of S100A4 calcium-binding protein in the periodontal ligament. Biochem Biophys Res Commun 1999;255:416-420.

15. Kato C, Kojima T, Komaki M, Mimori K, Duarte WR, Takenaga K et al. S100A4 inhibition by RNAi up-regulates osteoblast related genes in periodontal ligament cells. Biochem Biophys Res Commun 2005;326:147-153.

16. Yoshizawa T, Takizawa F, Iizawa F, Ishibashi O, Kawashima H, Matsuda A et al. Homeobox protein MSX2 acts as a molecular defense mechanism for preventing ossification in ligament fibroblasts. Mol Cell Biol 2004;24:3460-3472.

17. Komaki M, Karakida T, Abe M, Oida S, Mimori K, Iwasaki K et al. Twist negatively regulates osteoblastic differentiation in human periodontal ligament cells. J Cell Biochem 2007;100:303-314.

18. Tomoeda M, Yamada S, Shirai H, Ozawa Y, Yanagita M, Murakami S. PLAP-1/asporin inhibits activation of BMP receptor via its leucine-rich repeat motif. Biochem Biophys Res Commun 2008;371:191-196.

19. Yamada S, Tomoeda M, Ozawa Y, Yoneda S, Terashima Y, Ikezawa K et al. PLAP-1/asporin, a novel negative regulator of periodontal ligament mineralization. J Biol Chem 2007;282:23070-23080.

20. Chen WCW. An investigation into the role of osteoclasts and their precursors in an ankylosis model. B Sci Dent (Hons) Thesis. University of Adelaide, Adelaide; 2008.

21. Dreyer CW. Clast cell activity in a model of aseptic root resorption. PhD thesis, University of Adelaide; 2002.

22. Shaboodien SI. Traumatically induced dentoalveolar ankylosis in rats. D.Clin.Dent. thesis, University of Adelaide; 2005.

23. Curl L, Sampson, W. The presence of TNF-α and TNFR1 in aseptic root resorption. A preliminary study. Australian Orthodontic Journal 2011;27:102-109.

24. Di Iulio DS. Relationship of epithelial cells and nerve fibres to experimentally induced dentoalveolar ankylosis in the rat. D.Clin.Dent, University of Adelaide; 2007.

25. Dreyer CW, Pierce AM, Lindskog S. Hypothermic insult to the periodontium: a model for the study of aseptic tooth resorption. Endod Dent Traumatol 2000;16:9-15.

26. Andreasen JO, Skougaard MR. Reversibility of surgically induced dental ankylosis in rats. Int J Oral Surg 1972;1:98-102.

81

27. Andersson L. Dentoalveolar ankylosis and associated root resorption in replanted teeth. Experimental and clinical studies in monkeys and man. Swed Dent J Supplements 1988;56:1-75.

28. Blomlof L, Lindskog, S. Quality of periodontal healing. II: Dynamics of reparative cementum formation. Swed Dent J 1994;18:131-138.

29. Lee E-H, Park H-J., Jeong, J-H., Kim, Y-J., Cha, D-W., Kwon, D-K., Lee, S-H., Cho, J-Y. The role of Asporin in mineralization of human dental pulp stem cells. J. Cell. Physiol 2011;226:1676-2682.

30. Lorenzo P, Aspberg A, Onnerfjord P, Bayliss MT, Neame PJ, Heinegard D. Identification and characterization of asporin. a novel member of the leucine-rich repeat protein family closely related to decorin and biglycan. J Biol Chem 2001;276:12201-12211.

31. Henry SP, Takanosu M, Boyd TC, Mayne PM, Eberspaecher H, Zhou W et al. Expression pattern and gene characterization of asporin. a newly discovered member of the leucine-rich repeat protein family. J Biol Chem 2001;276:12212-12221.

32. Kou I, Nakajima, M., Ikegawa, S. Expression and regulation of the osteoarthritis-associated protein Asporin. The Journal of Biochemistry 2007;282:32193-32199.

33. Nakajima M, Kizawa H, Saitoh M, Kou I, Miyazono K, Ikegawa S. Mechanisms for asporin function and regulation in articular cartilage. J Biol Chem 2007;282:32185-32192.

34. Kizawa H, Kou, I., Iida, A., Sudo, A., Miyamoto, Y., Fukuda, A., Mabuchi, A., Kotani, A., Kawakami, A., Yamamoto, S., Uchida, A., Nakamura, K., Notoya, K., Nakamura, Y., Ikegawa, S. An aspartic acid repeat polymorphism in asporin inhibits chondrogenesis and increases susceptibility to osteoarthritis. Nature Genetics 2005;37:138-144.

35. Ikegawa S. Expression, regulation and function of asporin, a susceptibility gene in common bone and joint diseases. Curr Med Chem 2008;15:724-728.

36. Gruber HE, Ingram JA, Hoelscher GL, Zinchenko N, Hanley EN, Jr., Sun Y. Asporin, a susceptibility gene in osteoarthritis, is expressed at higher levels in the more degenerate human intervertebral disc. Arthritis Res Ther 2009;11:R47.

37. Dabbs DJ. Diagnostic immunohistochemistry. New York, Edinburgh: Churchill Livingstone; 2006.

38. Kalamajski S, Aspberg, A., Lindblom, K., Heinegard, D., Oldberg, A. Asporin competes with decorin for collagen binding, binds calcium and promotes osteoblast collagen mineralization. Biochem. J. 2009;423:53-59.

82

PAPER 2 (Prepared for submission in Archives of Oral Biology)

Title: Expression of Periodontal Ligament Associated Protein-1/Asporin in the Coronal Pulp Chamber of Rats Following Hypothermic Trauma. A Preliminary Report.

Authors: Chen, W.C., Sampson, W., Dreyer, C., Dharmapatni, K.

Affiliations:

Chen, W.C. – Postgraduate student, Orthodontic Unit, School of Dentistry, Faculty of

Health Sciences, University of Adelaide, South Australia.

Sampson, W. – Professor and P.R. Begg Chair, Orthodontic Unit, School of Dentistry,

Faculty of Health Sciences, University of Adelaide, South Australia.

Dreyer, C. – Associate Professor, Orthodontic Unit, School of Dentistry, Faculty of

Health Sciences, University of Adelaide, South Australia.

Dharmapatni, K. – Research Fellow, Hanson Institute, School of Medicine, Faculty of

Health Sciences, University of Adelaide, South Australia.

Abstract

Background: The dental pulp has several important functions including the formation

of the dentine, provision of nutrients, provision of neurosensory information as well

as the provision of protection via the formation of tertiary dentine. Recently, a

protein named PLAP-1/Asporin has been shown to be associated with the

differentiation and mineralisation of dental pulp stem cells. However, there are few

and conflicting reports regarding the role of PLAP-1/Asporin within mature dental

pulps. Furthermore, there are no reports regarding any association with the

formation of tertiary dentine.

83

Objectives: The aim of this investigation is to determine whether PLAP-1 protein is

expressed within the dentine-pulp complex and to determine its relationship with the

dentine-pulp complex following hypothermic trauma.

Materials & Methods: The maxillary right first molars of 30 Sprague-Dawley rats

were subjected to a single 20 minute application of dry ice to induce ankylosis.

Groups of five animals were sacrificed via cardiac perfusion after 7, 10, 14, 18, 21

and 28 days, respectively. The maxillae were dissected out and underwent routine

tissue fixation and processing. PLAP-1 expression was subsequently detected using

immunohistochemistry and analysed semi-quantitatively.

Results: PLAP-1/Asporin was found to be expressed exclusively within the pulp under

normal conditions and appeared to be associated with the odontoblastic and cell-rich

zone. Following trauma, PLAP-1/Asporin expression decreased marginally (not

statistically significant) alongside the dentine but increased significantly in the

central pulpal region where there was disruption and cellular breakdown.

Conclusions: PLAP-1/Asporin can be found within the mature dental pulp under

normal conditions and is particularly associated with the odontoblast layer and to a

lesser extent the cell rich zone. Under hypothermic trauma, PLAP-1/Asporin does not

appear to play a role in the formation of the tertiary dentine.

Key Words: periodontal ligament associated protein-1, asporin, dental pulp.

Introduction

The primary function of the dental pulp is the formation of dentine through the actions

of the odontoblasts. Other functions of the pulp include the provision of nutrients as

well as neurosensory information such as pain, pressure or temperature differences.

The dental pulp is also able to provide protection due to the formation of reparative

dentine following a traumatic episode.

84

The main cellular constituents are odontoblasts, fibroblasts, mesenchymal stem cells,

macrophages and lymphocytes. The extracellular components of the pulp are

composed of ground substance (eg. glycosaminoglycans) and collagen fibres (mainly

type I & III). Additionally, the pulp contains blood vessels, nerve fibres as well as

lymphatic vessels.1

Recently, a novel protein denoted Periodontal Ligament Associated Protein-1 (PLAP-

1/Asporin) has been found to be associated with the differentiation and mineralisation

of dental pulp stem cells.2 Furthermore, specific staining was found at the globular

calcific region in the junction of predentine and dentine. This finding was in

agreement with Kalamajski et al3 who also found Asporin to promote osteoblast-

driven collagen mineralization. However, contradictory results were reported by

Yamada et al4 who found PLAP-1 to be unique to the periodontal ligament as well as

being a negative regulator for PDL mineralisation.

PLAP-1 was originally characterised in 2001 and is considered a member of the

leucine-rich repeat proteoglycan family.5,6 It has been found to be similar to biglycan

and decorin (human Asporin is 50% identical and 70% similar to decorin and

biglycan)5. However, as PLAP-1 does not contain glycosaminoglycan attachment

sites6 and contains a unique sequence of aspartate residues, it is not considered a true

proteoglycan7. The normal structure of PLAP-1/Asporin contains a putative

propeptide, 4 aminoterminal cysteines, 10 leucine rich repeats and 2 C-terminal

cysteines.

Despite the pioneering work of Yamada et al4 and Lee et al2, there is a scarcity of

literature regarding asporin and its expression and association within the dental pulp.

Furthermore, there are no studies which report on the relationship of PLAP-1/Asporin

with the pulp and odontoblast-derived reparative dentine following a traumatic

episode.

85

In this article, the findings of PLAP-1 and its expression within the dental pulp under

normal as well as post-trauma conditions, is reported.

Aim

• To determine the normal expression of PLAP-1/Asporin within the pulp

chamber using immunohistochemistry.

• To investigate the relationship between PLAP-1/Asporin expression and the

coronal pulp following a hypothermic insult using immunohistochemistry.

Hypothesis

Null Hypothesis states that:

‘There is no change in the expression of PLAP-1 in traumatised pulp compared to

non-traumatised pulp.’

Materials & Methods

The materials utilised form the basis of another study which investigated the effect of

trauma on PLAP-1 within the periodontal ligament. The maxillary right first molars

of 30, 8-week-old Sprague-Dawley rats were subjected to a single 20 minute

application of dry ice which was considered adequate to induce ankylosis, while the

contralateral first molar served as a control. Anaesthesia was achieved via a 1:1

combination of Hypnorm® (Janssen-Cilag Ltd., Buckinghamshire, UK) and

Hypnovel® (Roche, Berne, Switzerland). Groups of five animals were sacrificed via

cardiac perfusion after 7, 10, 14, 18, 21 and 28 days, respectively. The maxillae were

dissected out and underwent routine tissue fixation and processing. The tissues were

embedded in paraffin wax and 7µm serial coronal sections were obtained, limited to

the furcation area.

86

The staining for PLAP-1 immunoreactivity was performed via the Labelled

StreptAvidin Biotin method. The primary antibody was an unlabelled rabbit

polyclonal antibody (450-31930, Sapphire Bioscience, NSW, Australia) diluted at 1 to

400 of stock concentration, incubated overnight in a wet chamber at room temperature

(RT). The linkage reagent was a biotinylated goat anti-rabbit secondary antibody

which is conjugated to horseradish peroxidase (K060911, LSAB®2-HRP, Dako,

Australia) and was incubated at 30 minutes, RT. Visualisation of the target protein

(PLAP-1) was through the use of AEC dye (K3469, Dako, Australia). Sections were

then counterstained with haematoxylin and lithium carbonate. Growing rat femoral

head containing cartilage was used as the positive controls as it has been previously

reported that chondrocytes express PLAP-1/Asporin.36 Sections incubated only with a

rabbit serum (N169987, Universal negative control, Dako, Australia) were used as the

negative control.

Stained slides were mounted and analysed subjectively using an Olympus B071

optical microscope with the images displayed by an attached personal computer with

a 24.0 inch monitor. The software programme Analysis (Olympus Soft Imaging

Solutions, Germany) was used for image processing and viewing. Various

magnifications including 4X, 10X and 20X were used to inspect the staining. Due to

the diffuse nature of the stain, a semi-quantitative scoring method was utilised (0=no

staining, 1=mild staining; 2=moderate staining, 3=intense staining) to measure the

activity of PLAP-1.

Fifteen sections 70µm apart per rat were stained and analysed. Wald statistics for

Type 3 GEE (General Estimating Equation) Analysis as well as Post-hoc comparisons

(see Appendices) were made. One section from each rat was randomly selected for

repeat scoring and the intra-observer reliability was calculated with a weighted Kappa

coefficient. All calculations performed by SAS Version 9.2 (SAS Institute Inc., Cary,

NC, USA).

Results

87

PLAP-1/Asporin was found to be positively expressed within the dental pulp. In

particular, PLAP-1 appears to be associated with the odontoblast layer as well as the

cell rich zone (Figure 1). There were no other regions in which the protein was

expressed within the crown of the rat molar (Figure 1).

Figure 1

a) PLAP-1 staining of the pulp chamber on the control side (10x magnification; 10 weeks old rat). Note the affiliation with the odontoblast/cell rich zone. b) 20 x magnification of the same section. c) Negative control of pulp staining displaying the lack of PLAP-1 staining (10 x magnification; 10 weeks old rat). d) Negative control of pulp staining displaying the lack of PLAP-1 staining (20 x magnification).

Unfortunately, the amount of trauma delivered may not have been equal amongst all

specimens as not all sections acquired ankylosis within the furcation region.

Moreover, some of the sections could not be included for analysis due to crucial

regions (especially the crown and pulp) being either lost or damaged during the

antigen retrieval process. In addition, it was found that despite identical staining

procedures and environmental conditions, there were great variations between animals

in the intensity of PLAP-1 expression.

a b

c d

dentine

dentine

pulp pulp

Note the increased affinity of PLAP-1 staining with the odontoblast/cell rich zone

pulp pulp

dentine dentine

200µm

100µm

200µm

100µm

88

Within the sections which did not display ankylosis, there were no consistent

statistically significant differences between control and experimental sides across the

various time points (results not shown but see page 113 in Appendices). This was

consistent for the area adjacent to the dentine and also in the central pulp. However,

there was still evidence of the freezing trauma due to the disorganisation and

breakdown of the pulpal cellular structures.

Figure 2

PLAP-1 pulp staining in control (a) and experimental side which did not achieve ankylosis (b). Note the lack of difference in staining intensity distribution. (Day 18 rat)

In the slides which did obtain ankylosis, the region adjacent to the dentine showed

statistically significant differences only when the staining intensity was grade 3.

However, the Odds Ratio consistently showed an increased likelihood of the control

side having more intense PLAP-1 reactivity than the experimental side (Table 1). The

region in the central part of the pulp showed consistent difference in that the

experimental pulp displayed statistically significantly more staining of PLAP-

1/Asporin than the control side (Table 2). A statistical analysis could not be

conducted for staining intensity of grade 3 due to a lack of sections with high intensity

staining within the central pulpal region.

a b

dentine

PDL

Alveolar bone

dentine

PDL

pulp pulp

Alveolar bone

200µm 200µm

89

Table 1. Comparison of PLAP-1 staining intensities within the pulp adjacent to the dentine in sections with ankylosis.

Model Contrast

P-value

Odds ratio

Lower 95% CI

Upper 95% CI

1. Predictors of intensity ≥ 1

Control vs. experimental

0.5450 1.473 0.420 5.159

2. Predictors of intensity ≥ 2

Control vs. experimental

0.2842 1.439 0.739 2.799

3. Predictors of intensity ≥ 3

Control vs. experimental

0.0095 1.533 1.110 2.116

Table 2. Comparison of PLAP-1 staining intensities within the central pulpal region in sections with ankylosis

Model Contrast

P-value

Odds ratio

Lower 95% CI

Upper 95% CI

1. Predictors of intensity ≥ 1

Control vs. experimental

<.0001 0.439 0.322 0.599

2. Predictors of intensity ≥ 2

Control vs. experimental

<.0001 0.102 0.063 0.166

3. Predictors of intensity ≥ 3

Control vs. experimental

Model did not converge

Figure 3

a) Increased PLAP-1 staining within the central pulpal region. b) The same section showing the largely homogenous PLAP-1 staining within the pulp chamber. Note the lack of obvious odontoblast/cell rich zones. (Day 18 rat)

b a

dentine

dentine

pulp

pulp Note the increase in staining in the central pulpal region

200µm

100µm

90

Regardless of whether ankylosis was achieved or not, there was evident formation of

tertiary dentine in the experimental side. This was true from day 10 onwards although

some sections from day 7 also displayed mild formation of tertiary dentine. Most of

the tertiary dentine exhibited regular structure with few cellular inclusions although

some sections displayed notable cellular inclusions (fig 4). Some of the cellular

inclusions stained positive for PLAP-1/Asoprin. No control side displayed the

formation of tertiary dentine.

Figure 4

Experimental section showing clear formation of tertiary dentine with obvious cellular inclusions. (Day 21 rat)

Comparisons between repeat measurements show a weighted Kappa coefficient of

0.95 (95% CI 0.92, 0.97), indicating very good agreement.

Discussion

The traumatic episode was generated by the application of a hypothermic insult as

part of another investigation.8 Whilst the aseptic injury from the dry ice application

maintains good tissue morphology and offers an insight to Asporin/PLAP-1

expression under trauma and reparative dentinogenesis, the results cannot be directly

translated to the classic process of reparative dentinogenesis caused by dental caries

which is due to bacterial insults.

dentine

Cellular inclusions

pulp

200µm

91

There is a lack of information regarding PLAP-1/Asporin and the dental pulp in the

dental literature. The result of this study is in contrast with Yamada et al4 who found

PLAP-1/Asporin to exist specifically within the periodontal ligament in the mouse

maxilla. The result also varies from that described by Lee et al2 who found PLAP-

1/Asporin to be found only in the globular calcific region at the junction of predentine

and dentine in post-natal mice. Differences in the experimental protocol likely

contributed to the differences. Park et al9 found the pulpal staining to be non-specific

and likely due to the antibody utilised or the high endogenous peroxidase activity.

However, in this study, there was obvious affinity of anti-PLAP-1/Asporin antibody

to the cell-rich zone of the pulp. This may be associated with fibroblasts and is in

agreement with Henry et al5 who also found prominent expression of PLAP-1/Asporin

associated with fibroblasts.

The increase in staining within the central region of the frozen pulp did not appear

histologically to relate to any particular cells or pattern. It appears that the increase in

staining is a result of the breakdown of cellular constituents within the pulp and their

structure thus leading to an increase in PLAP-1/Asporin staining within the centre and

a tendency to show decreased staining in the region adjacent to the dentine.

The lack of difference between the control and experimental sides in sections which

did not obtain ankylosis is likely due the lack of overall trauma delivered. The lack of

trauma possibly did not induce adequate damage to induce widespread cell death

within the pulp and periodontal ligament thus failing to obtain ankylosis as well as

significant cellular disruption within the dental pulp.

The formation of reparative dentine suggests that some odontoblasts survived the

hypothermic insult. However, judging by the morphological breakdown of cellular

structures within the pulp, it seems logical that a large number of odontoblasts would

not have survived but restoration was achieved through differentiation of

mesenchymal cells to odontoblasts. According to the findings of a previous study2,

PLAP-1/Asporin is expressed in the early stages of odontoblast differentiation and

92

may have a positive role in the minerlization of dental pulp stem cells under normal

developmental conditions. However, in this study, the lack of PLAP-1/Asporin

within the dentine, in addition to a decrease in staining intentisty adjacent to the

dentine, suggests that the protein is not involved in the mineralization of the tertiary

dentine. This suggests that although PLAP-1/Asporin may be involved in the

mineralization of dentine in normal development, it is not involved in the formation or

mineralization of tertiary dentine.

Conclusion

• PLAP-1/Asporin is positively expressed within the healthy dental pulp and is

particularly associated with the odontoblastic and cell-rich zones.

• PLAP-1/Asporin expression does appear to change (tendency to decrease in

peripheral regions and increase in central pulpal area) post hypothermic

trauma but is more likely to be due to the breakdown of the cellular structure.

• PLAP-1/Asporin is not directly associated with the formation of tertiary

dentine.

Acknolwedgements

We thank the Australian Dental Research Foundation and the Australian Society of

Orthodontists Foundation for Research and Education for their generous support in

funding this project. Also, we would like to thank Tom Sullivan from the Department

of Population Oral Health, University of Adelaide for his statistical advice.

Approval of the experimental procedures was provided by the Ethics Committee of

The University of Adelaide under ethics number M-01-2004.

There is no foreseeable conflict of interest

93

References

1. Ten Cate AR. Oral Histology: structure and function. St Louis: Mosby; 1989.

2. Lee E-H, Park H-J., Jeong, J-H., Kim, Y-J., Cha, D-W., Kwon, D-K., Lee, S-H., Cho, J-Y. The role of Asporin in mineralization of human dental pulp stem cells. J. Cell. Physiol 2011;226:1676-2682.

3. Kalamajski S, Aspberg, A., Lindblom, K., Heinegard, D., Oldberg, A. Asporin competes with decorin for collagen binding, binds calcium and promotes osteoblast collagen mineralization. Biochem. J. 2009;423:53-59.

4. Yamada S, Tomoeda M, Ozawa Y, Yoneda S, Terashima Y, Ikezawa K et al. PLAP-1/asporin, a novel negative regulator of periodontal ligament mineralization. J Biol Chem 2007;282:23070-23080.

5. Henry SP, Takanosu M, Boyd TC, Mayne PM, Eberspaecher H, Zhou W et al. Expression pattern and gene characterization of asporin. A newly discovered member of the leucine-rich repeat protein family. J Biol Chem 2001;276:12212-12221.

6. Lorenzo P, Aspberg A, Onnerfjord P, Bayliss MT, Neame PJ, Heinegard D. Identification and characterization of asporin. a novel member of the leucine-rich repeat protein family closely related to decorin and biglycan. J Biol Chem 2001;276:12201-12211.

7. Ikegawa S. Expression, regulation and function of asporin, a susceptibility gene in common bone and joint diseases. Curr Med Chem 2008;15:724-728.

8. Chen WCW. An Investigation into the Role of Periodontal Ligament Associated Protein-1 in the Maintenance of the Periodontal Ligament using an Ankylosis Model Adelaide, Doctor of Clinical Dentistry thesis, University of Adelaide; 2012.

9. Park ES, Cho, H.S., Kwon, T.G., Jang, S.N., Lee, S.N., An, C.H., Shin, H.I., Kim, J.Y., Cho, J.Y. Proteomics analysis of human dentin reveals distinct protein expression profiles. Journal of Proteome Research 2009;8:1338-1346.

94

Concluding Remarks The results from the first paper showed that PLAP-1 was expressed in the PDL, dental

pulp, blood vessel walls and the nasal cartilage. Not all sections obtained ankylosis.

Sections which did not obtain ankylosis demonstrated no significant PLAP-1

expression differences between control and experimental sides. Sections that did

obtain ankylosis yielded a tendency towards increased PLAP-1 reactivity especially

near the cementum. However, it was difficult to deduce whether the relationship of

PLAP-1 to the ankylotic union was associated with bone formation or resorptive

activities.

The results from paper two showed that PLAP-1/Asporin was expressed exclusively

within the pulp under normal conditions and appeared to be associated with the

odontoblastic and cell rich zone. Following trauma, PLAP-1/Asporin expression

decreased marginally (not statistically significant) alongside the dentine but increased

significantly in the central pulpal region possibly due to disruption and breakdown of

the cellular structures.

From the results derived, it can be concluded that PLAP-1/Asporin is indeed

expressed in several tissue/cell types and regions including the dental pulp and is not

exclusively associated with the periodontal ligament. In addition, PLAP-1 appears to

have a direct association with ankylosis although it is uncertain whether PLAP-1

exclusively facilitates bone mineralisation or resorption. The second null hypothesis

was also rejected although the change in expression of PLAP-1 within the pulp is

probably more morphological than physiological. Results from the study also suggest

that PLAP-1/Asporin does not appear to play a role in the formation of the tertiary

dentine.

Further research is required to elucidate the true role of PLAP-1 within the

periodontal ligament as well as the pulp.

95

APPENDICES Solutions

Anticoagulant - Heparin

Heparin Injection B.P. (containing no antiseptic) was supplied in 1ml plastic

ampoules (David Bull Laboratories, Mulgrave, Australia).

Contained 1000 units (IU) per 1ml.

Dosage: 0.02 ml of heparin sodium per 100 g of body weight

Route: Intravenous injection via femoral vein

Shelf: Discard unused heparin after vial seal is broken

Storage: Below 25º

Phosphate Buffer

Reagents:

Part A.

31.2g NaH2PO4.2H2O in 1 litre of distilled water (0.2M)

Part B.

28.39g Na2HPO4 in 1 litre distilled water (0.2M)

Procedure:

Mix 240ml of Part A and 760ml of Part B to make 1 litre

Phosphate Buffered Saline

Reagents:

Saline solution

8.79g NaCl in 1 L of distilled water (0.879%)

Part A phosphate buffer

Part B phosphate buffer

Procedure:

To 19ml of Part A and 81ml of Part B, add 100ml of saline solution to make

200ml with a pH of 7.4

Fixative : 4% paraformaldehyde / 0.1M phosphate buffer pH7.4

Reagents:

Paraformaldehyde

96

Sodium hydroxide

Distilled water

Preparation:

1) Heat up 1600ml of distilled water

2) Add 160g of paraformaldehyde

3) Add sodium hydroxide dropwise to clarify

4) Add 800ml of 0.4M sodium phosphate buffer pH 7.4

5) Add distilled water to make up 4000ml of solution in total.

Decalcifying agent

4% EDTA in phosphate buffer

Reagents:

Phosphate buffer - Parts A and B

EDTA - 80g

Procedure:

To 280ml of part A and 720ml of part B add 1litre of distilled water and

EDTA to give a pH of 7

Solutions for Immunostaining

Phosphate Buffered Saline

- Different to that used for tissue storage

- 5X PBS was made and then diluted when required

Reagents:

Na2HPO4 (anhydrous) 6.04g

NaH2PO4H20 3.93g

NaCl 45g

Mili Q water

Hydrochloric acid

Sodium hydroxide

Preparation:

1) Mix all reagents to dissolve

2) Adjust pH to 7.4 using either hydrochloric acid or sodium hydroxide (Bring

overall solution volume to 1L & store at room temperature)

97

Methanolic Hydrogen Peroxide Blocking Solution

- Courtesy of the Hanson Institute, Adelaide

- To be made on the day

Reagents & Preparation:

1) Obtain 250ml of absolute methanol and add

2) 4.15ml of 30% hydrogen peroxide

Tissue dehydration and paraffin embedding

The following automatic procedure was used for the impregnation of tissues with

paraffin wax prior to embedding, using a Shandon Citadel 2000 automatic processor

(Shandon Industries, Pittsburgh, Pennsylvania):

1. 70% ethyl alcohol 1 hour

2. 80% ethyl alcohol 3 hours

3. 90% ethyl alcohol 3 hours

4. 100% ethyl alcohol 4 hours

5. 100% ethyl alcohol 4 hours

6. 100% ethyl alcohol 4 hours

7. 100% histolene 4 hours

8. 100% histolene 5 hours

9. 100% histolene 5 hours

10. paraffin 7 hours

11. paraffin 7 hours

12. paraffin (under vacuum) 1 hour

Tissues were then further trimmed and embedded in Surgi Path® embedding media

using a Reichert Jung Histostat Embedding Centre.

Slide coating procedure

Slides were coated according to the following procedure using 3-

aminopropyltriethoxysilane (APT, Sigma code 36480).

1. Place slides in racks

2. Pre-rinse slides in 100% ethanol for 30 seconds, twice

3. Dip in 2% APT in ethanol for 2 minutes

4. Rinse in distilled water for 30 seconds, twice

5. Dry in 37° oven

98

Immunohistochemistry staining protocol

- Rat femur including articular cartilage as positive control

- Rabbit serum (N169987, Universal negative control, Dako, Australia) were

used as the negative control.

Day 1

1. Dewax all paraffin slides as usual

a. Histolene 1 for 10 minutes

b. Histolene 2 for 10 minutes

c. 95% ethanol for 5 minutes

d. 100% ethanol for 5 minutes

2. Then wash in Milli Q water for 2 x 5 minutes

3. Use pap pen to circle the sections

a. Use proteinase K 1/50 (100µg/ml)

b. Incubate in 37° D ro4 30 minutes

4. Wash in 1 x PBS 3 times for 5 minutes

5. Place methanolic hydrogen peroxide block on sections for 10 minutes

6. Wash in PBS 3 times for 5 minutes

7. Place normal horse serum (Vectastain, Australia) on slides for 60 minutes

8. Place polyclonal rabbit anti-Asporin primary antibody (LS-CS1930, Lifespan

Biosciences, Australia) at 1:400 dilution of stock concentration overnight in a

wet chamber at room temperature.

Day 2

9. Wash in PBS 3 times for 5 minutes

10. Place biotinylated goat anti-rabbit secondary antibody and incubate for 30

minutes (K060911, LSAB®2-HRP, Dako, Australia).

11. Wash in PBS 3 times for 5 minutes

12. Place streptavidin peroxidase (K060911, LSAB®2-HRP, Dako, Australia)

13. Wash in PBS 3 times for 5 minutes

14. Add AEC + substrate chromogen ( K346911, Dako, Australia) for 7 minutes

15. Wash with Mili Q water for 5 minutes

16. Counterstain

a. Hematoxylin for 10 seconds

99

b. Wash in tap water

c. Lithium carbonate for 30 seconds

d. Wash in tap water

17. Mount using Aquamount

Results

All calculations were performed using SAS Version 9.2 (SAS Institute Inc., Cary, NC, USA). 1. Analysis of PDL cementum intensity: no ankylosis in experimental side To compare PDL cementum intensity according to side of the mouth (experimental, control) and day of measurement, separate binary logistic generalised estimating equations were fitted to the data. In the models, side of the mouth, day and the interaction between side and day were included as predictor variables. Where the interaction term was not statistically significant, a second model excluding this term was fitted. Note that these models were chosen because: a) it was not reasonable to treat the outcome as being normally distributed with just 4 ordinal levels (hence the use of logistic models instead of ordinary linear regression models). b) a proportional odds model was explored but the proportionality assumption failed (hence used separate binary logistic regression models instead of a single proportional odds model). c) results from within the same rat were expected to be correlated, hence the use of generalised estimating equations instead of ordinary regression models. Model 1: predictors of intensity ≥ 1 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 145.54 <.0001

side 1 0.57 0.4484

day*side 4 1851.91 <.0001

The table 'Wald Statistics For Type 3 GEE Analysis' shows the significance of predictor variables in the model. Since the model includes an interaction effect, this is the only term that needs to be interpreted. The highly significant interaction effect (p < 0.0001) suggests that the odds of having an intensity reading ≥ 1 depended on both side of mouth and time. That is, the difference between the two sides of the mouth changed over time (or alternatively that changes over time differed according to side of the mouth).

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.1238 1.498 0.895 2.504

7 control 10 control 0.0173 1.923 1.123 3.293

100

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 10 experimental 0.0036 2.714 1.385 5.320

7 control 14 control 0.6779 1.192 0.521 2.725

7 control 14 experimental 0.0329 2.286 1.069 4.886

7 control 18 control 0.0914 2.036 0.892 4.647

7 control 18 experimental 0.2730 1.696 0.659 4.365

7 control 21 control <.0001 0.784 0.713 0.863

7 control 21 experimental 0.4117 0.711 0.315 1.606

7 control 28 control 0.2282 0.528 0.187 1.492

7 control 28 experimental 0.1778 0.465 0.153 1.416

7 experimental 10 control 0.4796 1.284 0.642 2.567

7 experimental 10 experimental 0.1954 1.813 0.737 4.459

7 experimental 14 control 0.7326 0.796 0.215 2.951

7 experimental 14 experimental 0.4361 1.526 0.527 4.424

7 experimental 18 control 0.5925 1.359 0.442 4.185

7 experimental 18 experimental 0.8302 1.133 0.362 3.542

7 experimental 21 control 0.0095 0.524 0.321 0.854

7 experimental 21 experimental 0.0207 0.475 0.253 0.892

7 experimental 28 control 0.1431 0.352 0.087 1.423

7 experimental 28 experimental 0.0588 0.311 0.092 1.045

10 control 10 experimental 0.2767 1.412 0.758 2.628

10 control 14 control 0.3495 0.620 0.228 1.688

10 control 14 experimental 0.5544 1.189 0.670 2.109

10 control 18 control 0.8440 1.059 0.599 1.871

10 control 18 experimental 0.7724 0.882 0.378 2.060

10 control 21 control 0.0002 0.408 0.256 0.651

10 control 21 experimental 0.0198 0.370 0.160 0.854

10 control 28 control 0.0249 0.275 0.089 0.849

10 control 28 experimental 0.0005 0.242 0.109 0.538

10 experimental 14 control 0.1124 0.439 0.159 1.213

10 experimental 14 experimental 0.7136 0.842 0.336 2.109

10 experimental 18 control 0.3978 0.750 0.385 1.461

10 experimental 18 experimental 0.1666 0.625 0.321 1.217

10 experimental 21 control <.0001 0.289 0.159 0.526

10 experimental 21 experimental 0.0068 0.262 0.099 0.691

10 experimental 28 control <.0001 0.194 0.101 0.373

10 experimental 28 experimental <.0001 0.171 0.087 0.339

14 control 14 experimental 0.0597 1.918 0.974 3.779

101

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

14 control 18 control 0.1677 1.708 0.798 3.656

14 control 18 experimental 0.4738 1.424 0.542 3.742

14 control 21 control 0.3376 0.658 0.280 1.548

14 control 21 experimental 0.5233 0.597 0.122 2.914

14 control 28 control 0.0667 0.443 0.185 1.058

14 control 28 experimental 0.2188 0.390 0.087 1.748

14 experimental 18 control 0.5590 0.891 0.604 1.314

14 experimental 18 experimental 0.4463 0.742 0.345 1.599

14 experimental 21 control 0.0043 0.343 0.165 0.715

14 experimental 21 experimental 0.0950 0.311 0.079 1.225

14 experimental 28 control 0.0092 0.231 0.077 0.696

14 experimental 28 experimental 0.0121 0.204 0.059 0.705

18 control 18 experimental 0.4792 0.833 0.503 1.381

18 control 21 control 0.0151 0.385 0.178 0.832

18 control 21 experimental 0.1291 0.349 0.090 1.359

18 control 28 control 0.0015 0.259 0.113 0.595

18 control 28 experimental 0.0042 0.229 0.083 0.628

18 experimental 21 control 0.0864 0.462 0.191 1.117

18 experimental 21 experimental 0.2491 0.419 0.095 1.839

18 experimental 28 control 0.0009 0.311 0.156 0.620

18 experimental 28 experimental 0.0336 0.274 0.083 0.904

21 control 21 experimental 0.8065 0.907 0.414 1.987

21 control 28 control 0.4410 0.673 0.246 1.843

21 control 28 experimental 0.3246 0.593 0.210 1.676

21 experimental 28 control 0.7099 0.742 0.155 3.565

21 experimental 28 experimental 0.3771 0.655 0.256 1.676

28 control 28 experimental 0.8390 0.882 0.261 2.972

Model 2: predictors of intensity ≥ 2 Same as above, but modelling the odds of the intensity score being ≥ 2. Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 13123.1 <.0001

side 1 0.00 0.9493

day*side 4 5.92 0.2050

102

Since the interaction effect was not statistically significant, there is no evidence to suggest that the difference between sides of the mouth depended on day of measurement. To be able to interpret the main effects of day and side, a second model excluding the interaction term was fitted. Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 1206.51 <.0001

side 1 0.00 0.9857

The model indicates there were no differences between the two sides, while there were significant changes over time (p < 0.0001).

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 10 0.0076 6.377 1.636 24.855

7 14 0.0129 3.155 1.276 7.801

7 18 0.1348 5.281 0.596 46.776

7 21 0.0261 1.877 1.078 3.270

7 28 <.0001 7.919 4.741 13.227

10 14 0.2044 0.495 0.167 1.467

10 18 0.8747 0.828 0.080 8.625

10 21 0.0614 0.294 0.082 1.060

10 28 0.7366 1.242 0.352 4.386

14 18 0.4895 1.674 0.388 7.219

14 21 0.1943 0.595 0.272 1.303

14 28 <.0001 2.510 1.602 3.933

18 21 0.3107 0.355 0.048 2.626

18 28 0.6401 1.499 0.274 8.193

21 28 <.0001 4.219 2.273 7.829

control experimental 0.9857 1.005 0.571 1.768

Model 3: predictors of intensity ≥ 3 Model did not converge (due to small number of observations that took the value 3). 2. Analysis of PDL bone intensity: no ankylosis in experimental side Model 1: predictors of intensity ≥ 1 Model did not converge. Model 2: predictors of intensity ≥ 2 Wald Statistics For Type 3 GEE Analysis

103

Source DF Chi-Square P value

day 4 75.86 <.0001

side 1 30.53 <.0001

day*side 4 25.84 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.2933 1.687 0.636 4.475

7 control 10 control 0.5128 0.765 0.344 1.705

7 control 10 experimental 0.1566 2.214 0.737 6.651

7 control 14 control 0.0322 1.246 1.019 1.523

7 control 14 experimental 0.0016 5.905 1.960 17.786

7 control 18 control 0.2995 1.703 0.623 4.658

7 control 18 experimental 0.1701 1.910 0.758 4.817

7 control 21 control 0.1582 3.030 0.650 14.130

7 control 21 experimental 0.0061 5.636 1.639 19.379

7 control 28 control 0.4258 1.546 0.529 4.522

7 control 28 experimental <.0001 2.847 1.863 4.351

7 experimental 10 control 0.0129 0.454 0.243 0.846

7 experimental 10 experimental 0.6323 1.312 0.431 3.999

7 experimental 14 control 0.5539 0.738 0.270 2.016

7 experimental 14 experimental 0.0123 3.500 1.313 9.331

7 experimental 18 control 0.9887 1.010 0.268 3.802

7 experimental 18 experimental 0.8478 1.132 0.318 4.028

7 experimental 21 control 0.5815 1.796 0.224 14.412

7 experimental 21 experimental 0.1683 3.341 0.601 18.585

7 experimental 28 control 0.8967 0.917 0.247 3.409

7 experimental 28 experimental 0.3315 1.688 0.587 4.851

10 control 10 experimental 0.0856 2.893 0.862 9.713

10 control 14 control 0.2623 1.627 0.695 3.812

10 control 14 experimental <.0001 7.714 3.224 18.459

10 control 18 control 0.2306 2.225 0.602 8.228

10 control 18 experimental 0.1058 2.496 0.824 7.560

10 control 21 control 0.1769 3.959 0.537 29.158

10 control 21 experimental 0.0192 7.364 1.385 39.151

10 control 28 control 0.3089 2.020 0.521 7.830

10 control 28 experimental 0.0065 3.719 1.443 9.584

10 experimental 14 control 0.2271 0.563 0.221 1.431

10 experimental 14 experimental 0.0008 2.667 1.501 4.737

10 experimental 18 control 0.4968 0.769 0.361 1.640

104

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

10 experimental 18 experimental 0.8239 0.863 0.235 3.168

10 experimental 21 control 0.6134 1.368 0.405 4.620

10 experimental 21 experimental 0.0381 2.545 1.053 6.155

10 experimental 28 control 0.0199 0.698 0.516 0.945

10 experimental 28 experimental 0.5022 1.286 0.617 2.679

14 control 14 experimental 0.0020 4.741 1.763 12.746

14 control 18 control 0.4959 1.368 0.555 3.367

14 control 18 experimental 0.3838 1.534 0.586 4.016

14 control 21 control 0.2128 2.433 0.601 9.851

14 control 21 experimental 0.0065 4.525 1.525 13.430

14 control 28 control 0.6342 1.242 0.509 3.028

14 control 28 experimental <.0001 2.286 1.809 2.889

14 experimental 18 control 0.0054 0.288 0.120 0.692

14 experimental 18 experimental 0.0540 0.324 0.103 1.020

14 experimental 21 control 0.3714 0.513 0.119 2.216

14 experimental 21 experimental 0.9371 0.955 0.301 3.029

14 experimental 28 control 0.0006 0.262 0.122 0.560

14 experimental 28 experimental 0.0969 0.482 0.204 1.141

18 control 18 experimental 0.7555 1.122 0.545 2.309

18 control 21 control 0.1670 1.779 0.786 4.027

18 control 21 experimental <.0001 3.309 2.022 5.415

18 control 28 control 0.7603 0.908 0.488 1.689

18 control 28 experimental 0.1690 1.671 0.804 3.475

18 experimental 21 control 0.5083 1.586 0.404 6.223

18 experimental 21 experimental 0.0595 2.950 0.958 9.091

18 experimental 28 control 0.7404 0.810 0.232 2.825

18 experimental 28 experimental 0.4162 1.490 0.570 3.899

21 control 21 experimental 0.0011 1.860 1.282 2.699

21 control 28 control 0.1598 0.510 0.200 1.304

21 control 28 experimental 0.9195 0.940 0.281 3.146

21 experimental 28 control <.0001 0.274 0.149 0.505

21 experimental 28 experimental 0.1304 0.505 0.208 1.224

28 control 28 experimental 0.0707 1.841 0.950 3.568

Model 3: predictors of intensity ≥ 3 Model did not converge.

105

3. Analysis of PDL mid intensity: no ankylosis in experimental side Model 1: predictors of intensity ≥ 1 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 21.37 0.0003

side 1 13.27 0.0003

day*side 4 1157.47 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.0036 3.185 1.461 6.943

7 control 10 control 0.0904 3.058 0.839 11.149

7 control 10 experimental 0.0004 5.609 2.173 14.475

7 control 14 control 0.5876 1.486 0.355 6.227

7 control 14 experimental 0.0002 3.185 1.731 5.863

7 control 18 control 0.0328 0.573 0.344 0.956

7 control 18 experimental 0.2327 1.877 0.667 5.280

7 control 21 control 0.9934 0.995 0.332 2.982

7 control 21 experimental 0.7613 1.251 0.294 5.318

7 control 28 control 0.0049 0.233 0.085 0.642

7 control 28 experimental 0.2834 0.377 0.064 2.240

7 experimental 10 control 0.9423 0.960 0.318 2.898

7 experimental 10 experimental 0.0646 1.761 0.966 3.209

7 experimental 14 control 0.1900 0.467 0.149 1.459

7 experimental 14 experimental 1.0000 1.000 0.403 2.482

7 experimental 18 control <.0001 0.180 0.088 0.370

7 experimental 18 experimental 0.4232 0.589 0.162 2.150

7 experimental 21 control 0.0288 0.313 0.110 0.887

7 experimental 21 experimental 0.0388 0.393 0.162 0.953

7 experimental 28 control <.0001 0.073 0.025 0.218

7 experimental 28 experimental 0.0024 0.118 0.030 0.470

10 control 10 experimental 0.0369 1.834 1.038 3.243

10 control 14 control 0.0004 0.486 0.326 0.724

10 control 14 experimental 0.9514 1.042 0.280 3.871

10 control 18 control 0.0004 0.188 0.074 0.475

10 control 18 experimental 0.4109 0.614 0.192 1.965

10 control 21 control <.0001 0.326 0.206 0.513

106

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

10 control 21 experimental 0.0091 0.409 0.209 0.801

10 control 28 control 0.0058 0.076 0.012 0.474

10 control 28 experimental 0.0100 0.123 0.025 0.607

10 experimental 14 control <.0001 0.265 0.150 0.469

10 experimental 14 experimental 0.2157 0.568 0.232 1.391

10 experimental 18 control <.0001 0.102 0.051 0.204

10 experimental 18 experimental 0.0706 0.335 0.102 1.096

10 experimental 21 control <.0001 0.177 0.102 0.309

10 experimental 21 experimental <.0001 0.223 0.129 0.385

10 experimental 28 control <.0001 0.042 0.011 0.161

10 experimental 28 experimental 0.0001 0.067 0.017 0.265

14 control 14 experimental 0.2563 2.143 0.575 7.988

14 control 18 control 0.1055 0.386 0.122 1.223

14 control 18 experimental 0.7602 1.263 0.282 5.649

14 control 21 control 0.1047 0.670 0.413 1.087

14 control 21 experimental 0.5760 0.842 0.460 1.539

14 control 28 control 0.0525 0.157 0.024 1.021

14 control 28 experimental 0.1060 0.254 0.048 1.338

14 experimental 18 control <.0001 0.180 0.082 0.395

14 experimental 18 experimental 0.4281 0.589 0.159 2.179

14 experimental 21 control 0.0359 0.313 0.105 0.926

14 experimental 21 experimental 0.1907 0.393 0.097 1.592

14 experimental 28 control <.0001 0.073 0.031 0.170

14 experimental 28 experimental 0.0130 0.118 0.022 0.638

18 control 18 experimental 0.0003 3.274 1.731 6.190

18 control 21 control 0.2349 1.736 0.699 4.314

18 control 21 experimental 0.1690 2.183 0.718 6.638

18 control 28 control 0.1095 0.407 0.135 1.224

18 control 28 experimental 0.5639 0.658 0.159 2.727

18 experimental 21 control 0.3396 0.530 0.144 1.949

18 experimental 21 experimental 0.5858 0.667 0.155 2.865

18 experimental 28 control 0.0051 0.124 0.029 0.535

18 experimental 28 experimental 0.0354 0.201 0.045 0.896

21 control 21 experimental 0.6198 1.257 0.509 3.105

21 control 28 control 0.1063 0.234 0.040 1.363

21 control 28 experimental 0.3052 0.379 0.059 2.422

21 experimental 28 control 0.0511 0.186 0.034 1.008

107

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

21 experimental 28 experimental 0.0548 0.301 0.089 1.025

28 control 28 experimental 0.4909 1.618 0.411 6.368

Model 2: predictors of intensity ≥ 2 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 262.31 <.0001

side 1 3.85 0.0496

day*side 4 36.87 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.7013 0.854 0.382 1.911

7 control 10 control 0.2749 1.625 0.680 3.885

7 control 10 experimental 0.1934 4.083 0.490 34.021

7 control 14 control 0.8472 0.900 0.308 2.628

7 control 14 experimental 0.7387 1.375 0.212 8.930

7 control 18 control 0.0007 0.250 0.113 0.555

7 control 18 experimental 0.6274 0.729 0.204 2.610

7 control 21 control 0.0303 0.400 0.175 0.916

7 control 21 experimental 0.1691 4.333 0.536 35.035

7 control 28 control 0.0002 0.514 0.360 0.734

7 control 28 experimental 0.0595 0.405 0.158 1.037

7 experimental 10 control 0.0541 1.902 0.989 3.660

7 experimental 10 experimental 0.2416 4.780 0.348 65.586

7 experimental 14 control 0.9279 1.054 0.339 3.271

7 experimental 14 experimental 0.5784 1.610 0.300 8.629

7 experimental 18 control 0.0593 0.293 0.082 1.049

7 experimental 18 experimental 0.8557 0.854 0.155 4.696

7 experimental 21 control 0.0011 0.468 0.297 0.739

7 experimental 21 experimental 0.0517 5.073 0.988 26.041

7 experimental 28 control 0.3498 0.602 0.207 1.745

7 experimental 28 experimental 0.2626 0.474 0.128 1.750

10 control 10 experimental 0.4582 2.513 0.220 28.665

10 control 14 control 0.3160 0.554 0.175 1.758

108

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

10 control 14 experimental 0.8576 0.846 0.137 5.244

10 control 18 control 0.0002 0.154 0.057 0.412

10 control 18 experimental 0.2386 0.449 0.118 1.701

10 control 21 control <.0001 0.246 0.142 0.426

10 control 21 experimental 0.3831 2.667 0.294 24.160

10 control 28 control 0.0181 0.316 0.122 0.822

10 control 28 experimental 0.0280 0.249 0.072 0.861

10 experimental 14 control 0.2229 0.220 0.019 2.509

10 experimental 14 experimental 0.4167 0.337 0.024 4.657

10 experimental 18 control 0.0006 0.061 0.012 0.303

10 experimental 18 experimental 0.0094 0.179 0.049 0.656

10 experimental 21 control 0.0492 0.098 0.010 0.992

10 experimental 21 experimental 0.9709 1.061 0.043 25.932

10 experimental 28 control 0.0259 0.126 0.020 0.780

10 experimental 28 experimental 0.0007 0.099 0.026 0.376

14 control 14 experimental 0.4216 1.528 0.543 4.295

14 control 18 control 0.0160 0.278 0.098 0.787

14 control 18 experimental 0.7681 0.810 0.200 3.282

14 control 21 control 0.0814 0.444 0.179 1.107

14 control 21 experimental 0.1296 4.815 0.631 36.744

14 control 28 control 0.3697 0.571 0.168 1.943

14 control 28 experimental 0.2930 0.450 0.101 1.994

14 experimental 18 control 0.0514 0.182 0.033 1.010

14 experimental 18 experimental 0.5120 0.530 0.080 3.531

14 experimental 21 control 0.0803 0.291 0.073 1.161

14 experimental 21 experimental 0.1900 3.152 0.566 17.542

14 experimental 28 control 0.3328 0.374 0.051 2.739

14 experimental 28 experimental 0.2026 0.294 0.045 1.931

18 control 18 experimental <.0001 2.917 1.781 4.775

18 control 21 control 0.3474 1.600 0.600 4.265

18 control 21 experimental 0.0224 17.333 1.497 200.726

18 control 28 control 0.0160 2.056 1.143 3.695

18 control 28 experimental 0.2046 1.619 0.769 3.409

18 experimental 21 control 0.3844 0.549 0.142 2.122

18 experimental 21 experimental 0.2081 5.943 0.371 95.286

18 experimental 28 control 0.4969 0.705 0.257 1.934

18 experimental 28 experimental 0.1987 0.555 0.226 1.362

109

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

21 control 21 experimental 0.0063 10.833 1.959 59.896

21 control 28 control 0.6165 1.285 0.482 3.425

21 control 28 experimental 0.9827 1.012 0.347 2.954

21 experimental 28 control 0.0715 0.119 0.012 1.206

21 experimental 28 experimental 0.0307 0.093 0.011 0.802

28 control 28 experimental 0.5261 0.788 0.377 1.647

Model 3: predictors of intensity ≥ 3 Model did not converge. 4. Analysis of PDL BV intensity: no ankylosis in experimental side Model 1: predictors of intensity ≥ 1 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 68.72 <.0001

side 1 63.77 <.0001

day*side 4 84.13 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental <.0001 2.982 2.398 3.708

7 control 10 control 0.6821 1.202 0.498 2.904

7 control 10 experimental 0.8373 1.177 0.248 5.576

7 control 14 control 0.5582 0.671 0.176 2.552

7 control 14 experimental 0.1999 2.485 0.618 9.994

7 control 18 control 0.5992 0.621 0.105 3.668

7 control 18 experimental 0.5766 1.464 0.384 5.586

7 control 21 control 0.5109 0.512 0.069 3.772

7 control 21 experimental 0.8166 0.867 0.259 2.900

7 control 28 control 0.0451 0.089 0.008 0.949

7 control 28 experimental 0.3127 0.403 0.069 2.353

7 experimental 10 control 0.0851 0.403 0.143 1.134

7 experimental 10 experimental 0.2536 0.395 0.080 1.947

7 experimental 14 control 0.0425 0.225 0.053 0.951

7 experimental 14 experimental 0.8093 0.833 0.189 3.665

7 experimental 18 control 0.0845 0.208 0.035 1.238

7 experimental 18 experimental 0.3308 0.491 0.117 2.059

110

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 experimental 21 control 0.1084 0.172 0.020 1.476

7 experimental 21 experimental 0.0868 0.291 0.071 1.196

7 experimental 28 control 0.0054 0.030 0.003 0.354

7 experimental 28 experimental 0.0449 0.135 0.019 0.955

10 control 10 experimental 0.9757 0.979 0.249 3.844

10 control 14 control 0.5294 0.558 0.091 3.438

10 control 14 experimental 0.3956 2.067 0.387 11.033

10 control 18 control 0.5434 0.517 0.061 4.347

10 control 18 experimental 0.7357 1.218 0.388 3.824

10 control 21 control 0.4595 0.425 0.044 4.094

10 control 21 experimental 0.5935 0.721 0.217 2.397

10 control 28 control 0.0090 0.074 0.010 0.522

10 control 28 experimental 0.2189 0.335 0.059 1.915

10 experimental 14 control 0.6156 0.570 0.064 5.115

10 experimental 14 experimental 0.3422 2.111 0.452 9.865

10 experimental 18 control 0.5149 0.528 0.077 3.612

10 experimental 18 experimental 0.8064 1.244 0.217 7.132

10 experimental 21 control 0.5670 0.435 0.025 7.535

10 experimental 21 experimental 0.7795 0.736 0.086 6.270

10 experimental 28 control <.0001 0.075 0.024 0.234

10 experimental 28 experimental 0.3512 0.342 0.036 3.259

14 control 14 experimental 0.0031 3.704 1.554 8.828

14 control 18 control 0.9119 0.926 0.237 3.622

14 control 18 experimental 0.3506 2.183 0.424 11.238

14 control 21 control 0.5691 0.763 0.300 1.939

14 control 21 experimental 0.6580 1.292 0.416 4.016

14 control 28 control 0.1202 0.132 0.010 1.697

14 control 28 experimental 0.3355 0.601 0.213 1.695

14 experimental 18 control 0.0020 0.250 0.104 0.603

14 experimental 18 experimental 0.4625 0.589 0.144 2.415

14 experimental 21 control 0.0538 0.206 0.041 1.026

14 experimental 21 experimental 0.1686 0.349 0.078 1.562

14 experimental 28 control 0.0003 0.036 0.006 0.216

14 experimental 28 experimental 0.0041 0.162 0.047 0.561

18 control 18 experimental 0.2455 2.357 0.554 10.021

18 control 21 control 0.8607 0.824 0.094 7.198

18 control 21 experimental 0.7601 1.395 0.164 11.836

111

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

18 control 28 control 0.0820 0.143 0.016 1.280

18 control 28 experimental 0.6600 0.649 0.094 4.464

18 experimental 21 control 0.3425 0.349 0.040 3.064

18 experimental 21 experimental 0.4963 0.592 0.131 2.681

18 experimental 28 control 0.0063 0.061 0.008 0.453

18 experimental 28 experimental 0.1314 0.275 0.051 1.471

21 control 21 experimental 0.3688 1.694 0.537 5.351

21 control 28 control 0.2513 0.173 0.009 3.459

21 control 28 experimental 0.5782 0.788 0.340 1.827

21 experimental 28 control 0.0688 0.102 0.009 1.192

21 experimental 28 experimental 0.0789 0.465 0.198 1.093

28 control 28 experimental 0.1834 4.541 0.489 42.182

Model 2: predictors of intensity ≥ 2 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 3.28 0.5128

side 1 11.85 0.0006

day*side 4 147.68 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental <.0001 3.964 2.591 6.066

7 control 10 control 0.7964 1.095 0.549 2.186

7 control 10 experimental 0.1209 2.204 0.812 5.984

7 control 14 control 0.4472 1.442 0.561 3.702

7 control 14 experimental 0.0060 6.643 1.721 25.646

7 control 18 control 0.5430 1.429 0.453 4.508

7 control 18 experimental 0.1314 1.929 0.822 4.527

7 control 21 control 0.1689 2.457 0.683 8.845

7 control 21 experimental <.0001 5.633 2.916 10.881

7 control 28 control 0.6251 1.311 0.443 3.883

7 control 28 experimental 0.1041 2.071 0.861 4.985

7 experimental 10 control 0.0175 0.276 0.096 0.799

7 experimental 10 experimental 0.1905 0.556 0.231 1.339

7 experimental 14 control 0.0502 0.364 0.132 1.001

112

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 experimental 14 experimental 0.3370 1.676 0.584 4.807

7 experimental 18 control 0.0891 0.360 0.111 1.169

7 experimental 18 experimental 0.1705 0.486 0.174 1.363

7 experimental 21 control 0.4567 0.620 0.176 2.185

7 experimental 21 experimental 0.2738 1.421 0.757 2.665

7 experimental 28 control 0.0473 0.331 0.111 0.987

7 experimental 28 experimental 0.2630 0.523 0.168 1.628

10 control 10 experimental 0.2422 2.012 0.623 6.497

10 control 14 control 0.6544 1.316 0.395 4.383

10 control 14 experimental 0.0554 6.065 0.959 38.349

10 control 18 control 0.6832 1.304 0.364 4.673

10 control 18 experimental 0.1693 1.761 0.786 3.946

10 control 21 control 0.2772 2.243 0.522 9.635

10 control 21 experimental 0.0026 5.143 1.769 14.948

10 control 28 control 0.7618 1.197 0.374 3.827

10 control 28 experimental 0.0951 1.891 0.895 3.997

10 experimental 14 control 0.4403 0.654 0.222 1.923

10 experimental 14 experimental 0.0322 3.014 1.098 8.272

10 experimental 18 control 0.3122 0.648 0.280 1.503

10 experimental 18 experimental 0.8078 0.875 0.298 2.565

10 experimental 21 control 0.8225 1.115 0.431 2.882

10 experimental 21 experimental 0.0666 2.556 0.938 6.965

10 experimental 28 control 0.0522 0.595 0.352 1.005

10 experimental 28 experimental 0.9055 0.940 0.337 2.618

14 control 14 experimental 0.0257 4.608 1.203 17.647

14 control 18 control 0.9706 0.991 0.612 1.604

14 control 18 experimental 0.5092 1.338 0.564 3.175

14 control 21 control 0.0426 1.705 1.018 2.854

14 control 21 experimental 0.0455 3.907 1.028 14.857

14 control 28 control 0.8132 0.909 0.414 1.999

14 control 28 experimental 0.3959 1.437 0.622 3.318

14 experimental 18 control 0.0172 0.215 0.061 0.761

14 experimental 18 experimental 0.1771 0.290 0.048 1.749

14 experimental 21 control 0.1372 0.370 0.100 1.373

14 experimental 21 experimental 0.7614 0.848 0.292 2.460

14 experimental 28 control 0.0048 0.197 0.064 0.610

14 experimental 28 experimental 0.1264 0.312 0.070 1.389

113

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

18 control 18 experimental 0.5500 1.350 0.505 3.611

18 control 21 control <.0001 1.720 1.343 2.204

18 control 21 experimental 0.0541 3.943 0.976 15.924

18 control 28 control 0.6837 0.918 0.607 1.387

18 control 28 experimental 0.3599 1.450 0.655 3.212

18 experimental 21 control 0.6463 1.274 0.453 3.584

18 experimental 21 experimental 0.1396 2.921 0.705 12.106

18 experimental 28 control 0.4729 0.680 0.237 1.951

18 experimental 28 experimental 0.8930 1.074 0.379 3.042

21 control 21 experimental 0.2972 2.292 0.482 10.906

21 control 28 control 0.0412 0.534 0.292 0.975

21 control 28 experimental 0.7455 0.843 0.301 2.364

21 experimental 28 control 0.0158 0.233 0.071 0.760

21 experimental 28 experimental 0.0646 0.368 0.127 1.063

28 control 28 experimental 0.2123 1.580 0.770 3.243

Model 3: predictors of intensity ≥ 3 Model did not converge. 5. Analysis of Pulp Dentine intensity: no ankylosis in experimental side Model 1: predictors of intensity ≥ 1 Model did not converge. Model 2: predictors of intensity ≥ 2 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 14.00 0.0073

side 1 1.40 0.2367

day*side 4 113.83 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.0092 4.655 1.463 14.814

7 control 10 control 0.0003 5.727 2.220 14.773

7 control 10 experimental 0.0016 4.875 1.819 13.066

7 control 14 control 0.0209 3.073 1.185 7.967

7 control 14 experimental <.0001 11.077 3.606 34.026

114

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 18 control 0.0055 5.276 1.632 17.060

7 control 18 experimental 0.0002 3.750 1.890 7.442

7 control 21 control 0.1352 1.884 0.821 4.323

7 control 21 experimental 0.0376 2.025 1.041 3.938

7 control 28 control 0.0152 5.000 1.364 18.326

7 control 28 experimental 0.0562 3.273 0.969 11.049

7 experimental 10 control 0.8234 1.230 0.199 7.593

7 experimental 10 experimental 0.9404 1.047 0.312 3.514

7 experimental 14 control 0.2708 0.660 0.315 1.382

7 experimental 14 experimental 0.0648 2.379 0.948 5.970

7 experimental 18 control 0.7775 1.133 0.476 2.700

7 experimental 18 experimental 0.5038 0.806 0.427 1.519

7 experimental 21 control 0.0882 0.405 0.143 1.145

7 experimental 21 experimental 0.0318 0.435 0.203 0.930

7 experimental 28 control 0.9037 1.074 0.338 3.416

7 experimental 28 experimental 0.6143 0.703 0.179 2.768

10 control 10 experimental 0.7927 0.851 0.256 2.831

10 control 14 control 0.4125 0.537 0.121 2.378

10 control 14 experimental 0.4157 1.934 0.395 9.470

10 control 18 control 0.9289 0.921 0.152 5.597

10 control 18 experimental 0.5532 0.655 0.161 2.655

10 control 21 control 0.1509 0.329 0.072 1.500

10 control 21 experimental 0.1044 0.354 0.101 1.240

10 control 28 control 0.8756 0.873 0.159 4.780

10 control 28 experimental 0.4139 0.571 0.149 2.188

10 experimental 14 control 0.4913 0.630 0.169 2.346

10 experimental 14 experimental 0.1594 2.272 0.724 7.127

10 experimental 18 control 0.8980 1.082 0.323 3.623

10 experimental 18 experimental 0.5737 0.769 0.308 1.919

10 experimental 21 control 0.2459 0.386 0.078 1.926

10 experimental 21 experimental 0.1361 0.415 0.131 1.319

10 experimental 28 control 0.9635 1.026 0.347 3.033

10 experimental 28 experimental 0.1418 0.671 0.395 1.142

14 control 14 experimental <.0001 3.604 2.064 6.295

14 control 18 control 0.1383 1.717 0.840 3.508

14 control 18 experimental 0.4648 1.220 0.716 2.081

14 control 21 control 0.1261 0.613 0.327 1.148

115

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

14 control 21 experimental 0.2908 0.659 0.304 1.429

14 control 28 control 0.3078 1.627 0.638 4.146

14 control 28 experimental 0.9398 1.065 0.208 5.446

14 experimental 18 control <.0001 0.476 0.333 0.680

14 experimental 18 experimental 0.0002 0.339 0.192 0.596

14 experimental 21 control 0.0014 0.170 0.057 0.505

14 experimental 21 experimental 0.0040 0.183 0.057 0.582

14 experimental 28 control <.0001 0.451 0.307 0.663

14 experimental 28 experimental 0.1315 0.295 0.061 1.441

18 control 18 experimental 0.2263 0.711 0.409 1.236

18 control 21 control 0.0735 0.357 0.116 1.103

18 control 21 experimental 0.1271 0.384 0.112 1.313

18 control 28 control 0.8202 0.948 0.596 1.506

18 control 28 experimental 0.5680 0.620 0.120 3.195

18 experimental 21 control 0.0969 0.502 0.223 1.132

18 experimental 21 experimental 0.0888 0.540 0.266 1.098

18 experimental 28 control 0.4729 1.333 0.608 2.925

18 experimental 28 experimental 0.8323 0.873 0.248 3.077

21 control 21 experimental 0.8456 1.075 0.519 2.226

21 control 28 control 0.1799 2.654 0.637 11.055

21 control 28 experimental 0.5582 1.737 0.273 11.039

21 experimental 28 control 0.2177 2.469 0.587 10.391

21 experimental 28 experimental 0.4459 1.616 0.470 5.553

28 control 28 experimental 0.5983 0.655 0.135 3.168

Model 3: predictors of intensity ≥ 3 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 4.80 0.3082

side 1 0.57 0.4491

day*side 4 25.42 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.5688 1.466 0.393 5.466

7 control 10 control 0.4851 1.846 0.330 10.324

7 control 10 experimental 0.3535 2.182 0.420 11.337

116

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 14 control 0.3174 1.495 0.680 3.285

7 control 14 experimental 0.0351 3.538 1.092 11.463

7 control 18 control 0.9887 1.007 0.385 2.634

7 control 18 experimental 0.4840 1.491 0.487 4.565

7 control 21 control 0.9941 0.997 0.441 2.254

7 control 21 experimental 0.9450 0.962 0.315 2.931

7 control 28 control 0.0060 3.923 1.479 10.407

7 control 28 experimental 0.3103 1.846 0.565 6.034

7 experimental 10 control 0.8431 1.259 0.128 12.345

7 experimental 10 experimental 0.6752 1.488 0.232 9.556

7 experimental 14 control 0.9658 1.019 0.424 2.451

7 experimental 14 experimental 0.3325 2.414 0.406 14.340

7 experimental 18 control 0.6199 0.687 0.156 3.030

7 experimental 18 experimental 0.9808 1.017 0.255 4.052

7 experimental 21 control 0.5208 0.680 0.210 2.207

7 experimental 21 experimental 0.1850 0.656 0.352 1.224

7 experimental 28 control 0.2116 2.676 0.571 12.535

7 experimental 28 experimental 0.7937 1.259 0.224 7.084

10 control 10 experimental 0.5811 1.182 0.653 2.139

10 control 14 control 0.7969 0.810 0.162 4.048

10 control 14 experimental 0.2974 1.917 0.564 6.516

10 control 18 control 0.4431 0.545 0.116 2.567

10 control 18 experimental 0.7411 0.808 0.227 2.868

10 control 21 control 0.4442 0.540 0.111 2.618

10 control 21 experimental 0.4941 0.521 0.080 3.378

10 control 28 control 0.3551 2.125 0.430 10.501

10 control 28 experimental 1.0000 1.000 0.515 1.940

10 experimental 14 control 0.5693 0.685 0.186 2.522

10 experimental 14 experimental 0.4530 1.622 0.459 5.734

10 experimental 18 control 0.2716 0.462 0.116 1.832

10 experimental 18 experimental 0.4390 0.683 0.261 1.792

10 experimental 21 control 0.2482 0.457 0.121 1.727

10 experimental 21 experimental 0.2698 0.441 0.103 1.889

10 experimental 28 control 0.4602 1.798 0.379 8.533

10 experimental 28 experimental 0.5975 0.846 0.455 1.573

14 control 14 experimental 0.1309 2.368 0.774 7.245

14 control 18 control 0.2680 0.674 0.335 1.355

117

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

14 control 18 experimental 0.9948 0.998 0.505 1.970

14 control 21 control 0.0654 0.667 0.434 1.026

14 control 21 experimental 0.0449 0.643 0.418 0.990

14 control 28 control 0.0615 2.625 0.955 7.219

14 control 28 experimental 0.7162 1.235 0.395 3.860

14 experimental 18 control 0.0682 0.285 0.074 1.099

14 experimental 18 experimental 0.2065 0.421 0.110 1.611

14 experimental 21 control 0.0583 0.282 0.076 1.045

14 experimental 21 experimental 0.0897 0.272 0.060 1.224

14 experimental 28 control 0.6569 1.109 0.703 1.748

14 experimental 28 experimental 0.2323 0.522 0.179 1.517

18 control 18 experimental 0.1245 1.481 0.897 2.443

18 control 21 control 0.9507 0.990 0.720 1.361

18 control 21 experimental 0.9215 0.955 0.381 2.392

18 control 28 control 0.0543 3.896 0.975 15.567

18 control 28 experimental 0.3371 1.833 0.532 6.321

18 experimental 21 control 0.0807 0.669 0.426 1.050

18 experimental 21 experimental 0.2855 0.645 0.288 1.442

18 experimental 28 control 0.1891 2.631 0.621 11.147

18 experimental 28 experimental 0.6552 1.238 0.485 3.161

21 control 21 experimental 0.9109 0.965 0.512 1.817

21 control 28 control 0.0361 3.935 1.093 14.166

21 control 28 experimental 0.2964 1.852 0.583 5.887

21 experimental 28 control 0.0472 4.080 1.017 16.362

21 experimental 28 experimental 0.3579 1.920 0.478 7.713

28 control 28 experimental 0.2471 0.471 0.131 1.686

6. Analysis of Pulp Middle intensity: no ankylosis in experimental side Model 1: predictors of intensity ≥ 1 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 114.97 <.0001

side 1 4.44 0.0350

day*side 4 39.27 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

118

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.0180 0.213 0.059 0.767

7 control 10 control 0.0006 2.656 1.523 4.634

7 control 10 experimental 0.3926 0.787 0.455 1.363

7 control 14 control 0.0698 0.472 0.210 1.063

7 control 14 experimental 0.7122 1.299 0.324 5.204

7 control 18 control 0.3125 0.668 0.305 1.462

7 control 18 experimental 0.2465 1.159 0.903 1.488

7 control 21 control 0.1705 0.386 0.099 1.505

7 control 21 experimental 0.0005 0.241 0.109 0.535

7 control 28 control 0.6039 1.308 0.475 3.603

7 control 28 experimental 0.2938 0.531 0.163 1.731

7 experimental 10 control 0.0010 12.500 2.775 56.306

7 experimental 10 experimental 0.0559 3.704 0.968 14.175

7 experimental 14 control 0.3741 2.222 0.382 12.928

7 experimental 14 experimental 0.0287 6.111 1.207 30.947

7 experimental 18 control 0.0004 3.143 1.659 5.955

7 experimental 18 experimental 0.0045 5.455 1.693 17.575

7 experimental 21 control 0.0531 1.818 0.992 3.332

7 experimental 21 experimental 0.8007 1.136 0.421 3.067

7 experimental 28 control 0.0028 6.154 1.872 20.226

7 experimental 28 experimental 0.0431 2.500 1.029 6.074

10 control 10 experimental 0.0017 0.296 0.138 0.634

10 control 14 control 0.0003 0.178 0.069 0.455

10 control 14 experimental 0.4180 0.489 0.087 2.763

10 control 18 control 0.0173 0.251 0.081 0.784

10 control 18 experimental 0.0290 0.436 0.207 0.919

10 control 21 control 0.0262 0.145 0.027 0.796

10 control 21 experimental <.0001 0.091 0.030 0.279

10 control 28 control 0.1424 0.492 0.191 1.269

10 control 28 experimental 0.0084 0.200 0.060 0.662

10 experimental 14 control 0.1333 0.600 0.308 1.169

10 experimental 14 experimental 0.5187 1.650 0.360 7.552

10 experimental 18 control 0.7363 0.849 0.326 2.207

10 experimental 18 experimental 0.1977 1.473 0.817 2.654

10 experimental 21 control 0.2693 0.491 0.139 1.735

10 experimental 21 experimental 0.0508 0.307 0.094 1.004

10 experimental 28 control 0.2837 1.662 0.657 4.204

119

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

10 experimental 28 experimental 0.6090 0.675 0.150 3.043

14 control 14 experimental 0.1146 2.750 0.783 9.663

14 control 18 control 0.5813 1.414 0.413 4.847

14 control 18 experimental 0.0248 2.455 1.120 5.377

14 control 21 control 0.8096 0.818 0.160 4.186

14 control 21 experimental 0.3116 0.511 0.139 1.875

14 control 28 control 0.0386 2.769 1.055 7.270

14 control 28 experimental 0.8980 1.125 0.186 6.808

14 experimental 18 control 0.2366 0.514 0.171 1.547

14 experimental 18 experimental 0.8476 0.893 0.280 2.845

14 experimental 21 control 0.0962 0.298 0.071 1.241

14 experimental 21 experimental 0.0014 0.186 0.066 0.522

14 experimental 28 control 0.9918 1.007 0.267 3.799

14 experimental 28 experimental 0.3219 0.409 0.070 2.398

18 control 18 experimental 0.0719 1.736 0.952 3.163

18 control 21 control 0.1401 0.579 0.280 1.197

18 control 21 experimental <.0001 0.362 0.224 0.584

18 control 28 control 0.1535 1.958 0.778 4.927

18 control 28 experimental 0.6205 0.795 0.321 1.968

18 experimental 21 control 0.0701 0.333 0.102 1.095

18 experimental 21 experimental <.0001 0.208 0.110 0.396

18 experimental 28 control 0.8020 1.128 0.439 2.896

18 experimental 28 experimental 0.1902 0.458 0.143 1.473

21 control 21 experimental 0.4336 0.625 0.193 2.027

21 control 28 control 0.0587 3.385 0.956 11.982

21 control 28 experimental 0.6591 1.375 0.334 5.660

21 experimental 28 control 0.0026 5.415 1.804 16.257

21 experimental 28 experimental 0.0612 2.200 0.964 5.022

28 control 28 experimental 0.1541 0.406 0.118 1.402

Model 2: predictors of intensity ≥ 2 Model did not converge. Model 1: predictors of intensity ≥ 3 Model did not converge. 7. Analysis of intensity: ankylosis in experimental side

120

Due to the small number of experimental teeth with ankylosis, all models failed to converge apart from: Analysis of Pulp Dentine intensity: ankylosis in experimental side Model : predictors of intensity ≥ 2 Wald Statistics For Type 3 GEE Analysis

Source DF Chi-Square P value

day 4 202.27 <.0001

side 1 7.31 0.0068

day*side 4 82.00 <.0001

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

7 control 7 experimental 0.4009 0.750 0.383 1.467

7 control 10 control 0.0003 5.727 2.220 14.773

7 control 10 experimental <.0001 9.000 4.649 17.423

7 control 14 control 0.0209 3.073 1.185 7.967

7 control 14 experimental 0.0186 7.200 1.390 37.291

7 control 18 control 0.0055 5.276 1.632 17.060

7 control 18 experimental <.0001 4.500 3.106 6.520

7 control 21 control 0.1352 1.884 0.821 4.323

7 control 21 experimental <.0001 9.000 4.649 17.423

7 control 28 control 0.0152 5.000 1.364 18.326

7 control 28 experimental <.0001 18.000 9.298 34.845

7 experimental 10 control 0.0111 7.636 1.592 36.633

7 experimental 10 experimental <.0001 12.000 7.561 19.046

7 experimental 14 control 0.0019 4.098 1.680 9.993

7 experimental 14 experimental 0.0002 9.600 2.975 30.975

7 experimental 18 control <.0001 7.034 3.280 15.088

7 experimental 18 experimental <.0001 6.000 3.082 11.682

7 experimental 21 control 0.0377 2.512 1.054 5.987

7 experimental 21 experimental <.0001 12.000 7.561 19.046

7 experimental 28 control 0.0005 6.667 2.306 19.275

7 experimental 28 experimental <.0001 24.000 15.121 38.093

10 control 10 experimental 0.5254 1.571 0.389 6.341

10 control 14 control 0.4125 0.537 0.121 2.378

10 control 14 experimental 0.8397 1.257 0.137 11.538

10 control 18 control 0.9289 0.921 0.152 5.597

10 control 18 experimental 0.6220 0.786 0.301 2.050

121

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

10 control 21 control 0.1509 0.329 0.072 1.500

10 control 21 experimental 0.5254 1.571 0.389 6.341

10 control 28 control 0.8756 0.873 0.159 4.780

10 control 28 experimental 0.1076 3.143 0.779 12.681

10 experimental 14 control 0.0013 0.341 0.177 0.658

10 experimental 14 experimental 0.6998 0.800 0.257 2.487

10 experimental 18 control 0.1195 0.586 0.299 1.148

10 experimental 18 experimental 0.0047 0.500 0.309 0.808

10 experimental 21 control 0.0005 0.209 0.087 0.506

10 experimental 21 experimental 1.0000 1.000 1.000 1.000

10 experimental 28 control 0.1950 0.556 0.228 1.352

10 experimental 28 experimental <.0001 2.000 2.000 2.000

14 control 14 experimental 0.1971 2.343 0.643 8.543

14 control 18 control 0.1383 1.717 0.840 3.508

14 control 18 experimental 0.2966 1.464 0.716 2.997

14 control 21 control 0.1261 0.613 0.327 1.148

14 control 21 experimental 0.0013 2.929 1.521 5.640

14 control 28 control 0.3078 1.627 0.638 4.146

14 control 28 experimental <.0001 5.857 3.041 11.281

14 experimental 18 control 0.3038 0.733 0.405 1.325

14 experimental 18 experimental 0.5041 0.625 0.157 2.482

14 experimental 21 control 0.1240 0.262 0.047 1.444

14 experimental 21 experimental 0.6998 1.250 0.402 3.887

14 experimental 28 control 0.2327 0.694 0.382 1.264

14 experimental 28 experimental 0.1134 2.500 0.804 7.773

18 control 18 experimental 0.7298 0.853 0.346 2.104

18 control 21 control 0.0735 0.357 0.116 1.103

18 control 21 experimental 0.1195 1.706 0.871 3.341

18 control 28 control 0.8202 0.948 0.596 1.506

18 control 28 experimental 0.0003 3.412 1.742 6.683

18 experimental 21 control 0.0452 0.419 0.179 0.981

18 experimental 21 experimental 0.0047 2.000 1.237 3.232

18 experimental 28 control 0.8304 1.111 0.424 2.914

18 experimental 28 experimental <.0001 4.000 2.475 6.465

21 control 21 experimental 0.0005 4.778 1.976 11.551

21 control 28 control 0.1799 2.654 0.637 11.055

21 control 28 experimental <.0001 9.556 3.952 23.102

122

Post-hoc comparisons

Day Side _Day _Side P-value Odds ratio Lower 95% CI Upper 95% CI

21 experimental 28 control 0.1950 0.556 0.228 1.352

21 experimental 28 experimental . 2.000 2.000 2.000

28 control 28 experimental 0.0047 3.600 1.480 8.758

Results of Ankylotic Sections 1. Analysis of PDL cementum intensity: ankylosis in experimental side To compare PDL cementum intensity according to side of the mouth (experimental, control & excluding time), separate binary logistic generalised estimating equations were fitted to the data. Model Contrast P-value Odds ratio Lower 95% CI Upper 95% CI

1. Predictors of intensity ≥ 1 Control vs. experimental <.0001 0.041 0.010 0.181

2. Predictors of intensity ≥ 2 Control vs. experimental <.0001 0.060 0.039 0.093

3. Predictors of intensity ≥ 3 Control vs. experimental <.0001 0.014 0.004 0.045

The odds of having staining intensity ≥ 1 was greatly reduced in control teeth compared to experimental teeth with ankylosis (OR = 0.04; 95% CI 0.01, 0.18; p < 0.0001). Similar results were observed for intensity ≥ 2 and intensity ≥ 3. Collectively the results indicate that the odds of staining intensity was greatly increased by ankylosis. 2. Analysis of PDL bone intensity: ankylosis in experimental side Model Contrast P-value Odds ratio Lower 95% CI Upper 95% CI

1. Predictors of intensity ≥ 1 Control vs. experimental 0.4950 1.375 0.551 3.432

2. Predictors of intensity ≥ 2 Control vs. experimental 0.0044 0.510 0.320 0.811

3. Predictors of intensity ≥ 3 Control vs. experimental 0.0763 0.381 0.131 1.107

3. Analysis of PDL mid intensity: ankylosis in experimental side Model Contrast P-value Odds ratio Lower 95% CI Upper 95% CI

1. Predictors of intensity ≥ 1 Control vs. experimental 0.2785 0.619 0.260 1.474

2. Predictors of intensity ≥ 2 Control vs. experimental <.0001 0.158 0.075 0.335

3. Predictors of intensity ≥ 3 Control vs. experimental Model did not converge

4. Analysis of PDL BV intensity: ankylosis in experimental side Model Contrast P-value Odds ratio Lower 95% CI Upper 95% CI

1. Predictors of intensity ≥ 1 Control vs. experimental <.0001 4.467 3.169 6.295

2. Predictors of intensity ≥ 2 Control vs. experimental 0.0003 2.548 1.541 4.214

3. Predictors of intensity ≥ 3 Control vs. experimental 0.8897 1.111 0.251 4.928

5. Analysis of Pulp Dentine intensity: ankylosis in experimental side

123

Model Contrast P-value Odds ratio Lower 95% CI Upper 95% CI

1. Predictors of intensity ≥ 1 Control vs. experimental 0.5450 1.473 0.420 5.159

2. Predictors of intensity ≥ 2 Control vs. experimental 0.2842 1.439 0.739 2.799

3. Predictors of intensity ≥ 3 Control vs. experimental 0.0095 1.533 1.110 2.116

Error Study

Table of measure1 by measure2

measure1 measure2

Frequency 0 1 2 3 Total

0 85 0 0 0 85

1 0 152 1 0 153

2 0 4 75 8 87

3 0 0 6 29 35

Total 85 156 82 37 360

Statistics for Table of measure1 by measure2

Test of Symmetry

Statistic (S) 2.0857

DF 6

Pr > S 0.9116

Kappa Statistics

Statistic Value ASE 95% Confidence Limits

Simple Kappa 0.9241 0.0167 0.8912 0.9569

Weighted Kappa 0.9465 0.0117 0.9235 0.9695

Sample Size = 360