Periodontology 2000, Vol. 24, 2000, 73–98 Copyright C Munksgaard 2000Printed in Denmark ¡ All rights reserved

PERIODONTOLOGY 2000ISSN 0906-6713

Molecular and cell biology ofcementumNAZAN E. SAYGIN, WILLIAM V. GIANNOBILE & MARTHA J. SOMERMAN

Cementum was first described in 1835 (53), and yetuntil recently has remained a poorly defined tissueat the cellular and molecular level. While researchthrough the years has demonstrated that cementumis a unique tissue histologically (19, 21), examinationof the proteins expressed by cells forming ce-mentum, cementoblasts has not resulted in defini-tive identification of proteins specific to cementum(141). In fact, unlike dentin and enamel, where thereare clear differences in the proteins present in thesetissues and the factors regulating their functionwhen compared with bone, cementum appears tocontain factors in common with those associatedwith bone and to be developmentally controlled bysimilar factors. Alternatively, it has been suggestedthat factors supplied by epithelial cells within the lo-cal environment may, at least in part, influence thedifferentiation of follicle cells into cementoblasts,thereby laying down cementum (67), or in fact thatepithelial root sheath cells undergo transformationinto cementoblasts and thus provide the appropriatematrix for cementum formation (Fig. 1) (18, 21).Thus, the dogma that follicle cells have the capacityto differentiate into an osteoblast, a cementoblast ora periodontal ligament fibroblast has never beenproven and is now being challenged. This chapterwill attempt to present existing knowledge as to thecellular and molecular properties of cementum in aframework that provides the clinician the infor-mation necessary for careful evaluation of agentsthat are being marketed for their ability to regenerateperiodontal tissues, with a specific focus on agentstargeted at regeneration of cementum.

There is accumulating histological evidence thatcementum formation is critical for appropriatematuration of the periodontium, both during devel-opment as well as that associated with regenerationof periodontal tissues. Despite many years of re-search and the importance that cementum isthought to play in the reparative process followingperiodontal disease, very little is known about the

73

cells responsible for formation of cementum, ce-mentoblasts. The wealth of what is known about ce-mentum comes from numerous, detailed studies ofits histology and composition, which will be touchedon only briefly here since the focus of this chapter ison the cellular and molecular properties of thistissue (19–22, 199). Light and electron microscopyhas enabled cementum to be classified into five dif-ferent subtypes based on the presence (cellular) orabsence (acellular) of cells and the source of collagenfibers (extrinsic versus intrinsic). All of these sub-types are quite different from bone, in that they arenot innervated, exhibit little or no remodeling andare avascular. Despite these differences, cementumis very similar to bone. First, diseases that affect theproperties of bone, often alter cementum’s prop-erties as well. For example, Paget’s disease results inhypercementosis, hypophosphatasia results in nocementum formation, with exfoliation of teeth, de-creased cementum is associated with hypopituitar-ism and defective cementum is seen in patients withcleidocranial dysplasia (175). Second, the compo-sition of cementum is similar to that of bone. Ce-mentum is approximately 50% hydroxyapatite and50% collagen and noncollagenous proteins. Proteinextracts of mature cementum promote cell attach-ment, migration and stimulate protein synthesis ofgingival fibroblasts and periodontal ligament cells.Investigation of these extracts revealed the presenceof bone sialoprotein, osteopontin, vitronectin andfibronectin. Immunocytochemistry and in situ hy-bridization confirmed the presence of these proteinsand further identified osteocalcin, g-carboxyglutam-ic acid, osteonectin, proteoglycans and severalgrowth factors. Two additional molecules, an ad-hesion molecule and a growth factor, have beenidentified and initial data suggest that they may beunique to cementum. Cementum attachment pro-tein may prove to be a cementum-specific collagen-like molecule, while a factor initially named ce-mentum-derived growth factor, now considered to

Saygin et al.

be an insulin-like growth factor-I–like molecule mayprove to have properties similar to those of insulin-like growth factor-I (see Table 1 for references).

In this chapter, the factors above are consideredin terms of their potential for having an active rolein triggering cementum formation, contrasting therole of specific factors during development of ce-mentum versus repair and regeneration of ce-mentum. Also, existing evidence as to the success ofsuch agents in regeneration of cementum in vivo willbe discussed. The next section is devoted to a dis-cussion on models used to investigate the propertiesof cementum, while the final section is more specu-lative in nature, providing a discussion on the im-pact of information obtained from ongoing studiesat the cellular and molecular level and at the tissueengineering level, to the design of attractive therap-ies for regeneration of periodontal tissues.

Regulators of cementogenesisOverview: events, cells and factorsassociated with cementum

Events

As shown in Fig. 1 and 2, while many of the eventsrequired for formation of cementum are well estab-lished, the actual cells and factors required to formthis tissue during development as well as during re-generation have yet to be defined. Ideal agents touse in attempts to regenerate tissues would includethose having the ability to promote migration andattachment of appropriate cells to the healing sitewith subsequent orchestration of cells to allow forcell differentiation, that is, to act as an osteoblast,cementoblast or periodontal ligament fibroblast.Equally important would be the ability of an agentto promote mineralization (new cementum) alongthe root surface, with insertion of periodontal liga-ment into cementum and opposing alveolar bone toform the periodontium. This simplified descriptionof events suggests that regeneration recapitulatesdevelopment (Fig. 2). However, it is important to rec-ognize that there are differences in both events andcells required for these two processes. For example,proper clot formation appears to be critical for ade-quate wound healing and subsequent regenerationof any tissue. In addition, during wound healingthere is a normal inflammatory response, resultingin release of several cytokines and growth factorsthat may not be associated with development of agiven tissue. In spite of these differences, it is poss-

74

ible that factors identified as having a role in regulat-ing the developing tooth root cementum, but notnecessarily associated with regeneration of thistissue, may prove of value for use in regenerativetherapies. As depicted in Table 1 and discussed be-low, while many of these factors are known to bepresent during development and regeneration ofperiodontal tissues, their precise functions in thesetissues are not clearly defined. Ongoing research tar-geted at defining the importance of these factors tocell function will help to clarify the role of these fac-tors in regulating periodontal tissues.

Cells

Histological examination of the healthy periodon-tium indicates that several types of mesenchymalcells are important for maintenance of a healthyperiodontium. Such cells include: periodontal liga-ment fibroblasts, responsible for ensuring a func-tional periodontal ligament region; osteoblasts andassociated progenitor cells, responsible for preserv-ing the surrounding alveolar bone; cementoblasts,root surface lining cells, that appear to be limited infunction in health but may be activated duringwound healing, and paravascular/marrow cells, thatare important for providing the required local nutri-ents at the site. In contrast, the cells responsible forformation of periodontal tissues, both developmen-tal and repair/regenerative aspects, are less definedand an area of intense investigation by several lab-oratories (Fig. 2).

At the developmental level there is existing evi-dence suggesting that ectomesenchymal cells, fol-licle cells and dental papilla cells, when triggered ap-propriately, have the capacity to act as cemento-blasts, or periodontal ligament fibroblasts orosteoblasts (4, 43, 172, 176, 226), (Fig. 1). In addition,although it is not the focus of this chapter it is im-portant to point out the strong evidence for the re-quirement of follicle cells and factors secreted bythese cells for normal tooth eruption, regardless ofroot formation (246, 248). Some of the factors thatmay trigger differentiation of follicle cells or possiblytransformation of Hertwig’s epithelial rooth sheathcells so as to function as cementoblasts are dis-cussed below and listed in Fig. 1 and Table 1. Thespecific cell type(s) having the capacity to functionas cementoblasts during wound healing remain(s)unknown as well. There is some suggestion that asmall population of periodontal ligament cells in themature periodontium have the capacity to undergodifferentiation toward an osteoblast or cementoblast

Molecular and cell biology of cementum

Fig. 1. Cementoblast maturation. Top. Depicts possible the intensity of expression). Abbreviations: OB: osteoblast;cell types involved in developing periodontal tissues. Bot- PDL: periodontal ligament cell; CM: cementoblast; EM:tom. Genes that are known to be expressed by cells at epithelial-mesenchymal; PTHrP: parathyroid hormone–various stages. Future research targeted at determining related protein; BMP: bone morphogenetic protein; Col.:the significance of these factors, as well as yet to be iden- collagen; CSF: colony-stimulating factor; EGF: epidermaltified factors, to cell activities will assist in the design of growth factor; ALKP: alkaline phosphatase; BSP: bone sia-predictable regenerative therapies. π or ª indicates loprotein; OC: osteocalcin; OPN: osteopontin.whether a factor is present or not (but does not indicate

phenotype and thereby act to lay down bone or ce-mentum (128, 167, 188). However, accumulating evi-dence exists to a support a role for periodontal liga-ment fibroblasts as inhibitors of mineralization (82,156, 169, 194). Thus, there may be distinct cell popu-lations within the periodontal ligament region thatcan both promote and inhibit mineral tissue forma-tion depending upon trigger factors. Also, it is highlylikely that other sources of cementobloast or osteo-blast progenitor cells include marrow stroma andparavascular and endosteal fibroblasts (151). At theresearch level there is a need to define both the cellsinvolved in regeneration as well as the trigger factorsfor controlling cell activities, in order to design pre-dictable regenerative therapies. Discussed below arespecific factors associated with periodontal tissuesduring development and in mature and regenerating

75

tissues that may prove to have potential as triggerfactors for use in regenerative therapies targeted atpromoting cementum formation.

Factors

Discussed below and outlined in Table 1 are factorsknown to be associated with cementum either dur-ing development and/or maturation and/or re-generation. Also included are molecules that havenot been fully characterized but have been reportedto be linked with cementum. Many of these factorshave been implicated as having a role in controllingseveral cell activities and thus can be listed undereach activity in Table 1; however, for simplicity, fac-tors are listed only under the activities currently con-sidered to be their major function.

Saygin et al.

Fig. 2. Regeneration versus development: events and cells. types involved and in some of the factors promoting cellWhile similar events occur during development and re- activities. ERS: Hertwig’s epithelial root sheath; PDL: peri-generation of tissues, there are clear differences in cell odontal ligament.

Adhesion and chemotactic factors

Central to the interaction of cells with the local en-vironment is the ability of cells, through receptor-ligand interactions, to be attracted to and remain atthe site of action. Several ligands have been iden-tified that appear to have an important role in at-tracting cells to specific sites during tissue develop-ment. Fibronectin is one of the most extensivelystudied of such molecules that, in addition to its rolein tissue development, also is purported to have annotable role in attracting and maintaining appropri-ate cells at healing sites. There are several compre-hensive, excellent reviews that present the import-ance of adhesion molecules and associated receptorsto a variety of cell activities, and the reader is re-ferred to these for in-depth information on thesemolecules (71, 130). The discussion here will belimited to adhesion and chemotactic factors thathave been regarded as having a role during develop-ment and/or regeneration of cementum.

76

DevelopmentData from in situ hybridization analyses indicatethat the adhesion molecules osteopontin and bonesialoprotein are expressed by cells along the root sur-face, cementoblasts, during early stages of tooth rootdevelopment. In contrast, cells expressing collagenare noted throughout the surrounding soft connec-tive tissue, follicle cells and periodontal ligamentfibroblasts (139, 142). Bone sialoprotein messengerRNA and protein remain localized to the root surfacein the mature tooth, while osteopontin protein isnoted within the periodontal ligament region in themature tooth (136, 142, 154, 253). In addition tothese adhesion molecules, laminin has been iden-tified on the dentin surface at the initiation of ce-mentum formation, where it has been speculatedthat this protein serves a role in attracting appropri-ate ‘‘cementoblast-like’’ cells to the root surface (145,146, 235). Further known factors, as well as yet to beidentified novel factors, secreted by epithelial cellsmay promote migration and/or adhesion of appro-

Molecular and cell biology of cementum

Tab

le1.

Mo

lecu

lar

fact

ors

asso

ciat

edw

ith

cem

entu

m.

Incl

ud

edin

this

tab

lear

eb

oth

esta

bli

shed

fact

ors

asw

ell

asfa

cto

rssu

gges

ted

(bu

tst

ill

un

der

inve

stig

atio

n)

tob

eim

po

rtan

tfo

rd

evel

op

men

t/m

ain

ten

ance

/reg

ener

atio

no

fce

men

tum

.M

od

ified

and

up

dat

edfr

om

Mac

Nei

let

al.

(141

)

Pro

po

sed

acti

vity

Dev

elo

pin

gce

men

tum

Mat

ure

cem

entu

mR

egen

erat

ive

cem

entu

m

a.A

dh

esio

no

rP

rote

ogl

ycan

s(1

1)P

rote

ogl

ycan

s(1

0,12

,41

,42

)N

ot

esta

bli

shed

chem

oat

trac

tio

nO

steo

po

nti

n(2

9,21

6,22

4)O

steo

po

nti

n(1

55)

Ost

eop

on

tin

(125

)B

on

esi

alo

pro

tein

(36,

37,

140,

142,

218)

Bo

ne

sial

op

rote

in(1

25,

137,

140,

142,

214,

215)

Bo

ne

sial

op

rote

in(1

25)

Fib

ron

ecti

n(2

34)

Fib

ron

ecti

n(1

33,

215)

No

tes

tab

lish

edLa

min

in(2

34)

No

tes

tab

lish

edN

ot

esta

bli

shed

No

tes

tab

lish

edC

emen

tum

adh

esio

np

rote

in(1

49,

171,

249)

No

tes

tab

lish

edC

olla

gen

I,II

I,X

II(1

40,

165,

227)

Co

llage

nI,

III,

XII

(56)

Co

llage

nI,

III

No

tes

tab

lish

edTe

nas

cin

(133

)N

ot

esta

bli

shed

Her

twig

’sep

ith

elia

lro

oth

shea

thfa

cto

rsH

ertw

ig’s

epit

hel

ial

roo

thsh

eath

fact

ors

No

tes

tab

lish

ed

b.M

ito

gen

esis

Gro

wth

ho

rmo

ne

(126

,25

7)N

ot

esta

bli

shed

No

tes

tab

lish

edTr

ansf

orm

ing

grow

thfa

cto

r-b

(60,

227)

Tran

sfo

rmin

ggr

owth

fact

or-

b(7

4)N

ot

esta

bli

shed

Insu

lin

-lik

egr

owth

fact

or-

I(2

27)

Cem

entu

mgr

owth

fact

or/

Insu

lin

-lik

egr

owth

fact

or-

IN

ot

esta

bli

shed

(159

,25

4,25

5)

c.D

iffe

ren

tiat

ion

Para

thyr

oid

ho

rmo

ne–

rela

ted

pro

tein

(180

,22

8)N

ot

esta

bli

shed

No

tes

tab

lish

edTr

ansf

orm

ing

grow

thfa

cto

r-b

(60,

227,

247)

Tran

sfo

rmin

ggr

owth

fact

or-

b(7

4)N

ot

esta

bli

shed

Bo

ne

mo

rph

oge

net

icp

rote

ins

(1)

Bo

ne

mo

rph

oge

net

icp

rote

ins

(202

)N

ot

esta

bli

shed

Her

twig

’sep

ith

elia

lro

oth

shea

thfa

cto

rs(6

7,12

1,21

0,21

1)H

ertw

ig’s

epit

hel

ial

roo

thsh

eath

fact

ors

(67)

No

tes

tab

lish

edO

steo

bla

st-s

pec

ific

tran

scri

pti

on

fact

or

(113

)O

steo

bla

st-s

pec

ific

tran

scri

pti

on

fact

or

No

tes

tab

lish

ed

d.

Min

eral

izat

ion

Bo

ne

sial

op

rote

in(1

02,

103,

140,

142)

Bo

ne

sial

op

rote

in(1

02,

103,

140,

142)

Bo

ne

sial

op

rote

in(1

25)

Ost

eoca

lcin

(56)

Ost

eoca

lcin

(29,

224)

No

tes

tab

lish

edO

steo

po

nti

n(2

9,10

2,10

3,21

6,22

4)O

steo

po

nti

n(2

9,10

2,10

3,21

6,22

4)O

steo

po

nti

n(1

25)

Co

llage

nI,

XII

(116

,13

9,16

5)C

olla

gen

I,X

II(2

9,13

9,14

0)C

olla

gen

IP

rote

ogl

ycan

s(1

1)P

rote

ogl

ycan

s(1

0,12

,41

,42

)N

ot

esta

bli

shed

77

Saygin et al.

priate cells to the root surface (19, 21, 67, 101, 121).Importantly, the ratonale behind the use of enamelmatrix derivative (129: see entire issue) is that en-amel matrix proteins may promote cementoblast ac-tivity including proliferation, migration/adhesion, aswell as cell differentiation. With regard to cell differ-entiation this would be analogous to the known abil-ity of ameloblasts-odontoblasts to secrete productsrequired for the development of both dentin and en-amel (epithelial-mesenchymal interactions).

Suggested roles for bone sialoprotein includeacting as an adhesion molecule to maintain appli-cable cells at the root surface and as an initiator ofmineral formation along the root surface (102, 104,170, 213, 215, 221). Importantly, the temporal andspatial expression of bone sialoprotein during ce-mentogenesis and bone formation is consistent witha role for this molecule in promoting mineral forma-tion (16, 36–39, 56, 73, 137, 140, 142, 153). Like bonesialoprotein, the phosphoglycoprotein osteopontincontains the well recognized adhesion domain, argi-nine-glycine-aspartic acid (RGD) targeted to specificintegrin receptors, as well as other adhesion regionsthat act to promote migration and cell adhesion invitro (52, 213, 241, 250, 251). However, unlike bonesialoprotein, which is selective to mineralizedtissues, osteopontin is expressed by several cells andlocalized to several tissues where its function mayrelate to post-translational modifications within spe-cific tissues. With regard to bone and cementum, os-teopontin is expressed during periods of osteogenic/cementogenic activity in situ (29, 31, 36–38, 56, 137,138, 142, 152, 154, 155, 224), and also is associatedwith resorptive activity (58, 147, 182, 203, 224). Oste-opontin has been linked to regulation of ectopiccrystal formation, where evidence supports a role forosteopontin in controlling the extent of hydroxyapa-tite crystal nucleation and/or growth (17, 76, 83,102–104, 204, 209, 219). Also, osteopontin has beenreported to inhibit apoptotic events such as thoseassociated with inflammation (51, 65), and this abil-ity may have some significance in the regulation ofcells at sites during development of cementum andalso during wound healing. Thus, while both bonesialoprotein and osteopontin may have a role in re-cruitment and maintenance of selective cells at theroot surface, an equally important role may be re-lated to control of mineralization along the root sur-face (see below under mineralization).

Type I collagen, the most abundant protein of ce-mentum, is known to promote cell attachment, butalso is a critical molecule for maintaining the integ-rity of both soft and hard connective tissues, during

78

development as well as in repair. Less clear is therole of type XII collagen in developing and in main-taining a functional periodontal ligament. Studies byKarimbux (116) and by MacNeil et al. (139) demon-strated that type XII collagen is expressed by peri-odontal ligament cells during later stages of root for-mation but not by follicle cells prior to root develop-ment. Thus, while it is not clear whether thiscollagen has a role in promoting cell adhesion, itmay act through association with type I collagen information and maintenance of a functional peri-odontal ligament, thus preventing ankylosis.

RegenerationMore recent in vivo studies, using rat periodontal de-fect models, have shown that both bone sialoproteinand osteopontin are expressed by cells linked to for-mation of mineralized tissues, while osteopontinalso is expressed by cells within the newly formingperiodontal ligament (125). Future studies directedat overexpressing or blocking expression of thesemolecules during periodontal wound healing shouldprovide additional information required to establishthe function of these molecules in mineralizedtissues and also determine the value of using suchagents clinically for enhancing regeneration of peri-odontal tissues.

MaturationMature cementum contains the adhesion moleculesmentioned above as well as vitronectin and ce-mentum attachment protein. Data to date indicatethat cementum attachment protein shares signifi-cant homology to type I collagen (249). The actualspecificity of cementum attachment protein to ce-mentum, or in fact whether this is a unique protein,awaits further research and availability of DNAprobes to determine cells expressing cementumattachment protein during tooth root developmentand maturation (9, 249).

In addition to the classical adhesion moleculeslisted in Table 1, proteoglycans, ubiquitous to allconnective tissues, most likely play a role in regulat-ing cell-cell and cell-matrix interactions both duringdevelopment as well as in regeneration of cementum(11). Proteoglycans are linked to several aspects ofthe cell from the matrix, to cell surfaces, to cell or-ganelles. They are highly anionic molecules impli-cated in playing a role in several cell functions in-cluding tissue hydration, storage for growth factorsthereby regulating activity of growth factors, and incontrolling collagen fiber formation, cell adhesion,cell growth and cell differentiation.

Molecular and cell biology of cementum

Mitogens

Details on the potential role of mitogens in control-ling cell function are provided later. It is clear thatfactors that promote cell proliferation offer a criticalmass of cells at a given site, as required for sub-sequent synthesis and secretion of molecules toform the extracellular matrix needed for cell differ-entiation and for mineralization.

As shown in Table 1, mitogens that have beenmapped during tooth root development include themembers of the transforming growth factor-b super-family, growth hormone, insulin-like growth factor-I/II and parathyroid hormone–related protein. Manymore growth factors have been identified duringodontogenesis, and these include fibroblast growthfactor, platelet-derived growth factor, and growthhormone (59, 60, 231, 256–258), but their relation-ship to root formation have not been detailed. Whilethese molecules have been called growth factors,their role during tooth development does not appearto be associated with proliferative activity. In factsome transforming growth factor-bs and parathyroidhormone–related protein may have a role in regulat-ing cell differentiation and subsequently mineraliza-tion, as discussed in the next section.

In addition to the factors listed above, Narayanan etal. (107, 108) extracted a factor from mature ce-mentum they called cementum-derived growth fac-tor. This factor has been shown to enhance prolifer-ation of gingival fibroblasts and periodontal ligamentcells (107, 108). Cementum-derived growth factorbears substantial homology to insulin-like growth fac-tor-I, and thus additional research is required to de-termine the uniqueness of this molecule to specifictissues. In this regard, studies from Waters’ group sug-gest that both insulin-like growth factor and growthhormone may have independent roles in promotingcementogenesis, including perhaps the ability of bothof these molecules to induce expression of bone mor-phogenetic proteins (47, 126, 257).

Differentiation

Significant efforts have been placed on determiningfactors regulating differentiation of cells during rootdevelopment into cells that can function as peri-odontal ligament cells, cementoblasts and/or osteo-blasts, with the goal of yielding information that mayprove valuable for developing ideal agents for pro-moting regeneration of periodontal tissues. While thefactors listed in Table 1 are attractive candidates andsome have shown promise at the clinical level (see

79

later) the exact mechanisms, cells or factors requiredfor promoting formation of periodontal tissues hasyet to be determined. The discussion below providesa brief overview of specific factors. The readers are re-ferred to the original studies and detailed reviews formore information in this area (4, 77, 89, 129 – see en-tire issue, 245 – see entire issue).

DevelopmentIt is well established that epithelial-mesenchymalsignaling through specific molecules is required forformation of many tissues, including lung, heart,hair follicles, and tooth crown (enamel and dentin)(135, 231, 233). In contrast, the exact mechanisms orfactors involved in regulating formation of ce-mentum remain undefined. For example, strong evi-dence exists to support a role for both bone morpho-genetic protein 2 and 4 as molecules critical to epi-thelial-mesenchymal signaling events required forenamel and dentin formation (15, 239); however,there is no indication that bone morphogenetic pro-teins 2, 4 or 7, based on expression pattern, serve asepithelial-mesenchymal signaling molecules or haveany other role in formation of cementum (1). In con-trast, bone morphogenetic protein-3 may play a rolein cementum formation since bone morphogeneticprotein-3 transcripts are expressed by follicle cellssurrounding murine first molars (1). More recently,Thesleff’s laboratory in collaboration with our lab-oratory, using the murine first molar model, havemapped expression of bone morphogenetic proteinsat later stages of root development. Notably, of thebone morphogenetic proteins examined (2, 3, 4 and7) only bone morphogenetic protein-3 was ex-pressed at later stages of root development. Bonemorphogenetic protein-3 was selective to root liningcells and was not expressed by cells within the peri-odontal ligament region (unpublished data). Whilethis is speculative at this point, it is possible thatbone morphogenetic protein-3 plays an importantrole in differentiation of follicle cells along the ce-mentoblast pathway.

In terms of the importance of epithelial productsin influencing cementoblast phenotype severalstudies, including immunohistochemical, in situhybridization and recombination studies, providedata that support a role for products secreted bythe epithelial root sheath in regulating cementumformation. These factors, some as yet to be iden-tified, may attract cells to the root surface, such aslaminin (145, 146, 235), or promote differentiationof follicle cells along the cementoblast pathway,

Saygin et al.

such as sheathlin (also called ameloblastin andamelin) (67, 101, 121).

Alkaline phosphatase is known to regulate forma-tion of mineralized tissues, including cementum. Re-cent studies focused on histological examination ofteeth from alkaline phosphatase knock out mice sug-gest that alkaline phosphatase may have a morecritical role in controlling formation of acellular ce-mentum versus cellular cementum (14). This is aninteresting and potential important finding in termsof efforts directed at regeneration of cementum,where the cementum formed appears to be cellularin nature. Thus, factors controlling formation ofcellular cementum may be different than those in-volved in formation of acellular cementum.

Another molecule that may prove to be importanttoward expression of the cementoblast phenotype isparathyroid hormone–related protein. There is ac-cumulating evidence that parathyroid hormone–re-lated protein has a role in regulating early stages oftooth development and tooth eruption (50, 180).Rescued parathyroid hormone–related proteinknockout mice are small in stature, exhibit cranialchondrodystrophy and a failure of tooth eruption(180). The role for this molecule in tooth root devel-opment is less clear. Parathyroid hormone–relatedprotein messenger RNA and the associated para-thyroid hormone/parathyroid hormone–related pro-tein type I receptor have been identified within de-veloping periodontal tissues (13, 174, 180, 228), andan extract of parathyroid hormone was shown to en-hance both tooth eruption and orthodontic toothmovement (50, 180).

Other factors such as insulin-like growth factorand proteoglycans, present in developing and ma-ture cementum may serve multiple functions includ-ing monitoring mineralization and controlling celldifferentiation.

In addition to the factors discussed above, severalstudies suggest that the transcription factor, corebinding transcription factor 1/osteoblast-specifictranscription factor 2 is a key activator of osteoblastdifferentiation (8, 62, 119, 173). Importantly, duringtooth development in mice, osteoblact-specific tran-scription factor 2 is expressed by mesenchymal cells,including follicle cells and cementoblasts, as well asby ameloblasts (113).

RegenerationAs presented later, the factors discussed above mayprove important for promoting regeneration andthus it is not surprising that some of these agentshave been used in periodontal regenerative models.

80

Platelet-derived growth factor and platelet-derivedgrowth factor-receptor have been localized to peri-odontal tissues during wound healing (85). In thisregard, an apparent limitation to current regenera-tive therapies is considered to be related to main-taining sufficient quantities of a specific factor at thelocal site. Thus, studies examining the quantity ofspecific growth factors and their associated recep-tors at a given site during wound healing, such asa non-critical size defect, may help to establish theessential factors, both type and quantity, required forpromoting regeneration at healing sites.

MaturationAs with other mineralized tissues, mature cementumcontains several factors that were either secreted bycementoblasts or absorbed by hydroxyapatite duringroot formation (11).

MineralizationAlthough several factors have been implicated ashaving an important role in controlling mineraliza-tion, the specific role of individual molecules has yetto be established. Investigations targeted at mappingthe expression pattern for specific molecules devel-opmentally, as well as use of knock-out mice, haveenhanced understanding of these molecules in themineralization process, including cementogenesis.

DevelopmentAs noted in Table 1 several factors that have beenimplicated as having a role in regulating mineraliza-tion may also have other functions. In this context,the potential roles for bone sialoprotein, osteopontinand type I and XII collagen in regulating crystalgrowth, both during development and in regenera-tion of periodontal tissues, were discussed under theheading of adhesion molecules.

In addition to these molecules both osteocalcinand proteoglycans are considered to be importantfor regulating mineralization in various tissues in-cluding cementum. We and others have demon-strated the selective expression of osteocalcin to rootlining cells, cementoblasts, with this profile of ex-pression being maintained in the mature tissue, thatis, periodontal ligament fibroblasts do not expressosteocalcin. Expression pattern for matrix ‘‘gla’’ pro-tein during root development and in mature tissueshas not been studied. The temporal and spatial ex-pression pattern for osteocalcin during cementogen-esis (56), and the selectivity of osteocalcin to min-eralized tissues suggests that it has a prominent rolein root development. Studies by Ducy et al. (61), in-

Molecular and cell biology of cementum

dicate that transgenic mice lacking the osteocalcingene develop a phenotype marked by enhancedmineral density, suggesting that osteocalcin may bea controller of mineral formation. Thus, it is possiblethat osteocalcin is one of the molecules that controlthe mineral-to-ligament ratio, allowing for formationof a periodontal ligament region versus an ankylosissituation.

RegenerationAs discussed above under adhesion/migration fac-tors, both osteopontin and bone sialoprotein are ex-pressed in cells/tissues associated with periodontalwound healing (125). Also, osteocalcin and types I,III and XII collagen have been associated withwound healing and/or stress to the periodontaltissues associated with orthodontic tooth movement(238). These molecules, as well as proteoglycans,most likely play a critical role during wound healingin balancing formation of cementum and surround-ing bone, with development of a functional peri-odontal ligament.

MaturationMature cementum contains all of the non-colla-genous molecules discussed above. Mature ce-mentum has a high percentage of type I collagen,with less amounts reported for other collagens, in-cluding type III and type XII collagen. The smallamount of other collagens identified in cementummay be produced by periodontal ligament fibro-blasts, where the formed collagen fibers are insertedinto cementum.

Factor-mediated cell activities

A key for designing agents for use in regeneration oftissues is an understanding of the molecular conse-quences resulting from interactions between factorsand corresponding receptors. Substantial progresshas been made in understanding of regulators of cellfunction in health, as well as in diseased states. How-ever, until recently, few studies have been performedwith cells considered to be associated with root de-velopment, such as epithelial root sheath cells, fol-licle cells and root surface cells/cementoblasts dueto lack of availability of such cell types. In contrast,more information is available on cells associatedwith the mature periodontium, such as osteoblastsand periodontal ligament cells. In this section funda-mental concepts related to ligand–receptor interac-tions are presented as a means of stressing the im-

81

portance, as well as the complexities, for under-standing basic mechanisms regulating cell activitiesto designing of regenerative therapies. For example,using a hypothetical model, it may be possible thatfactor A is capable of promoting mineralization ofmature osteoblasts but not of periodontal fibroblastsor preosteoblasts. Further, the reason for this may berelated to the lack of appropriate receptors on thelatter cells; however, addition of factor B may primethese cells so that they now express the receptor re-quired for responding to factor A.

Provided below is a broad overview of currentknowledge as to growth factor–cell interactions, witha discussion on how this may relate to control of de-velopment and/or regeneration of periodontaltissues. For in-depth reviews of this area, the readeris referred to Sastry & Horwitz (198); Lafrenie & Yam-ada (123); Force & Bonventre (68); and Ingber (109).

Growth factors have numerous effects on cells de-pending on both the factor and cell type and stage ofmaturation. For example, growth factors can initiateDNA synthesis, modulate differentiation and alterthe cytoskeleton. The short half-life of growth factorsand their association with extracellular matrix andgrowth factor–binding proteins ensure their local ef-fects. The extracellular matrix molecules and growthfactors exert effects through specific cell surface re-ceptors, and when the receptor is bound it interactswith cytoplasmic effector molecules to initiate acomplex cascade of intracellular events leading to analteration in gene function (see Fig. 3 and 4 for anoverview of selected pathways).

Receptor tyrosine kinases

Growth factor receptors are prototypic members ofa family of cell surface receptors characterized by anextracellular binding domain, a single transmem-brane portion and a large intracellular catalytic do-main. Specifically, those that possess intrinsic tyro-sine kinase domains belong to the hydrophilic re-ceptor family and are referred as receptor tyrosinekinases. The signaling molecules for this group of re-ceptors include epidermal growth factor, insulin-likegrowth factor, fibroblast growth factor, platelet-de-rived growth factor, nerve growth factor and insulin.As the ligand binds to the receptor, the kinase do-main is autophosphorylated and transmission of themolecular signal from the ligand to the cytoplasmicdomain of the receptor occurs by a dimerization ofthe receptor complex (161). Receptor activation trig-gers intracellular cascades of protein phosphoryla-tions that promote protein interactions. These inter-

Saygin et al.

actions result in the transmission of specific signalsto the nucleus capable of controlling gene express-ion, with ultimate alteration of cell function. A goodexample that demonstrates the complexities of theseactivities is the mitogen-activated protein kinasepathway. Mitogen-activated protein kinase phos-phorylates transcription factors resulting in the acti-vation of specific genes. Mitogen-activated proteinkinase has 3 isoforms (mitogen-activated protein ki-nase and extracellular signal–regulated kinase-1and -2), which can be activated in response togrowth factors and other mitogens. Extracellular sig-nal–regulated kinase-1 and -2 proteins are presentin human osteoblasts, human bone marrow stromalcells, rat osteoblastic cells (ROS 17/2.8 and UMR-106), and mouse osteoblastic cells (MC3T3-E1) (34).Insulin-like growth factor-I, fibroblast growth factor-2 and platelet-derived growth factor-BB were foundto activate extracellular signal–regulated kinase-2.Receptor tyrosine phosphorylation also promotes in-teractions with a number of other molecules such asphospholipase-Cg, non-receptor Src family of ki-

Fig. 3. Putative receptor-cementoblast interactions. derived growth factor; IGF: insulin-like growth factor;Ligand-receptor binding results in activation of effector FGF: fibroblast growth factor; PG: prostaglandin; PTH/molecules that act as intracellular relay systems to trigger PTHrP: parathyroid hormone/parathyroid hormone–re-gene expression. Abbreviations: TGF: transforming growth lated protein. Adapted from Fuller & Shields (71).factor; BMP: bone morphogenetic protein; PDGF: platelet-

82

nases, tyrosine phosphatases and phosphoinositide-3-kinase (30, 71).

Several growth factors have been shown to inter-act with periodontal fibroblasts and/or osteoblastsand these include members of the transforminggrowth factor-b super family, platelet-derived growthfactor, insulin-like growth factor, epidermal growthfactor, fibroblast growth factor, prostaglandin E, andparathyroid hormone/parathyroid hormone–relatedprotein. In 1991 Narayanan’s group isolated a factorfrom cementum, which they called cementum de-rived growth factor (164, 254). In subsequent studiesthey reported that cementum-derived growth factoris an insulin-like growth factor-I-like molecule (107).Insulin-like growth factor-I has been shown to po-tentiate the effects of several growth factors andother molecules on cell activity (81, 197). Similar toresponses noted with growth factors such as insulin-like growth factor-I, cementum-derived growth fac-tor was found to cause a transient increase in cyto-plasmic Ca2π concentration, promote phosphoinosi-tol-phosphate hydrolysis, activate the protein ki-

Molecular and cell biology of cementum

nase-C cascade and increase expression of cellularprotooncogenes in gingival fibroblasts (255).

Growth factor interactions with their specific re-ceptors, coupled with their complex interactionswith other molecules, including other growth factorsand integrins, result in a complex set of cell re-sponses (198). Thus, the effect of a growth factor ona specific cell in vitro may be very different from theeffects of that factor on another cell type in vitro orin fact on the same cell type in vivo. Few studieshave examined the role of growth factors in regulat-ing cementogenesis, where more extensive studieshave been performed using animal models, versusin vitro or in situ models, and these are discussedlater.

Examining literature on in situ models, usingradioautography and 14-day-old rats, Cho et al. (45)investigated the role of epidermal growth factor ondifferentiation of cementoblasts after injecting 125I-epidermal growth factor. They reported that duringdifferentiation of cementoblasts a very low level ofepidermal growth factor–binding sites were presenton the mesenchymal cells in dental follicle proper,precementoblasts and cementoblasts, indicating thelimited effect of epidermal growth factor on ce-mentoblast differentiation. However, preosteoblasts,prechondroblasts, perifollicular cells and matureperiodontal ligament fibroblasts exhibited epidermalgrowth factor–binding sites in vivo (44). Thus, whileinteractions between epidermal growth factor andepidermal growth factor-receptor are not involveddirectly in the regulation of cementoblast activities,these studies, as well as those by other groups (158,178, 179, 230), indicate that epidermal growth factorand associated interactions with other cells duringtooth/periodontal development are critical for for-mation of a functional periodontium.

G-protein-coupled receptors

G-protein-coupled receptors are associated with theability of growth factors to activate the mitogen-acti-vated protein kinase cascade and other tyrosine ki-nase pathways. G-proteins act as the initial effectoractivating substrate for multipass receptors. Themajor downstream effector molecules controlled byG-proteins are adenyl cyclase, phospholipase C,phospholipase A2, phosphoinositide 3-kinase, and b-adrenegic receptor kinase. The activated secondmessengers, such as cyclic adenosine monophos-phate, diacylglycerol, inositol triphosphate and cal-cium, are small molecules that amplify receptor-acti-vated signals. Further, many growth factor–G protein

83

Fig. 4: Growth factor signaling cascade via MAPK. Abbrevi-ations: TK: tyrosine kinase; SOS: son of sevenless; MAPK:mitogen-activated kinase; MAP2K: mitogen-activated ki-nase kinase; TF: transcription factor. Note: this is onlyone aspect of the GF-signaling pathway.

interactions, such as platelet-derived growth factorand insulin-like growth factor-II, are known to acti-vate the mitogen-activated protein-kinase pathway(148, 197). With regard to cementum, it is known thatcementoblasts express parathyroid hormone–relatedprotein receptors (228), and thus it is not surprisingthat both parathyroid hormone and parathyroid hor-mone–related protein promote an increase in cyclicadenosine monophosphate in these cells (56, 57,174). Parathyroid hormone/parathyroid hormone–related protein, through G-protein-linked receptors,evoke multiple parallel signaling events that includeactivation of adenyl cyclase (protein kinase-A path-way), phospholipase C (protein kinase-C pathway)and cytosolic free calcium transients (225).

Saygin et al.

Serine-threonine receptor kinases

Transforming growth factor-b superfamily moleculesare known to elicit their effects through interactionswith serine/threonine kinase receptors. Subsequentto their interactions with and activation of the recep-tor, a group of signaling molecules, Smads, are phos-phorylated selectively by bone morphogenetic pro-teins and form heteromeric complexes (195). Someof these Smads enter the nucleus, where they inducetranscriptional activation and subsequently, altercell behavior. Bone morphogenetic proteins havenumerous functions, including regulation of cellgrowth, differentiation and apoptosis in a variety ofcell types, such as osteoblasts, neural cells, epithelialcells (184, 195, 201, 229, 232, 239); however, their ex-act role in cementum formation is only beginning tobe explored (1). Their potential role in regenerationof periodontal tissues is discussed later.

Integrins

A required event for development and regenerationof cementum is attachment of appropriate cells onthe root surface. While several adhesion moleculeshave been identified on the root surface at variousstages of root development, such as fibronectin, la-minin, type I collagen, bone sialoprotein, cementumattachment protein and osteopontin (Table 1, Fig. 1),the importance of these molecules for controllingcell adhesion and differentiation during tooth rootformation are not known. Further, certain growthfactors that have been implicated in controlling celladhesion, for example, platelet-derived growth fac-tor-BB and insulin-like growth factor-I, have beenshown to regulate expression of integrins (40) andproteoglycans (87). Many adhesion molecules inter-act with cells through specific integrins on the cellsurface. Integrin–extracellular matrix binding acti-vates signal transduction pathways and hence regu-lates gene expression (71, 130). Particularly centralto integrin signaling are focal adhesion kinases,which have been reported to bind directly to inte-grins. A number of other signaling molecules thenbind to focal adhesion kinases and are phosphoryl-ated by it, linking focal adhesion kinases to the mito-gen-activated protein kinase pathway and to the 85-kDa subunit of phosphoinositide 3-kinase (189),thus linking the integrin signaling pathway with sev-eral other pathways.

Few studies have focused on the role of these inte-grins during tooth development (196, 252), and thusthe specific integrins required for cell adhesion at

84

the local root surface site, as well as the ligands in-volved, remain unknown. Studies by Saito & Naray-anan (193), demonstrated that molecules extractedfrom mature cementum that promote adhesion offibroblasts induce characteristic signaling events,such as activation of c-fos, focal adhesion kinasesand extracellular signal–regulated kinase-2. It ispossible that such molecules may play a role in re-cruitment of specific cell types to the local site dur-ing wound healing. Future studies directed at deter-mining the specific adhesion molecules and sig-naling pathways regulating cementoblast maturationwill aid in enhancing understanding of root develop-ment and thus in the design of appropriate therapiesfor activating cementoblasts.

Regenerative therapies andcementogenesis in vivo

This section focuses on preclinical and clinical pro-gress toward using growth factors for stimulatingperiodontal regeneration, with an emphasis on ce-mentum regeneration. Regulators of periodontaltissue regeneration that stimulate formation of bone,periodontal ligament and cementum include manydifferent agents categorized as follows: I) cell occlus-ive membranes (including guided tissue regenera-tion) (242); II) bone replacement grafts (such asautografts, allografts, xenografts and alloplasts)(157); III) root conditioning agents (such as citricacid, ethylenediaminetetraacetic acid) (75); and IV)growth and attachment factors (such as bone mor-phogenetic proteins, platelet-derived growth factorand enamel matrix derivative) (88, 150).

A multitude of studies testing therapies to stimu-late periodontal repair have been published. How-ever, the evidence to support the use of these therap-ies as modulators of complete periodontal regenera-tion is quite limited. Table 2 highlights studiesfocusing on therapeutics measuring cementogenesisin vivo in both preclinical and clinical settings.

The results of preclinical and clinical studies willbe reviewed, and the ability of these molecules tonot only stimulate cementogenesis but also governbone and periodontal ligament regeneration will bediscussed.

Therapies based on platelet-derived growth factorand insulin-like growth factor

Some of the earliest in vivo studies assessing the roleof growth factors on periodontal regeneration fo-cused on a combination of platelet-derived growth

Molecular and cell biology of cementum

Table 2. Agents demonstrated to promote cementogenesis in preclinical and/or clinical studies

Evidence of cementogenesis

Therapy Preclinical (animal) Human histology References

Growth factorsPlatelet-derived growth factor or platelet-derived Yes No (79, 100, 134, 190)growth factor/insulin-like growth factor-IBone morphogenetic protein 2 Yes No (118, 207)Bone morphogenetic protein 3 No Yes (23)Bone morphogenetic protein 4 Yes No (186)Bone morphogenetic protein 7 (OP-1) Yes No (80)

Bone allografts Yes Yes (24, 25, 26, 183)

Xenogenic bone grafts Yes Yes (32)

Autogenous bone grafts Yes Yes (70, 97)

Citric acid demineralization Yes Yes (3)

Enamel matrix derivative Yes Yes (88, 94)

factor alone or combined with insulin-like growthfactor-I. Results using natural disease lesions in dogsand ligature-induced lesions in nonhuman primatesshowed that this growth factor combination pro-moted formation of new bone, cementum and peri-odontal ligament (79, 134, 190).

Park et al. reported results in a dog model usingplatelet-derived growth factor–modulated guidedtissue regeneration therapy. Findings from this studydemonstrated the promotion of new bone, ce-mentum and periodontal ligament, measured at 5and 11 weeks after platelet-derived growth factor/guided tissue regeneration treatment in class III fur-cation defects (177). Interestingly, when platelet-de-rived growth factor was applied to tooth roots di-rectly (following citric acid demineralization of theroot surface without the use of a barrier membrane),extensive ankylosis ensued (46). The ankylotic unionof bone to tooth without intervening cementum andperiodontal ligament was found in 100% of thespecimens evaluated at 5 weeks and 83% of thespecimens at 8 weeks. Further investigations areneeded to explore this result, since platelet-derivedgrowth factor-induced ankylosis has not been de-scribed in other studies.

The first human clinical trial testing the safetyand efficacy of recombinant human platelet-de-rived growth factor/recombinant human insulin-like growth factor-I was completed in 1997 (100).This study examined 38 patients possessing moder-ate to severe periodontal disease treated with a)150 mg/ml each of recombinant human platelet-de-rived growth factor-BB and recombinant human in-sulin-like growth factor-I in a methylcellulose ve-hicle or b) vehicle alone or c) surgery alone. Theresults revealed patients treated with recombinant

85

human platelet-derived growth factor/recombinanthuman insulin-like growth factor-I responded with42.5% osseous defect fill, while the control groupconsisting of pooled vehicle and surgery alonedemonstrated only 18.5% osseous defect fill. Thegrowth factors were shown to be safe and well tol-erated by the subjects. Furcation lesions respondedmost favorably to treatment with nearly a four-foldincrease in bone volume compared with pairedcontrols. These results in humans were found tobe highly consistent when compared to preclinicalstudies in nonhuman primates (77). Ongoingstudies, using these growth factors with a varietyof delivery systems, should provide the informationrequired to determine the clinical feasibility ofusing platelet-derived growth factor/insulin-likegrowth factor-I for regeneration of periodontaltissues in humans.

Transforming growth factor-b family members:bone morphogenetic proteins 2, 3 (osteogenin),4, and 7 (OP-1)

The bone morphogenetic proteins have been evalu-ated extensively in orthopedic models for their abil-ity to induce osteogenesis (181). Bone morphogen-etic protein-2 is the most thoroughly researchedmember of the transforming growth factor-b super-family for the promotion of periodontal and peri-implant bone regeneration (48, 91, 92, 99, 118, 191,206–208). This molecule has demonstrated potenteffects in stimulating cementogenesis. Sigurdssonet al. reported the effects of recombinant humanbone morphogenetic protein-2 on periodontal re-generation and found that bone morphogeneticprotein-2 applied in synthetic bioabsorbable poly-

Saygin et al.

mer greatly stimulated new bone and cementumformation (207). These results were achieved 8weeks following bone morphogenetic protein-2 ap-plication. Close to 95% of the bone in surgicallycreated class III furcation lesions was regenerated.However, a near 4-fold increase in ankylosis wasfound in bone morphogenetic protein-2-treatedsites as compared to vehicle. Ripamonti et al. dem-onstrated potent stimulation of cementum andbone regeneration in class II furcation defects inbaboons using topical bone morphogenetic pro-tein-4 application (186).

The first human study using a bone morphogen-etic protein to promote periodontal regenerationutilized a single application of bone morphogeneticprotein-3 (osteogenin) combined with demineral-ized bone allograft in a submerged tooth model(23). The investigators found increased bone andcementum deposition around periodontally in-volved submerged teeth with bone morphogeneticprotein-3 treatment as assessed by human his-tology. However, the bone morphogenetic protein-3 was not significantly better than carrier alone(bone graft group). Additionally, pinpoint ankylosiswas observed in submerged teeth grafted with bonemorphogenetic protein-3 plus bone grafts. More re-cently, our group demonstrated closure of class IIIfurcation defects with the application of bone mor-phogenetic protein-7/OP-1 applied to periodontaldefects, using an animal model (80). Heightenedstimulation of the formation of cellular ‘‘regenera-tive cementum’’ could be noted at 8 weeks postbone morphogenetic protein-7/OP-1 application.Interestingly, some root surfaces exhibited dentinresorptive pits with regenerative cementum form-ing over the irregular root surface (Fig. 5). Anotherinteresting observation was that ankylosis was notobserved in areas where regenerative cementumoccurred. While this needs to be examined morecarefully, this suggests that identification of factorsthat can promote new mineralization consistentlyon the root surface is an important area for re-search development.

Transforming growth factor-b

The first published report examining transforminggrowth factor-b in periodontal defects examined itsrole in combination with insulin-like growth factor-II and basic fibroblast growth factor (200). The re-sults in surgically created fenestration defects indogs failed to show a benefit by the applicationof these factors. The authors reported the use of

86

nanogram quantities of growth factor, which mayin part explain the negative results. More recently,Wikesjö et al. reported the use of transforminggrowth factor-b in periodontal repair and stimula-tion of cementum formation (244). They tested theability of transforming growth factor-b1 coupledwith expanded polytetrafluoroethylene barriers tostimulate periodontal regeneration. The results ofthe study failed to reveal any benefit with the com-bined therapy. However, the authors did not use agroup containing transforming growth factor-bwithout barriers, and the exclusion of periostealcells may have been a factor in results achieved.Cochran et al. demonstrated an inhibition of bonepromotion around implant fixtures when bonemorphogenetic protein-2 was used in combinationwith barrier membranes (48). Mixed reports on theability of transforming growth factor-b to promoteregeneration of mineralized tissues in orthopedicmodels have been reported as well (33). At thepresent time, transforming growth factor-b needs tobe further explored for its potential use in recon-structive periodontal therapy.

Enamel matrix derivative

Enamel matrix derivative, where the principal pro-tein is amelogenin, is approved for human use inter-nationally. Enamel matrix derivative has been evalu-ated, in a variety of preclinical and clinical settings,for its ability to stimulate cementum and peri-odontal attachment structures. In a preclinical buc-cal dehiscence model in nonhuman primates, Ham-marström et al. tested the ability of enamel matrixderivative to affect periodontal wound healing (90).When compared with carrier and ethylenedi-aminetetraacetic acid root conditioning, enamel ma-trix derivative treated defects demonstrated signifi-cant enhancements of bone, cementum and peri-odontal ligament. In human studies, enamel matrixderivative has demonstrated partial regeneration byhuman histology (94), safety in a multicenter trial of10 test centers and 107 patients, and efficacy asshown by a placebo-controlled human trial using 33subjects with paired intrabony defects (95). The laterstudy assessed enamel matrix derivative therapycoupled with modified Widman flap surgery on peri-odontal wound healing as measured by clinicalattachment level change and subtraction radi-ography (95). The enamel matrix derivative therapypromoted 66% defect fill 36 months post-therapy,while paired control defects failed to show a changein radiographic bone level. Thus, this study con-

Molecular and cell biology of cementum

Fig. 5. BMP7/OP-1 stimulates cementogenesis in vivo: larly from the root surface are anchored to the adjacentA. A representative lesion treated with 7.5 mg/g recombin- alveolus (Goldner’s, ¿200 magnification). C. Occasionalant human OP-1 in a collagen carrier. Pronounced new resorptive pits found on the dentinal surface contiguousbone formation can be seen with numerous embedded with the adjacent cemental dentinal interface (Goldner’s,osteocytes in the mineralized matrix (Goldner’s, ¿40 mag- ¿400 magnification). The interface in C can be seen undernification). B. Cellular regenerative cementum at high ultraviolet light to further demonstrate the cemental/den-magnification of the lower box in A. A thick layer of ce- tinal line of reversal (ultraviolet unstained section, ¿400mentum with collagen-like fibers oriented perpendicu- magnification).

cluded that topical enamel matrix derivative appli-cation in conjunction with modified Widman flapstimulates periodontal regeneration and is stable forup to 36 months. Expanded studies in larger patientpopulations will be needed to further assess the ef-fects of enamel matrix derivative treatment in intra-bony defects.

Delivery systems

The rate limiting step in re-engineering of peri-odontal structures (including cementum) is the tar-geting of agents/cells to the tooth root surface. Thehealing of a periodontal wound is complicated byseveral factors that limit predictable delivery ofagents to the root surface, such as: 1) the transmu-cosal environment of the mineralized tooth surfacetraversing keratinized gingiva; 2) a complex micro-biota contaminates wounds at the soft-hard tissueinterface and may affect the release kinetics of deliv-ery devices; 3) occlusal forces on the tooth complexin transverse and axial planes may modulate thehealing response and disrupt the stability of deliverydevices; and 4) the complexity of the attachment ap-paratus includes several stromal/cellular interac-tions that will require proper orientation of vehicleswithin these wounds. Hence, the use of various poly-mer delivery systems will need to circumvent thesechallenges. To date, various biodegradable polymericmatrices have been developed for use in guidedtissue regeneration (242). The above characteristicsof periodontal wounds place many limitations onthe successful use of these devices. For the delivery

87

of bioactive molecules including growth factors,genes or cells, materials such as copolymers of poly-glycolic acid and polylactic acid have been utilizedas delivery vehicles in medicine and dentistry. Amajor focus of research in tissue transplantation,growth factor and gene transfer is the developmentof novel delivery systems to allow the extended re-lease of these agents to the target tissue (tooth rootsurface).

Models to study cementogenesis

Investigations targeted at understanding the cellularand molecular mechanisms controlling develop-ment and regeneration of periodontal tissues haveutilized both in vitro and in vivo models. Commonlyused animals for regenerative studies include ro-dents, canines, felines and nonhuman primates (78,122, 185, 190). Investigations focused on determin-ing regulators of tooth/periodontal developmentoften use rodents and follow molar and/or incisordevelopment (5, 29, 56, 96, 135, 140, 143, 187, 192,232, 236, 237, 240, 253). In vitro models include cellcultures where cells are obtained from animal tissuesand cells can be manipulated in various ways tomimic the in vivo environment, such as growingcells on or within selective matrices or using a var-iety of co-culture models (49, 69, 117, 132, 168).

While in vivo models reflect the complexities ofhost-cell interactions, and thus may more accuratelyreflect the regenerative activities in humans (80),versus in vitro models, there are significant limi-tations. Thus, there is a need for both in vitro and in

Saygin et al.

vivo studies, where in vitro models, using cell sys-tems, provide the tools required to understand theresponse of cells to specific factors and the molecu-lar factors controlling these responses, and in vivomodels provide the tools required to establish ‘‘proofof concept’’. Limitations of in vivo models includethe inability to design a model that harbors the exacttype of defect seen in human disease. Further, manymodels use an acute defect which, while valuable fordetermining whether a factor does elicit a response,may not reflect the response for chronic situations,such as those found with most individuals who haveperiodontal disease. Moreover there are distinct gen-etic, anatomic, biochemical, immune and microbialdifferences between the species. In attempts to over-come these drawbacks genetically engineered ani-mals, as well as cells in vitro and in vivo are beingused. Discussed below are animal and cell modelsthat are being used to understand periodontal dis-ease or that have potential for use in future studies.

In vitro models

Excellent studies at the light and electron micro-scopic level have provided a detailed analysis of ce-mentum at various stages of development and also,during regeneration of periodontal tissues sub-sequent to disease (20, 21). Further, immunocyto-chemical studies and in situ hybridization studieshave provided information as to the factors ex-pressed by cells associated with the periodontium.However, these studies present only indirect infor-mation as to the factors critical to formation of theperiodontium (19, 22, 29, 56, 137, 138, 140, 142, 144,153, 216, 217, 227). As discussed later, transgenic/knock-out animals offer another tool to assist in de-termining the role of specific molecules in control-ling tissue function, but with limitations since suchanimals often do not survive, or show no phenotype,or show a complex phenotype that requires furtheranalysis. Cell cultures provide an additional toolwhere advances in cell and molecular techniquesallow for selected manipulation of specific cell types.In addition, cells isolated in culture can be reintro-duced into a specific site in animals and the activityof the cell type confirmed in vivo.

An important aspect of cell culture is to be able todescribe the cell phenotype at some level prior toisolation, since with passage cells in culture oftenlose their phenotype. In this regard, studies focusedon establishing factors expressed by cementoblastsand precementoblasts (follicle cells) in situ allowedfor the isolation of cementoblasts where cells in vitro

88

could be analyzed to ensure that they expressed thesame genes identified for these cells prior to iso-lation (see Fig. 1, selective stage markers). Fortu-nately, using these tools our laboratory has success-fully isolated and cultured cells from the developingroot surface of mice using a variety of techniques(54–57, 212). Briefly, as a first approach, a mixedpopulation of periodontal ligament cells and ce-mentoblasts were isolated by collagenase-trypsin di-gestion, and cells were shown to exhibit propertiesassociated with these cells in vivo, however cells didnot survive continued passages (56). Therefore, sev-eral approaches were used to establish cells thatwould survive passage and maintain phenotype, in-cluding immortalization of primary cell culturesusing wild-type SV40 (28), using immortalized trans-genic mice, ‘‘immorto-mice’’ (54, 57, 111, 112, 212),and using osteocalcin promoter–driven SV40 Tagmice (35). The advantage of the latter is that onlyroot surface cells expressing osteocalcin, cemento-blasts, will survive in vitro, thus excluding peri-odontal ligament cells from the population. The het-erogeneous cell populations also are of value sincethey reflect the local environment in vivo and bysubcloning mixed populations, both periodontalligament cell lines and cementoblast populationscan be established. Cementoblasts in vitro expresstranscripts for bone sialoprotein, osteocalcin, osteo-pontin, alkaline phosphatase, osteoblast-specifictranscription factor, parathyorid hormone/para-thyroid hormone–related protein receptor 1 and typeI collagen (54, 56, 57). In addition, these cells re-spond to parathyroid hormone and vitamin D in afashion similar to that noted for osteoblasts and alsopromote mineral nodule formation both in vitro andin vivo (54). Additionally, these cells were found tobe responsive to several growth factors, includingplatelet-derived growth factor and insulin-likegrowth factor (unpublished data).

In vitro models with human cementoblasts, as re-ported by Grzesik et al. (86), will provide additionaltools for understanding the behavior of periodontaltissue. While several groups have been successful inisolating and characterizing cultured human peri-odontal ligament cells, it is known that these aremixed populations, and also that they change pheno-type with culture. Thus, using a similar approach offirst establishing cell phenotype in vivo and then isol-ating, immortalizing and subcloning human popula-tions will allow for studies on these cells. Unfortu-nately, markers selective for periodontal ligamentcells have not been established. Nevertheless, immor-talized periodontal ligament cell lines are available,

Molecular and cell biology of cementum

where future studies will help to clarify if these cellsreflect the in vivo cell type (98).

In addition, cells from a human cementoblastomawere isolated and cultured. However, since the ‘‘ce-mentoblastoma’’ is poorly defined, the exact cellsobtained for culture are unclear. Cells cultured fromthis tumor produced bone sialoprotein, cementumattachment protein and collagen types I and V andalso promote mineralized nodules in vitro (6, 7). Asspecific markers for cementoblasts are established,these cells may prove to be an excellent source forhuman cells. Importantly, with the availability of cellcultures, critical issues regarding the mechanismsand factors regulating cementoblast function, in-cluding identification of genes expressed selective bycementoblasts, can be addressed. Beyond determin-ing the properties of cementoblasts in vitro, thesecells can be used for targeted gene therapy as dis-cussed later (166).

In vivo: genetically engineered animals

Reverse genetic techniques, including gene knock-outs and transgenesis, allow defined mutations to beintroduced into the mouse genome and provide in-sight into gene function (84, 106, 205). The mouse isparticularly useful as its genome is very well char-acterized, and genes can therefore be manipulatedwith relative ease (72). Understanding the factors con-trolling periodontal diseases and specifically ce-mentogenesis may benefit from the use of transgenicor knock-out animals, as well as animals with spon-taneous mutations that reflect systemic diseases orsyndromes associated with alterations in periodontaltissues, such as Paget’s disease, hypophosphatasia,

Table 3. Methods for delivering genes into mammalian cells and likely applications in gene therapy

Application in gene therapy Transient orMethod Ex vivo In vivo stable expression

ViralRetrovirus π ? StableAdenovirus ∫ π TransientAdeno-associated virus π ? StableHerpes virus ∫ π Not predictableVaccina virus ∫ π TransientPolio virus ∫ π TransientSindbis or other RNA viruses ∫ π Transient

NonviralLigand-DNA conjugates ª π TransientAdenovirus-ligand-DNA conjugates ª π TransientLipofection ∫ π TransientDirect injection of DNA ª π TransientCaPO4 precipitation ∫ ª Stable

(π): Major application, (∫): some application. (ª): little or no application. (?): safety concern. Adapted from Mulligan (162).

89

cleidocranial dysplasia, hypopituitarism and osteo-petrosis. Animal models demonstrating alterations intooth structure most often are reported as failure oftooth eruption, where the major defect is related tolack of osteoclast activity, and include op/op mice (2,64, 163, 222), c-fos knock-outs (110, 114, 243), srcknock-outs (27, 220), and osteoprotegerin ligandknock-outs (120, 223). Rescue attempts toward re-placement of missing factors have provided a meansof maintaining animals that would normally not sur-vive. For example, parathyroid hormone/parathyroidhormone–related protein knock-out mice have severealterations in skeletal development and do not sur-vive (66, 115, 124). Rescue of these animals by a trans-gene for chondrocytic parathyroid hormone–relatedprotein expression failed to recover eruption and rootformation, while rescue using a keratin-driven trans-gene/parathyroid hormone–related protein correctedthe defect of eruption and tooth formation (180).

Another area of focus has been to overexpress orunderexpress selective genes at various stages of de-velopment and then to determine the effect of thesemanipulations on animal function. Ibaraki-O’Conn-or et al. (105) have established a transgenic mouseline to investigate the function of amelogenin duringmineralization. Importantly, future studies using thiscell line may provide insight as to the role of amelog-enin-like molecules, including enamel matrix deriva-tive, in controlling cells associated with the peri-odontium.

Cell applications: regional gene therapy

The application of putative molecules to induce ormodulate periodontal regeneration is an area of in-

Saygin et al.

tensive interest. However the short half-lives of thesemolecules at the healing site may reduce their effectsin vivo. Therefore, methods that provide stability ofexogenous molecules at the healing site may be ad-vantageous toward maximizing wound repair. Theability to transfer genetic material into specific cellsoffers an approach that may help sustain the effectof the targeted factor. Several investigators have fo-cused on developing ideal methods for delivery ofgenes to cells for subsequent use in in vivo models.

Gene transfer can be performed with strategies foreither ex vivo or in vivo transfer of the desired trans-gene. In ex vivo transfer, cDNA is transferred to cells inculture, and the genetically modified cells are ex-panded and then administered to the recipient site(162). For example, bone marrow cells can be re-moved from an individual, transduced ex vivo andthen the genetically modified cells reimplanted to thesite of interest (127). The alternative is the in vivotechnique, where the gene is transferred directly intothe target tissues of the recipient by microseeding ofplasmid or viral DNA (77). The attractiveness of usingregional gene therapy to induce repair and formationis that genes can be delivered to the appropriate ana-tomic site, and the duration of protein expression canbe determined by selecting the appropriate vectorand/or promoter (162). Both viral and nonviralmethods are used for gene transfer (Table 3). In thecase of retroviral and adeno-associated viral vectors,the transferred DNA sequences are stably integratedinto the chromosomal DNA of the target cell. In gen-eral, this approach is used for ex vivo application.However, in certain situations, gene expression is highbut transient, and such conditions favor in vivo trans-fer. Although retroviral gene transfer is ideal for ex vivoapplications, several features of the gene transfermethod may limit its applicability, particularly withregard to in vivo applications. The entry of the retro-virus depends on the existence of the viral receptor. Inaddition, the replication of the target cell is necessaryfor proviral integration to occur.

The most important advance for viruses as genetransfer vectors was the generation of ‘‘packagingcells’’ that permit the production of high titers of repli-cation-defective recombinant virus, free of wild-typevirus; also called gutless viruses (63, 93, 131, 160).Adenoviruses are capable of efficiently infecting non-dividing and dividing cells and expressing largeamounts of gene products (162). The transient natureof adenoviruses allows them to be used in high titerssince their effect will gradually diminish and remainunintegrated (as extrachromosomal DNA) due in partto a T-cell-mediated immune response.

90

The gene transfer approach presents an attractivealternative to the conventional topical application ofmolecules to the complex periodontal wound site.The use of viral constructs containing one or moreresponse modifier transgenes for long-term deliverymay modulate the cellular response in the periodon-tium. For these studies, the first step is to determinethe gene transfer efficiency of putative cells isolatedfrom the periodontium, such as periodontal liga-ment cells or cementoblasts. Although the mechan-isms that coordinate major regenerative events re-main obscure, migration, proliferation, attachment,differentiation and maturation of participating cellscan be modulated by gene transfer methods. Factorsthat modulate angiogenesis, proliferation, attach-ment and differentiation of the cells may be deliver-ed into the wound by gene therapy. The use of genetransfer techniques should enable one to modulateperiodontal regeneration, as well as assist in en-hancing understanding of the mechanisms involvedin wound healing.

Future directions: unknowns

This is a dynamic time for researchers and cliniciansdevoted to optimizing periodontal/implant regene-rative therapies. The explosion in understanding ofregulators of cell function, coupled with tools thatallow researchers to engineer cells so as to expressspecific factors, added to improved delivery systemsfor controlling release of cells/factors at a given site,now allows treatment modalities to be designedbased on sound scientific data. Clearly, a first step isevaluating engineered cell types/delivery systems inin vitro and in vivo models. This would includeexamining their ability to enhance cell proliferationand/or mineral nodule formation and/or specific os-teoblast- and cementoblast-associated genes in vitroand formation of new bone, new root surface min-eral and a functional periodontal ligament region invivo. Information gained from these studies shouldprovide the foundation required for designing morepredictable regenerative therapies when comparedwith those available at present.

References

1. Aberg T, Wozney J, Thesleff I. Expression patterns of bonemorphogenetic proteins (BMPs) in the developing mousetooth suggest roles in morphogenesis and cell differen-tiation. Dev Dyn 1997: 210: 383–396.

Molecular and cell biology of cementum

2. Aharinejad S, Marks SC Jr, Bock P, Mason-Savas A,MacKay CA, Larson EK, Jackson ME, Luftensteiner M,Wiesbauer E. CSF-1 treatment promotes angiogenesis inthe metaphysis of osteopetrotic (toothless, tl) rats. Bone1995: 16: 315–324.

3. Albair WB, Cobb CM, Killoy WJ. Connective tissue attach-ment to periodontally diseased roots after citric acid de-mineralization. J Periodontol 1982: 53: 515–526.

4. Amar S. Implications of cellular and molecular biologyadvances in periodontal regeneration. Anat Rec 1996: 245:361–373.

5. Andujar MB, Couble P, Couble ML, Magloire H. Differen-tial expression of type I and type III collagen genes duringtooth development. Development 1991: 111: 691–698.

6. Arzate H, Alvarez-Perez MA, Aguilar-Mendoza ME, Alvar-ez-Fregoso O. Human cementum tumor cells have differ-ent features from human osteoblastic cells in vitro. J Peri-odontal Res 1998: 33: 249–258.

7. Arzate H, Olson SW, Page RC, Narayanan AS. Isolation ofhuman tumor cells that produce cementum proteins inculture. Bone Miner 1992: 18: 15–30.

8. Banerjee C, McCabe LR, Choi JY, Hiebert SW, Stein JL,Stein GS, Lian JB. Runt homology domain proteins in os-teoblast differentiation: AML3/CBFA1 is a major compo-nent of a bone-specific complex. J Cell Biochem 1997: 66:1–8.

9. Bar-Kana I, Savion N, Narayanan AS, Pitaru S. Cementumattachment protein manifestation is restricted to the min-eralized tissue forming cells of the periodontium. Eur JOral Sci 1998: 106(suppl 1): 357–364.

10. Bartold PM, Miki Y, McAllister B, Narayanan AS, Page RC.Glycosaminoglycans of human cementum. J PeriodontalRes 1988: 23: 13–17.

11. Bartold PM, Narayanan AS. Biology of the periodontal con-nective tissues. Chicago, IL: Quintessence, 1998: 173–195.

12. Bartold PM, Reinboth B, Nakae H, Narayanan AS, PageRC. Proteoglycans of bovine cementum: isolation andcharacterization. Matrix 1990: 10: 10–19.

13. Beck F, Tucci J, Russell A, Senior PV, Ferguson MW. Theexpression of the gene coding for parathyroid hormone-related protein (PTHrP) during tooth development in therat. Cell Tissue Res 1995: 280: 283–290.

14. Beertsen W, VandenBos T, Everts V. Root development inmice lacking functional tissue non-specific alkaline phos-phatase gene: inhibition of acellular cementum forma-tion. J Dent Res 1999: 78: 1221–1229.

15. Begue-Kirn C, Smith AJ, Loriot M, Kupferle C, Ruch JV,Lesot H. Comparative analysis of TGFbs, BMPs, IGF1,Msxs, fibronectin, osteonectin and bone sialoproteingene expression during normal and in vitro inducedodontoblast differentiation. Int J Dev Biol 1994: 38: 405–420.

16. Bianco P, Fisher LW, Young MF, Termine JD, Robey PG.Expression of bone sialoprotein (BSP) in developing hu-man tissues. Calcif Tissue Int 1991: 49: 421–426.

17. Boskey AL, Maresca M, Ullrich W, Doty SB, Butler WT,Prince CW. Osteopontin-hydroxyapatite interactions invitro: inhibition of hydroxyapatite formation and growthin a gelatin-gel. Bone Miner 1993: 22: 147–159.

18. Bosshardt DD, Nanci A. Morphological and immunocyto-chemical evidence for epithelial-mesenchymal trans-formation of Hertwig’s root sheath. J Dent Res 1996: 75:abstr 1943.

91

19. Bosshardt DD, Nanci A. Immunodetection of enamel- andcementum-related (bone) proteins at the enamel-freearea and cervical portion of the tooth in rat molars. J BoneMiner Res 1997: 12: 367–379.

20. Bosshardt DD, Schroeder HE. Cementogenesis reviewed:a comparison between human premolars and rodent mo-lars. Anat Rec 1996: 245: 267–292.

21. Bosshardt DD, Selvig KA. Dental cementum: the dynamictissue covering of the root. Periodontol 2000 1997: 13: 41–75.

22. Bosshardt DD, Zalzal S, McKee MD, Nanci A. Develop-mental appearance and distribution of bone sialoproteinand osteopontin in human and rat cementum. Anat Rec1998: 250: 13–33.

23. Bowers G, Felton F, Middleton C, Glynn D, Sharp S, Mel-lonig J, Corio R, Emerson J, Park S, Suzuki J, Ma S, Rom-bert E, Reddi AH. Histologic comparison of regenerationin human intrabony defects when osteogenin is com-bined with demineralized freeze-dried bone allograft andwith purified bovine collagen. J Periodontol 1991: 62: 690–702.

24. Bowers GM, Chadroff B, Carnevale R, Mellonig J, Corio R,Emerson J, Stevens M, Romberg E. Histologic evaluationof new attachment apparatus formation in humans. PartI. J Periodontol 1989: 60: 664–674.

25. Bowers GM, Chadroff B, Carnevale R, Mellonig J, Corio R,Emerson J, Stevens M, Romberg E. Histologic evaluationof new attachment apparatus formation in humans. PartII. J Periodontol 1989: 60: 675–682.

26. Bowers GM, Chadroff B, Carnevale R, Mellonig J, Corio R,Emerson J, Stevens M, Romberg E. Histologic evaluationof new attachment apparatus formation in humans. PartIII. J Periodontol 1989: 60: 683–693.

27. Boyce BF, Chen H, Soriano P, Mundy GR. Histomorpho-metric and immunocytochemical studies of src-relatedosteopetrosis. Bone 1993: 14: 335–340.

28. Brockman WW, Christensen JB, Ryan KW, Souwaidane M,Imperiale MJ. Fate and expression of simian virus 40 DNAafter introduction into murine cells under nonselectiveconditions. Virology 1987: 158: 118–125.

29. Bronckers AL, Farach-Carson MC, Van Waveren E, ButlerWT. Immunolocalization of osteopontin, osteocalcin, anddentin sialoprotein during dental root formation andearly cementogenesis in the rat. J Bone Miner Res 1994:9: 833–841.

30. Burgess AW, Thumwood CM. The Sixth George SwansonChristie Memorial Lecture: growth factors and their re-ceptors: new opportunities for cancer treatment. Pathol-ogy 1994: 26: 453–463.

31. Butler WT, Ridall AL, McKee MD. Osteopontin. In: Bilezik-ian JP, Raisz LG, Rodan GA, ed. Principles of bone biology.San Diego: Academic Press, 1996: 167–182.

32. Camelo M, Nevins ML, Schenk R, Rasperini G, Simion M,Lynch SE, Nevins M. Clinical, radiographic and histologicevaluation of human periodontal defects treated withBio-oss and Bio-Gide. Int J Periodontics Restorative Dent1998: 18: 321–331.

33. Centrella M, Horowitz MC, Wozney JM, McCarthy TL.Transforming growth factor-beta gene family membersand bone. Endocr Rev 1994: 15: 27–39.

34. Chaudhary LR, Avioli LV. Activation of extracellular signal-regulated kinases 1 and 2 (ERK1 and ERK2) by FGF-2 andPDGF-BB in normal human osteoblastic and bone mar-

Saygin et al.

row stromal cells: differences in mobility and in-gel renat-uration of ERK1 in human, rat, and mouse osteoblasticcells. Biochem Biophys Res Commun 1997: 238: 134–139.

35. Chen D, Chen H, Feng JQ, Windle JJ, Koop BA, Harris MA,Bonewald LF, Boyce BF, Wozney JM, Mundy GR, Harris SE.Osteoblastic cell lines derived from a transgenic mousecontaining the osteocalcin promoter driving SV40 T-anti-gen. Mol Cell Differentiation 1995: 3: 193–212.

36. Chen J, McCulloch CAG, Sodek J. Bone sialoprotein in de-veloping porcine dental tissues: cellular expression andcomparison of tissue localization with osteopontin andosteonectin. Arch Oral Biol 1993: 38: 241–249.

37. Chen J, Shapiro HS, Sodek J. Developmental expressionof bone sialoprotein mRNA in rat mineralized connectivetissues. J Bone Miner Res 1992: 7: 987–997.

38. Chen J, Singh K, Mukherjee BB, Sodek J. Developmentalexpression of osteopontin (OPN) mRNA in rat tissues: evi-dence for a role for OPN in bone formation and resorp-tion. Matrix 1993: 13: 113–123.

39. Chen JK, Shapiro HS, Wrana JL, Reimers S, Heersche JN,Sodek J. Localization of bone sialoprotein (BSP) express-ion to sites of mineralized tissue formation in fetal rattissues by in situ hybridization. Matrix 1991: 11: 133–143.

40. Chen Q, Kinch MS, Lin TH, Burridge K, Juliano RL. Inte-grin-mediated cell adhesion activates mitogen-activatedprotein kinases. J Biol Chem 1994: 269: 26602–26605.

41. Cheng H, Caterson B, Neame PJ, Lester GE, Yamauchi M.Differential distribution of lumican and fibromodulin intooth cementum. Connect Tissue Res 1996: 34: 87–96.

42. Cheng H, Caterson B, Yamauchi M. Identification and im-munolocalization of chondroitin sulfate proteoglycans intooth cementum. Connect Tissue Res 1999: 40: 37–47.

43. Cho MI, Garant PR. Radioautographic study of [3H]man-nose utilization during cementoblast differentiation, for-mation of acellular cementum, and development of peri-odontal ligament principal fibers. Anat Rec 1989: 223:209–222.

44. Cho MI, Garant PR. Expression and role of epidermalgrowth factor receptors during differentiation of ce-mentoblasts, osteoblasts, and periodontal ligamentfibroblasts in the rat. Anat Rec 1996: 245: 342–360.

45. Cho MI, Lin WL, Garant PR. Occurrence of epidermalgrowth factor-binding sites during differentiation of ce-mentoblasts and periodontal ligament fibroblasts of theyoung rat: a light and electron microscopic radioauto-graphic study. Anat Rec 1991: 231: 14–24.

46. Cho MI, Lin WL, Genco RJ. Platelet-derived growth factor-modulated guided tissue regenerative therapy. J Peri-odontol 1995: 66: 522–530.

47. Clayden AM, Young WG, Zhang CZ, Harbrow D, Ro-maniuk K, Waters MJ. Ultrastructure of cementogenesisas affected by growth hormone in the molar periodon-tium of the hypophysectomized rat. J Periodontal Res1994: 29: 266–275.

48. Cochran DL, Schenk R, Buser D, Wozney JM, Jones AA.Recombinant human bone morphogenetic protein-2stimulation of bone formation around endosseous dentalimplants. J Periodontol 1999: 70: 139–150.

49. Conway-Myers BA. Co-culture update: creating an em-bryotrophic environment in vitro. Semin Reprod Endocri-nol 1998: 16: 175–182.

50. Davidovitch Z, Musich D, Doyle M. Hormonal effects onorthodontic tooth movement in cats – a pilot study. Am JOrthod 1972: 62: 95–96.

92

51. Denhardt DT, Chambers AF. Overcoming obstacles tometastasis – defenses against host defenses: osteopontin(OPN) as a shield against attack by cytotoxic host cells. JCell Biochem 1994: 56: 48–51.

52. Denhardt DT, Lopez CA, Rollo EE, Hwang SM, An XR,Walther SE. Osteopontin-induced modifications of cellu-lar functions. Ann N Y Acad Sci 1995: 760: 127–142.

53. Denton GB. The discovery of cementum. J Dent Res 1939:18: 239–242.

54. D’Errico JA, Berry JE, Ouyang H, Strayhorn CL, Windle JJ,Somerman MJ. Employing a transgenic animal model toobtain cementoblasts, in vitro. J Periodontol 2000: 71: 63–72.

55. D’Errico JA, MacNeil RL, Strayhorn CL, Piotrowski BT,Somerman MJ. Models for the study of cementogenesis.Connect Tissue Res 1995: 33: 9–17.

56. D’Errico JA, MacNeil RL, Takata T, Berry J, Strayhorn C,Somerman MJ. Expression of bone associated markers bytooth root lining cells, in situ and in vitro. Bone 1997: 20:117–126.

57. D’Errico JA, Ouyang H, Berry JE, MacNeil RL, StrayhornCL, Imperiale MJ, Harris NL, Goldberg H, Somerman MJ.Immortalized cementoblasts and periodontal ligamentcells in culture. Bone 1999: 25: 39–47.

58. Dodds RA, Connor JR, James IE, Rykaczewski EL, Appel-baum E, Dul E, Gowen M. Human osteoclasts, not osteo-blasts, deposit osteopontin onto resorption surfaces: anin vitro and ex vivo study of remodeling bone. J BoneMiner Res 1995: 10: 1666–1680.

59. D’Souza RN, Flanders K, Butler WT. Colocalization of TGF-beta 1 and extracellular matrix proteins during rat toothdevelopment. Proc Finn Dent Soc 1992: 88: 419–426.

60. D’Souza RN, Happonen RP, Ritter NM, Butler WT. Tem-poral and spatial patterns of transforming growth factor-beta 1 expression in developing rat molars. Arch Oral Biol1990: 35: 957–965.

61. Ducy P, Desbois C, Boyce B, Pinero G, Story B, DunstanC, Smith E, Bonadio J, Goldstein S, Gundberg C, BradleyA, Karsenty G. Increased bone formation in osteocalcin-deficient mice. Nature 1996: 382: 448–452.

62. Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. Osf2/Cbfa1: a transcriptional activator of osteoblast differen-tiation. Cell 1997: 89: 747–754.

63. Duisit G, Salvetti A, Moullier P, Cosset FL. Functionalcharacterization of adenoviral/retroviral chimeric vectorsand their use for efficient screening of retroviral producercell lines. Hum Gene Ther 1999: 10: 189–200.

64. Felix R, Cecchini MG, Hofstetter W, Elford PR, Stutzer A,Fleisch H. Impairment of macrophage colony-stimulatingfactor production and lack of resident bone marrowmacrophages in the osteopetrotic op/op mouse. J BoneMiner Res 1990: 5: 781–789.

65. Feng B, Rollo EE, Denhardt DT. Osteopontin (OPN) mayfacilitate metastasis by protecting cells from macrophageNO-mediated cytotoxicity: evidence from cell lines down-regulated for OPN expression by a targeted ribozyme. ClinExp Metastasis 1995: 13: 453–462.

66. Foley J, Longely BJ, Wysolmerski JJ, Dreyer BE, BroadusAE, Philbrick WM. PTHrP regulates epidermal differen-tiation in adult mice. J Invest Dermatol 1998: 111: 1122–1128.

67. Fong CD, Slaby I, Hammarström L. Amelin: an enamel-related protein, transcribed in the cells of epithelial rootsheath. J Bone Miner Res 1996: 11: 892–898.

Molecular and cell biology of cementum

68. Force T, Bonventre JV. Growth factors and mitogen-acti-vated protein kinases. Hypertension 1998: 31: 152–161.

69. Friedman CD, Costantino PD, Takagi S, Chow LC.BoneSource hydroxyapatite cement: a novel biomaterialfor craniofacial skeletal tissue engineering and recon-struction. J Biomed Mater Res 1998: 43: 428–432.

70. Froum SJ, Thaler R, Scopp IW, Stahl SS. Osseous auto-grafts. II. Histological responses to osseous coagulum-bone blend grafts. J Periodontol 1975: 46: 656–661.

71. Fuller GM, Shields D. Molecular basis of medical cell bi-ology. 1st edn. Stamford, CT: Appleton & Lange, 1998:148–220.

72. Fuster V, Poon M, Willerson JT. Learning from the trans-genic mouse: endothelium, adhesive molecules, and neo-intimal formation. Circulation 1998: 97: 16–18.

73. Ganss B, Kim RH, Sodek J. Bone sialoprotein. Crit RevOral Biol Med 1999: 10: 79–98.

74. Gao J, Symons AL, Bartold PM. Expression of trans-forming growth factor-beta 1 (TGF-beta1) in the develop-ing periodontium of rats. J Dent Res 1998: 77: 1708–1716.

75. Garrett S. Periodontal regeneration around natural teeth.Ann Periodontol 1996: 1: 621–666.

76. Giachelli CM, Liaw L, Murry CE, Schwartz SM, AlmeidaM. Osteopontin expression in cardiovascular diseases.Ann N Y Acad Sci 1995: 760: 109–126.

77. Giannobile WV. Periodontal regeneration by polypeptidegrowth factors and gene transfer. In: Lynch SE, Genco RJ,Marx RE, ed. Tissue engineering: applications in maxillo-facial surgery and periodontics. Chicago, IL: Quintess-ence, 1999: 231–243.

78. Giannobile WV, Finkelman RD, Lynch SE. Comparison ofcanine and non-human primate animal models for peri-odontal regenerative therapy: results following a singleadministration of PDGF/IGF-I. J Periodontol 1994: 65:1158–1168.

79. Giannobile WV, Hernandez RA, Finkelman RD, Ryan S, Ki-ritsy CP, D’Andrea M, Lynch SE. Comparative effects ofplatelet-derived growth factor-BB and insulin-like growthfactor-I, individually and in combination, on periodontalregeneration in Macaca fascicularis. J Periodontal Res1996: 31: 301–312.

80. Giannobile WV, Ryan S, Shih MS, Su DL, Kaplan PL, ChanTC. Recombinant human osteogenic protein-1 (OP-1)stimulates periodontal wound healing in class III fur-cation defects. J Periodontol 1998: 69: 129–137.

81. Giannobile WV, Whitson SW, Lynch SE. Non-coordinatecontrol of bone formation displayed by growth factorcombinations with IGF-I. J Dent Res 1997: 76: 1569–1578.

82. Giniger MS, Norton L, Sousa S, Lorenzo JA, Bronner F. Ahuman periodontal ligament fibroblast clone releases abone resorption inhibition factor in vitro. J Dent Res 1991:70: 99–101.

83. Goldberg HA, Hunter GK. The inhibitory activity of osteo-pontin on hydroxyapatite formation in vitro. Ann N YAcad Sci 1995: 760: 305–308.

84. Gordon JW. Transgenic animals. Int Rev Cytol 1989: 115:171–229.

85. Green RJ, Usui ML, Hart CE, Ammons WF, Narayanan AS.Immunolocolization of platelet-derived growth factor Aand B chains and PDGF-alpha and beta receptors in hu-man gingival wounds. J Periodontal Res 1997: 32: 209–214.

86. Grzesik WJ, Kuzentsov SA, Uzawa K, Mankani M, Robey

93

PG, Yamauchi M. Normal human cementum-derivedcells: isolation, clonal expansion, and in vitro and in vivocharacterization. J Bone Miner Res 1998: 13: 1547–1554.

87. Haase HR, Clarkson RW, Waters MJ, Bartold PM. Growthfactor modulation of mitogenic responses and proteo-glycan synthesis by human periodontal fibroblasts. J CellPhysiol 1998: 174: 353–361.

88. Hammarström L. The role of enamel matrix proteins inthe development of cementum and periodontal tissues.Ciba Found Symp 1997: 205: 246–255.

89. Hammarström L, Alatli I, Fong CD. Origins of cementum.Oral Dis 1996: 2: 63–69.

90. Hammarström L, Heijl L, Gestrelius S. Periodontal re-generation in a buccal deshiscence model in monkeysafter application of enamel matrix proteins. J Clin Peri-odontol 1997: 24: 669–677.

91. Hanisch O, Tatakis DN, Boskovic MM, Rohrer MD,Wikesjö UM. Bone formation and reosseointegration inperi-implantitis defects following surgical implantation ofrhBMP-2. Int J Oral Maxillofac Implants 1997: 12: 604–610.

92. Hanisch O, Tatakis DN, Rohrer MD, Wohrle PS, WozneyJM, Wikesjö UM. Bone formation and osseointegrationstimulated by rhBMP-2 following subantral augmentationprocedures in nonhuman primates. Int J Oral MaxillofacImplants 1997: 12: 785–792.

93. Hardy S, Kitamura M, Harris-Stansil T, Dai Y, Phipps ML.Construction of adenovirus vectors through Cre-lox re-combination. J Virol 1997: 71: 1842–1849.

94. Heijl L. Periodontal regeneration with enamel matrix de-rivative in one human experimental defect. A case report.J Clin Periodontol 1997: 24: 693–696.

95. Heijl L, Heden G, Svardstrom G, Ostgren A. Enamel matrixderivative (EMDOGAIN) in the treatment of intrabonyperiodontal defects. J Clin Periodontol 1997: 24: 705–714.

96. Helder MN, Karg H, Bervoets TJ, Vukicevic S, Burger EH,D’Souza RN, Woltgens JH, Karsenty G, Bronckers AL.Bone morphogenetic protein-7 (osteogenic protein-1, OP-1) and tooth development. J Dent Res 1998: 77: 545–554.

97. Hiatt WH, Schallhorn RG. Intraoral transplants of cancel-lous bone and marrow in periodontal lesions. J Peri-odontol 1973: 44: 194–208.

98. Hoang AM, Chen D, Oates TW, Jiang C, Harris SE,Cochran DL. Development and characterization of atransformed human periodontal ligament cell line. J Peri-odontol 1997: 68: 1054–1062.

99. Howell TH, Fiorellini J, Jones A, Alder M, Nummikoski P,Lazaro M, Lilly L, Cochran D. A feasibility study evaluatingrhBMP-2/absorbable collagen sponge device for local al-veolar ridge preservation or augmentation. Int J Peri-odontics Restorative Dent 1997: 17: 124–139.

100. Howell TH, Fiorellini JP, Paquette DW, Offenbacher S, Gi-annobile WV, Lynch SE. A phase I/II clinical trial to evalu-ate a combination of recombinant human platelet-de-rived growth factor-BB and recombinant human insulin-like growth factor-I in patients with periodontal disease.J Periodontol 1997: 68: 1186–1193.

101. Hu CC, Fukae M, Uchida T, Qian Q, Zhang CH, Ryu OH,Tanabe T, Yamakoshi Y, Murakami C, Dohi N, Shimizu M,Simmer JP. Sheathlin: cloning, cDNA/polypeptide se-quences, and immunolocalization of porcine enamelsheath proteins. J Dent Res 1997: 76: 648–657.

102. Hunter GK, Goldberg HA. Nucleation of hydroxyapatite

Saygin et al.

by bone sialoprotein. Proc Natl Acad Sci U S A 1993: 90:8562–8565.

103. Hunter GK, Goldberg HA. Modulation of crystal forma-tion by bone phosphoproteins: role of glutamic acid-richsequences in the nucleation of hydroxyapatite by bonesialoprotein. Biochem J 1994: 302: 175–179.

104. Hunter GK, Kyle CL, Goldberg HA. Modulation of crystalformation by bone phosphoproteins: structural specificityof the osteopontin-mediated inhibition of hydroxyapatiteformation. Biochem J 1994: 300: 723–728.

105. Ibaraki-O’Connor K, Nakata K, Young MF. Toward under-standing the function of amelogenin using transgenicmice. Adv Dent Res 1996: 10: 208–214.

106. Ignelzi MA Jr, Liu YH, Maxson RE Jr, Snead ML. Genetic-ally engineered mice: tools to understand craniofacial de-velopment. Crit Rev Oral Biol Med 1995: 6: 181–201.

107. Ikezawa K, Hart CE, Williams DC, Narayanan AS. Char-acterization of cementum derived growth factor as an in-sulin-like growth factor-I like molecule. Connect TissueRes 1997: 36: 309–319.

108. Ikezawa K, Ohtsubo M, Norwood TH, Narayanan AS. Roleof cyclin E and cyclin E-dependent kinase in mitogenicstimulation by cementum-derived growth factor in hu-man fibroblasts. FASEB J 1998: 12: 1233–1239.

109. Ingber D. In search of cellular control: signal transductionin context. J Cell Biochem 1998: 31(suppl): 232–237.

110. Jacenko O. c-fos and bone loss: a proto-oncogene regu-lates osteoclast lineage determination. Bioessays 1995: 17:277–281.

111. Jat PS, Noble MD, Ataliotis P, Tanaka Y, Yannoutsos N,Larsen L, Kioussis D. Direct derivation of conditionallyimmortal cell lines from an H-2Kb- tsA58 transgenicmouse. Proc Natl Acad Sci U S A 1991: 88: 5096–5100.

112. Jat PS, Sharp PA. Cell lines established by a temperature-sensitive simian virus 40 large-T-antigen gene are growthrestricted at the nonpermissive temperature. Mol Cell Biol1989: 9: 1672–1681.

113. Jiang H, Sodek J, Karsenty G, Thomas H, Ranly D, Chen J.Expression of core binding factor Osf2/Cbfa-1 and bonesialoprotein in tooth development. Mech Dev 1999: 81:169–173.

114. Johnson RS, Spiegelman BM, Papaioannou V. Pleiotropiceffects of a null mutation in the c-fos proto-oncogene.Cell 1992: 71: 577–586.

115. Karaplis AC, Kronenberg HM. Physiological roles for para-thyroid hormone-related protein: lessons from geneknockout mice. Vitam Horm 1996: 52: 177–193.

116. Karimbux NY, Nishimura I. Temporal and spatial express-ions of type XII collagen in the remodeling periodontalligament during experimental tooth movement. J DentRes 1995: 74: 313–318.

117. Kikuchi A, Okuhara M, Karikusa F, Sakurai Y, Okano T.Two-dimensional manipulation of confluently culturedvascular endothelial cells using temperature-responsivepoly(N-isopropylacrylamide)-grafted surfaces. J BiomaterSci Polym Ed 1998: 9: 1331–1348.

118. Kinoshita A, Oda S, Takahashi K, Yokota S, Ishikawa I.Periodontal regeneration by application of recombinanthuman bone morphogenetic protein-2 to horizontal cir-cumferential defects created by experimental peri-odontitis in beagle dogs. J Periodontol 1997: 68: 103–109.

119. Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, De-guchi K, Shimizu Y, Bronson RT, Gao YH, Inada M, Sato M,

94

Okamoto R, Kitamura Y, Yoshiki S, Kishimoto T. Targeteddisruption of Cbfa1 results in a complete lack of boneformation owing to maturational arrest of osteoblasts.Cell 1997: 89: 755–764.

120. Kong YY, Yoshida H, Sarosi I, Tan HL, Timms E, CapparelliC, Morony S, Oliveira-dos-Santos AJ, Van G, Itie A, KhooW, Wakeham A, Dunstan CR, Lacey DL, Mak TW, BoyleWJ, Penninger JM. OPGL is a key regulator of osteo-clastogenesis, lymphocyte development and lymph-nodeorganogenesis. Nature 1999: 397: 315–323.

121. Krebsbach PH, Lee SK, Matsuki Y, Kozak CA, Yamada KM,Yamada Y. Full-length sequence, localization, and chro-mosomal mapping of ameloblastin – a novel tooth-speci-fic gene. J Biol Chem 1996: 271: 4431–4435.

122. Kuboki Y, Sasaki M, Saito A, Takita H, Kato H. Regenera-tion of periodontal ligament and cementum by BMP-ap-plied tissue engineering. Eur J Oral Sci 1998: 106(suppl 1):197–203.

123. Lafrenie RM, Yamada KM. Integrin-dependent signaltransduction. J Cell Biochem 1996: 61: 543–553.

124. Lanske B, Divieti P, Kovacs CS, Pirro A, Landis WJ, KraneSM, Bringhurst FR, Kronenberg HM. The parathyroid hor-mone (PTH)/PTH-related peptide receptor mediates ac-tions of both ligands in murine bone. Endocrinology1998: 139: 5194–5204.

125. Lekic P, Sodek J, McCulloch CAG. Osteopontin and bonesialoprotein expression in regenerating rat periodontalligament and alveolar bone. Anat Rec 1996: 244: 50–58.

126. Li H, Bartold PM, Zhang CZ, Clarkson RW, Young WG,Waters MJ. Growth hormone and insulin-like growth fac-tor I induce bone morphogenetic proteins 2 and 4: a me-diator role in bone and tooth formation? Endocrinology1998: 139: 3855–3862.

127. Lieberman JR, Le LQ, Wu L, Finerman GA, Berk A, WitteON, Stevenson S. Regional gene therapy with a BMP-2-producing murine stromal cell line induces heterotopicand orthotopic bone formation in rodents. J Orthop Res1998: 16: 330–339.

128. Lin WL, McCulloch CAG, Cho MI. Differentiation of peri-odontal ligament fibroblasts into osteoblasts duringsocket healing after tooth extraction in the rat. Anat Rec1994: 240: 492–506.

129. Lindhe J, ed. EmdogainA: a biological approach to peri-odontal regeneration. J Clin Periodontol 1997: 24: 675–714.

130. Liu YK, Uemura T, Nemoto A, Yabe T, Fujii N, Ushida T,Tateishi T. Osteopontin involvement in integrin-mediatedcell signaling and regulation of expression of alkalinephosphatase during early differentiation of UMR cells.FEBS Lett 1997: 420: 112–116.

131. Lochmuller H, Jani A, Huard J, Prescott S, Simoneau M,Massie B, Karpati G, Acsadi G. Emergence of early region1-containing replication-competent adenovirus in stocksof replication-defective adenovirus recombinants (deltaE1πdelta E3) during multiple passages in 293 cells. HumGene Ther 1994: 5: 1485–1491.

132. Loomer PM, Ellen RP, Tenenbaum HC. Effects of Porphyr-omonas gingivalis 2561 extracts on osteogenic and osteo-clastic cell function in co-culture. J Periodontol 1998: 69:1263–1270.

133. Lukinmaa PL, Mackie EJ, Thesleff I. Immunohistochem-ical localization of the matrix glycoproteins – tenascinand the ED-sequence-containing form of cellular

Molecular and cell biology of cementum

fibronectin – in human permanent teeth and periodontalligament. J Dent Res 1991: 70: 19–26.

134. Lynch SE, Williams RC, Polson AM, Howell TH, Reddy MS,Zappa UE, Antoniades HN. A combination of platelet-de-rived and insulin-like growth factors enhances peri-odontal regeneration. J Clin Periodontol 1989: 16: 545–548.

135. Maas R, Bei M. The genetic control of early tooth develop-ment. Crit Rev Oral Biol Med 1997: 8: 4–39.

136. MacNeil RL, Berry J, D’Errico J, Strayhorn C, Piotrowski B,Somerman MJ. Role of two mineral-associated adhesionmolecules, osteopontin and bone sialoprotein, during ce-mentogenesis. Connect Tissue Res 1995: 33: 1–7.

137. MacNeil RL, Berry J, D’Errico J, Strayhorn C, SomermanMJ. Localization and expression of osteopontin in min-eralized and nonmineralized tissues of the periodontium.Ann N Y Acad Sci 1995: 760: 166–176.

138. MacNeil RL, Berry JE, D’ Errico JA, Strayhorn CL, Piotrow-ski B, Somerman MJ. Role of two mineral-associated ad-hesion molecules, osteopontin and bone sialoprotein,during cementogenesis. Connect Tissue Res 1995: 33: 1–7.

139. MacNeil RL, Berry JE, Strayhorn CL, Shigeyama Y, Somer-man MJ. Expression of type I and XII collagen during de-velopment of the periodontal ligament in the mouse.Arch Oral Biol 1998: 43: 779–787.

140. MacNeil RL, Berry JE, Strayhorn CL, Somerman MJ. Ex-pression of bone sialoprotein mRNA by cells lining themouse tooth root during cementogenesis. Arch Oral Biol1996: 41: 827–835.

141. MacNeil RL, D’Errico JA, Ouyang H, Berry J, Strayhorn C,Somerman MJ. Isolation of murine cementoblasts:unique cells or uniquely-positioned osteoblasts? Eur JOral Sci 1998: 106(suppl 1): 350–356.

142. MacNeil RL, Sheng N, Strayhorn CL, Fisher LW, Somer-man MJ. Bone sialoprotein is localized to the root surfaceduring cementogenesis. J Bone Miner Res 1994: 9: 1597–1606.

143. MacNeil RL, Somerman MJ. Development and regenera-tion of the periodontium: parallels and contrasts. Peri-odontol 2000 1999: 19: 8–20.

144. MacNeil RL, Strayhorn CL, Berry JE, Bautista DS, Harris J,Chambers A, Somerman MJ. Osteopontin is localized andexpressed at specific sites during tooth eruption. In: Davi-dovitch Z, Norton LA, ed. The biological mechanisms oftooth movement and craniofacial adaptation. Boston:Harvard Society for the Advancement of Orthodontics,1996: 123–131.

145. MacNeil RL, Thomas HF. Development of the murineperiodontium. I. Role of basement membrane in forma-tion of a mineralized tissue on the developing root dentinsurface. J Periodontol 1993: 64: 95–102.

146. MacNeil RL, Thomas HF. Development of the murineperiodontium. II. Role of the epithelial root sheath in for-mation of the periodontal attachment. J Periodontol 1993:64: 285–291.

147. Maeda H, Kukita T, Akamine A, Kukita A, Iijima T. Localiz-ation of osteopontin in resorption lacunae formed by os-teoclast-like cells: a study by a novel monoclonal anti-body which recognizes rat osteopontin. Histochemistry1994: 102: 247–254.

148. Malarkey K, Belham CM, Paul A, Graham A, McLees A,Scott PH, Plevin R. The regulation of tyrosine kinase sig-

95

nalling pathways by growth factor and G-protein-coupledreceptors. Biochem J 1995: 309: 361–375.

149. McAllister B, Narayanan AS, Miki Y, Page RC. Isolation ofa fibroblast attachment protein from cementum. J Peri-odontal Res 1990: 25: 99–105.

150. McCauley LK, Somerman MJ. Biologic modifiers in peri-odontal regeneration. Dent Clin North Am 1998: 42: 361–387.

151. McCulloch CAG, Melcher AH. Cell migration in the peri-odontal ligament of mice. J Periodontal Res 1983: 18: 339–352.

152. McKee MD, Nanci A. Osteopontin and the bone remodel-ing sequence. Colloidal-gold immunocytochemistry of aninterfacial extracellular matrix protein. Ann N Y Acad Sci1995: 760: 177–189.

153. McKee MD, Nanci A. Postembedding colloidal-gold im-munocytochemistry of noncollagenous extracellular ma-trix proteins in mineralized tissues. Microsc Res Tech1995: 31: 44–62.

154. McKee MD, Nanci A. Osteopontin at mineralized tissueinterfaces in bone, teeth, and osseointegrated implants:ultrastructural distribution and implications for mineral-ized tissue formation, turnover, and repair. Microsc ResTech 1996: 33: 141–164.

155. McKee MD, Zalzal S, Nanci A. Extracellular matrix in toothcementum and mantle dentin: localization of osteopontinand other noncollagenous proteins, plasma proteins, andglycoconjugates by electron microscopy. Anat Rec 1996:245: 293–312.

156. Melcher AH. Repair of wounds in the periodontium of therat. Influence of periodontal ligament on osteogenesis.Arch Oral Biol 1970: 15: 1183–1204.

157. Mellonig JT. Bone allografts in periodontal therapy. ClinOrthop 1996: 324: 116–125.

158. Miettinen PJ, Berger JE, Meneses J, Phung Y, Pedersen RA,Werb Z, Derynck R. Epithelial immaturity and multiorganfailure in mice lacking epidermal growth factor receptor.Nature 1995: 376: 337–341.

159. Miki Y, Narayanan AS, Page RC. Mitogenic activity of ce-mentum components to gingival fibroblasts. J Dent Res1987: 66: 1399–1403.

160. Miller AD. Retrovirus packaging cells. Hum Gene Ther1990: 1: 5–14.

161. Morrison DK, Kaplan DR, Rapp U, Roberts TM. Signaltransduction from membrane to cytoplasm: growth fac-tors and membrane-bound oncogene products increaseRaf-1 phosphorylation and associated protein kinase ac-tivity. Proc Natl Acad Sci U S A 1988: 85: 8855–8859.

162. Mulligan RC. The basic science of gene therapy. Science1993: 260: 926–932.

163. Myint YY, Miyakawa K, Naito M, Shultz LD, Oike Y, Yama-mura K, Takahashi K. Granulocyte/macrophage colony-stimulating factor and interleukin-3 correct osteopetrosisin mice with osteopetrosis mutation. Am J Pathol 1999:154: 553–566.

164. Nakae H, Narayanan AS, Raines E, Page RC. Isolation andpartial characterization of mitogenic factors from ce-mentum. Biochemistry 1991: 30: 7047–7052.

165. Narayanan AS, Bartold PM. Biochemistry of periodontalconnective tissues and their regeneration: a current per-spective. Connect Tissue Res 1996: 34: 191–201.

166. Noble M, Groves AK, Ataliotis P, Ikram Z, Jat PS. The H-2KbtsA58 transgenic mouse: a new tool for the rapid gen-

Saygin et al.

eration of novel cell lines. Transgenic Res 1995: 4: 215–225.

167. Nohutcu RM, McCauley LK, Shigeyama Y, Somerman MJ.Expression of mineral-associated proteins by periodontalligament cells: in vitro vs. ex vivo. J Periodontal Res 1996:31: 369–372.

168. Oberpenning F, Meng J, Yoo JJ, Atala A. De novo recon-stitution of a functional mammalian urinary bladder bytissue engineering. Nat Biotechnol 1999: 17: 149–155.

169. Ogiso B, Hughes FJ, Melcher AH, McCulloch CA. Fibro-blasts inhibit mineralised bone nodule formation by ratbone marrow stromal cells in vitro. J Cell Physiol 1991:146: 442–450.

170. Oldberg A, Franzen A, Heinegard D. The primary structureof a cell-binding bone sialoprotein. J Biol Chem 1988: 263:19430–19432.

171. Olson S, Arzate H, Narayanan AS, Page RC. Cell attach-ment activity of cementum proteins and mechanism ofendotoxin inhibition. J Dent Res 1991: 70: 1272–1277.

172. Osborn JW, Price DG. An autoradiographic study of peri-odontal development in the mouse. J Dent Res 1988: 67:455–461.

173. Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC,Rosewell IR, Stamp GW, Beddington RS, Mundlos S, OlsenBR, Selby PB, Owen MJ. Cbfa1, a candidate gene for clei-docranial dysplasia syndrome, is essential for osteoblastdifferentiation and bone development. Cell 1997: 89: 765–771.

174. Ouyang H, McCauley LK, Berry JE, D’Errico JA, StrayhornCL, Somerman MJ. Response of immortalized murine ce-mentoblasts/periodontal ligament cells to parathyroidhormone and parathyroid hormone-related protein in vi-tro. Arch Oral Biol 2000: 45: 293–303.

175. Page RC, Baab DA. A new look at the etiology and patho-genesis of early-onset periodontitis. Cementopathia re-visited. J Periodontol 1985: 56: 748–751.

176. Palmer RM, Lumsden AG. Development of periodontalligament and alveolar bone in homografted recombi-nations of enamel organs and papillary, pulpal and fol-licular mesenchyme in the mouse. Arch Oral Biol 1987:32: 281–289.

177. Park JB, Matsuura M, Han KY, Norderyd O, Lin WL, GencoRJ, Cho MI. Periodontal regeneration in class III furcationdefects of beagle dogs using guided tissue regenerativetherapy with platelet-derived growth factor. J Periodontol1995: 66: 462–477.

178. Partanen AM. Epidermal growth factor and transforminggrowth factor-alpha in the development of epithelial-mesenchymal organs of the mouse. Curr Top Dev Biol1990: 24: 31–55.

179. Partanen AM, Alaluusua S, Miettinen PJ, Thesleff I, Tuom-isto J, Pohjanvirta R, Lukinmaa PL. Epidermal growth fac-tor receptor as a mediator of developmental toxicity ofdioxin in mouse embryonic teeth. Lab Invest 1998: 78:1473–1481.

180. Philbrick WM, Dreyer BE, Nakchbandi IA, Karaplis AC.Parathyroid hormone-related protein is required for tootheruption. Proc Natl Acad Sci U S A 1998: 95: 11846–11851.

181. Reddi AH. Initiation of fracture repair by bone morpho-genetic proteins. Clin Orthop 1998: 355(suppl): S66–S72.

182. Reinholt FP, Hultenby K, Oldberg A, Heinegard D. Osteo-pontin – a possible anchor of osteoclasts to bone. ProcNatl Acad Sci U S A 1990: 87: 4473–4475.

96

183. Reynolds MA, Bowers GM. Fate of demineralized freeze-dried bone allografts in human intrabony defects. J Peri-odontol 1996: 67: 150–157.

184. Ripamonti U. Induction of cementogenesis and peri-odontal ligament regeneration by bone morphogeneticproteins. In: Lindholm TS, ed. Bone morphogenetic pro-teins: biology, biochemistry and reconstructive surgery.Austin, TX: R. G. Landes Company, 1996: 189–198.

185. Ripamonti U, Heliotis M, Rueger DC, Sampath TK. Induc-tion of cementogenesis by recombinant human osteo-genic protein-1 (hOP-1/BMP-7) in the baboon (Papio ur-sinus). Arch Oral Biol 1996: 41: 121–126.

186. Ripamonti U, Heliotis M, van den Heever B, Reddi AH.Bone morphogenetic proteins induce periodontal re-generation in the baboon (Papio ursinus). J PeriodontalRes 1994: 29: 439–445.

187. Ritchie HH, Berry JE, Somerman MJ, Hanks CT, BronckersAL, Hotton D, Papagerakis P, Berdal A, Butler WT. Dentinsialoprotein (DSP) transcripts: developmentally-sustainedexpression in odontoblasts and transient expression inpre-ameloblasts. Eur J Oral Sci 1997: 105: 405–413.

188. Roberts WE, Wood HB, Chambers DW, Burk DT. Vascu-larly oriented differentiation gradient of osteoblast pre-cursor cells in rat periodontal ligament: implications forosteoblast histogenesis and periodontal bone loss. J Peri-odontal Res 1987: 22: 461–467.

189. Ruoslahti E, Obrink B. Common principles in cell ad-hesion. Exp Cell Res 1996: 227: 1–11.

190. Rutherford RB, Niekrash CE, Kennedy JE, Charette MF.Platelet-derived and insulin-like growth factors stimulateregeneration of periodontal attachment in monkeys. JPeriodontal Res 1992: 27: 285–290.

191. Rutherford RB, Sampath TK, Rueger DC, Taylor TD. Useof bovine osteogenic protein to promote rapid osseoin-tegration of endosseous dental implants. Int J Oral Maxil-lofac Implants 1992: 7: 297–301.

192. Sahlberg C, Hormia M, Airenne T, Thesleff I. Laminingamma2 expression is developmentally regulated duringmurine tooth morphogenesis and is intense in amelo-blasts. J Dent Res 1998: 77: 1589–1596.

193. Saito M, Narayana AS. Signaling reactions induced inhuman fibroblasts during adhesion to cementum-de-rived attachment protein. J Bone Miner Res 1999: 14:65–72.

194. Saito S, Rosol TJ, Saito M, Ngan PW, Shanfeld J, Davidov-itch Z. Bone-resorbing activity and prostaglandin E pro-duced by human periodontal ligament cells in vitro. JBone Miner Res 1990: 5: 1013–1018.

195. Sakou T. Bone morphogenetic proteins: from basicstudies to clinical approaches. Bone 1998: 22: 591–603.

196. Salmivirta K, Gullberg D, Hirsch E, Altruda F, Ekblom P.Integrin subunit expression associated with epithelial-mesenchymal interactions during murine tooth develop-ment. Dev Dyn 1996: 205: 104–113.

197. Sandy J, Davies M, Prime S, Farndale R. Signal pathwaysthat transduce growth factor-stimulated mitogenesis inbone cells. Bone 1998: 23: 17–26.

198. Sastry SK, Horwitz AF. Adhesion-growth factor interac-tions during differentiation: an integrated biological re-sponse. Dev Biol 1996: 180: 455–467.

199. Schroeder HE. The periodontium. In: Oksche A, BollrathL, ed. Handbook of microscopic anatomy. Berlin:Springer, 1986: 23–129.

Molecular and cell biology of cementum

200. Selvig KA, Wikesjö UM, Bogle GC, Finkelman RD. Im-paired early bone formation in periodontal fenestrationdefects in dogs following application of insulin-likegrowth factor. II. Basic fibroblast growth factor and trans-forming growth factor beta 1. J Clin Periodontol 1994: 21:380–385.

201. Sharpe PT. Homeobox genes and orofacial development.Connect Tissue Res 1995: 32: 17–25.

202. Shigeyama Y, D’Errico JA, Stone R, Somerman MJ. Com-mercially-prepared allograft material has biological activ-ity in vitro. J Periodontol 1995: 66: 478–487.

203. Shigeyama Y, Grove TK, Strayhorn C, Somerman MJ. Ex-pression of adhesion molecules during tooth resorptionin feline teeth: a model system for aggressive osteoclasticactivity. J Dent Res 1996: 75: 1650–1657.

204. Shiraga H, Min W, VanDusen WJ, Clayman MD, Miner D,Terrell CH, Sherbotie JR, Foreman JW, Przysiecki C, Neil-son EG, Hoyer JR. Inhibition of calcium oxalate crystalgrowth in vitro by uropontin: another member of the as-partic acid-rich protein superfamily. Proc Natl Acad SciU S A 1992: 89: 426–430.

205. Shuldiner AR. Transgenic animals. N Engl J Med 1996:334: 653–655.

206. Sigurdsson TJ, Fu E, Tatakis DN, Rohrer MD, Wikesjö UM.Bone morphogenetic protein-2 for peri-implant bone re-generation and osseointegration. Clin Oral Implants Res1997: 8: 367–374.

207. Sigurdsson TJ, Lee MB, Kubota K, Turek TJ, Wozney JM,Wikesjö UME. Periodontal repair in dogs: recombinanthuman bone morphogenetic protein-2 significantly en-hances periodontal regeneration. J Periodontol 1995: 66:131–138.

208. Sigurdsson TJ, Nygaard L, Tatakis DN, Fu E, Turek TJ, JinL, Wozney JM, Wikesjö UM. Periodontal repair in dogs:evaluation of rhBMP-2 carriers. Int J Periodontics Restora-tive Dent 1996: 16: 524–537.

209. Singh RP, Patarca R, Schwartz J, Singh P, Cantor H. Defi-nition of a specific interaction between the early Tlymphocyte activation 1 (Eta-1) protein and murinemacrophages in vitro and its effect upon macrophages invivo. J Exp Med 1990: 171: 1931–1942.

210. Slavkin HC, Bessem C, Fincham AG, Bringas P Jr, SantosV, Snead ML, Zeichner-David M. Human and mouse ce-mentum proteins immunologically related to enamel pro-teins. Biochim Biophys Acta 1989: 991: 12–18.

211. Slavkin HC, Bringas P Jr, Bessem C, Santos V, NakamuraM, Hsu MY, Snead ML, Zeichner-David M, Fincham AG.Hertwig’s epithelial root sheath differentiation and initialcementum and bone formation during long-term organculture of mouse mandibular first molars usingserumless, chemically-defined medium. J Periodontal Res1989: 24: 28–40.

212. Somerman MJ, Berry JE, Ouyang H, Strayhorn CL,McCauley LK, Saygin NE, Barald K, D’Errico JA. Definingthe cementoblast: from in situ to in vitro. J Am Acad Or-thop Surg (in press).

213. Somerman MJ, Fisher LW, Foster RA, Sauk JJ. Humanbone sialoprotein I and II enhance fibroblast attachmentin vitro. Calcif Tissue Int 1988: 43: 50–53.

214. Somerman MJ, Morrison GM, Alexander MB, Foster RA.Structure and composition of cementum. In: Bowan W,Tabak L, ed. Cariology of the 1990’s. New York: Universityof Rochester Press, 1993: 155–171.

97

215. Somerman MJ, Sauk JJ, Foster RA, Norris K, Dickerson K,Argraves WS. Cell attachment activity of cementum: bonesialoprotein II identified in cementum. J Periodontal Res1991: 26: 10–16.

216. Somerman MJ, Shroff B, Agraves WS, Morrison G, CraigAM, Denhardt DT, Foster RA, Sauk JJ. Expression ofattachment proteins during cementogenesis. J Biol Buc-cale 1990: 18: 207–214.

217. Somerman MJ, Shroff B, Foster RA, Butler WT, Sauk JJ.Mineral-associated adhesion proteins are linked to rootformation. Proc Finn Dent Soc 1992: 88(suppl 1): 451–461.

218. Sommer B, Bickel M, Hofstetter W, Wetterwald A. Express-ion of matrix proteins during the development of min-eralized tissues. Bone 1996: 19: 371–380.

219. Sorensen ES, Hojrup P, Petersen TE. Posttranslationalmodifications of bovine osteopontin: identification oftwenty-eight phosphorylation and three O-glycosylationsites. Protein Sci 1995: 4: 2040–2049.

220. Soriano P, Montgomery C, Geske R, Bradley A. Targeteddisruption of the c-src proto-oncogene leads to osteope-trosis in mice. Cell 1991: 64: 693–702.

221. Stubbs JT 3rd, Mintz KP, Eanes ED, Torchia DA, Fisher LW.Characterization of native and recombinant bone sialop-rotein: delineation of the mineral-binding and cell ad-hesion domains and structural analysis of the RGD do-main. J Bone Miner Res 1997: 12: 1210–1222.

222. Sundquist KT, Jackson ME, Hermey DC, Marks SC Jr. Os-teoblasts from the toothless (osteopetrotic) mutation inthe rat are unable to direct bone resorption by normalosteoclasts in response to 1,25-dihydroxyvitamin D.Tissue Cell 1995: 27: 569–574.

223. Takahashi N, Udagawa N, Suda T. A new member of tu-mor necrosis factor ligand family, ODF/OPGL/TRANCE/RANKL, regulates osteoclast differentiation and function.Biochem Biophys Res Commun 1999: 256: 449–455.

224. Takano-Yamamoto T, Takemura T, Kitamura Y, Nomura S.Site-specific expression of mRNAs for osteonectin, osteo-calcin, and osteopontin revealed by in situ hybridizationin rat periodontal ligament during physiological toothmovement. J Histochem Cytochem 1994: 42: 885–896.

225. Takasu H, Guo J, Bringhurst FR. Dual signaling and ligandselectivity of the human PTH/PTHrP receptor. J BoneMiner Res 1999: 14: 11–20.

226. Ten Cate AR, Mills C, Solomon G. The development of theperiodontium. A transplantation and autoradiographicstudy. Anat Rec 1971: 170: 365–379.

227. Tenorio D, Cruchley A, Hughes FJ. Immunocytochemicalinvestigation of the rat cementoblast phenotype. J Peri-odontal Res 1993: 28: 411–419.

228. Tenorio D, Hughes FJ. An immunohistochemical investiga-tion of the expression of parathyroid hormone receptors inrat cementoblasts. Arch Oral Biol 1996: 41: 299–305.

229. Thesleff I, Nieminen P. Tooth morphogenesis and cell dif-ferentiation. Curr Opin Cell Biol 1996: 8: 844–850.

230. Thesleff I, Partanen AM, Vainio S. Epithelial-mesenchy-mal interactions in tooth morphogenesis: the roles ofextracellular matrix, growth factors, and cell surface re-ceptors. J Craniofac Genet Dev Biol 1991: 11: 229–237.

231. Thesleff I, Sahlberg C. Growth factors as inductive signalsregulating tooth morphogenesis. Annu Rev Cell Dev Biol1996: 7: 185–193.

232. Thesleff I, Sharpe P. Signalling networks regulating dentaldevelopment. Mech Dev 1997: 67: 111–123.

Saygin et al.

233. Thesleff I, Vaahtokari A, Vainio S, Jowett A. Molecularmechanisms of cell and tissue interactions during earlytooth development. Anat Rec 1996: 245: 151–161.

234. Thomas HF, Kollar EJ. Tissue interactions in normal mu-rine root development. In: Davidovitch Z, ed. The biologi-cal mechanisms of tooth eruption and root resorption.Birmingham, UK: EBSCO Media, 1988: 145–151.

235. Thomas HF, MacNeil RL, Haydenbilt R. The role of epi-thelium in the developing and adult murine periodon-tium. In: Davidovitch Z, ed. The biological mechanisms oftooth eruption, resorption and replacement by implants.Boston: Harvard Society for the Advancement of Ortho-dontics, 1994: 317–323.

236. Tucker AS, Matthews KL, Sharpe PT. Transformation oftooth type induced by inhibition of BMP signaling.Science 1998: 282: 1136–1138.

237. Tureckova J, Sahlberg C, Aberg T, Ruch JV, Thesleff I,Peterkova R. Comparison of expression of the msx-1, msx-2, BMP-2 and BMP-4 genes in the mouse upper diastemaland molar tooth primordia. Int J Dev Biol 1995: 39: 459–468.

238. Urias SE, Gluhak J, Ramirez L, Goldman E, Pavlin D. Effectof mechanical stress on bone markers in cellular ce-mentum. J Dent Res 1999: 78S: 500.

239. Vainio S, Karavanova I, Jowett A, Thesleff I. Identificationof BMP-4 as a signal mediating secondary induction be-tween epithelial and mesenchymal tissues during earlytooth development. Cell 1993: 75: 45–58.

240. Vainio S, Thesleff I. Sequential induction of syndecan,tenascin and cell proliferation associated with mesenchy-mal cell condensation during early tooth development.Differentiation 1992: 50: 97–105.

241. van Dijk S, D’Errico JA, Somerman MJ, Farach-CarsonMC, Butler WT. Evidence that a non-RGD domain in ratosteopontin is involved in cell attachment. J Bone MinerRes 1993: 8: 1499–1506.

242. Wang HL, MacNeil RL. Guided tissue regeneration. Ab-sorbable barriers. Dent Clin North Am 1998: 42: 505–522.

243. Wang ZQ, Ovitt C, Grigoriadis AE, Mohle-Steinlein U,Ruther U, Wagner EF. Bone and haematopoietic defectsin mice lacking c-fos. Nature 1992: 360: 741–745.

244. Wikesjö UM, Razi SS, Sigurdsson TJ, Tatakis DN, Lee MB,Ongpipattanakul B, Nguyen T, Hardwick R. Periodontalrepair in dogs: effect of recombinant human transforminggrowth factor-beta1 on guided tissue regeneration. J ClinPeriodontol 1998: 25: 475–481.

98

245. Wikesjö UM, Selvig KA, ed. Periodontal wound healingand regeneration. Periodontol 2000 1999: 19: 1–172.

246. Wise GE. The biology of tooth eruption. J Dent Res 1998:77: 1576–1579.

247. Wise GE, Fan W. Immunolocalization of transforminggrowth factor beta in rat molars. J Oral Pathol Med 1991:20: 74–80.

248. Wise GE, Lin F. The molecular biology of initiation oftooth eruption. J Dent Res 1995: 74: 303–306.

249. Wu D, Ikezawa K, Parker T, Saito M, Narayanan AS. Char-acterization of a collagenous cementum-derived attach-ment protein. J Bone Miner Res 1996: 11: 686–692.

250. Xuan JW, Hota C, Chambers AF. Recombinant GST-humanosteopontin fusion protein is functional in RGD-depend-ent cell adhesion. J Cell Biochem 1994: 54: 247–255.

251. Xuan JW, Hota C, Shigeyama Y, D’Errico JA, SomermanMJ, Chambers AF. Site-directed mutagenesis of the argi-nine-glycine-aspartic acid sequence in osteopontin de-stroys cell adhesion and migration functions. J Cell Bio-chem 1995: 57: 680–690.

252. Yamada S, Yamada KM, Brown KE. Integrin regulatoryswitching in development: oscillation of beta 5 integrinmRNA expression during epithelial-mesenchymal inter-actions in tooth development. Int J Dev Biol 1994: 38:553–556.

253. Yamamoto T, Domon T, Takahashi S, Wakita M. Compara-tive study of the initial genesis of acellular and cellularcementum in rat molars. Anat Embryol (Berl) 1994: 190:521–527.

254. Yonemura K, Narayanan AS, Miki Y, Page RC, Okada H.Isolation and partial characterization of a growth factorfrom human cementum. Bone Miner 1992: 18: 187–198.

255. Yonemura K, Raines EW, Ahn NG, Narayanan AS. Mito-genic signaling mechanisms of human cementum-de-rived growth factors. J Biol Chem 1993: 268: 26120–26126.

256. Young WG. Growth hormone and insulin-like growth fac-tor-I in odontogenesis. Int J Dev Biol 1995: 39: 263–272.

257. Zhang CZ, Young WG, Li H, Clayden AM, Garcia-AragonJ, Waters MJ. Expression of growth hormone receptor byimmunocytochemistry in rat molar root formation andalveolar bone remodeling. Calcif Tissue Int 1992: 50: 541–546.

258. Zhang R, Ducy P, Karsenty G. 1,25-dihydroxyvitamin D3

inhibits osteocalcin expression in mouse through an in-direct mechanism. J Biol Chem 1997: 272: 110–116.