Dy sregulation of Indian hedgehog - Parathyroid hormone related protein signalling in cartilaginous neoplasia

Sevan Hopyan

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy, Graduate Department of the

Institute of Medical Science, University of Toronto

O Copyright by Sevan Hopyan 2001

National Library u d C m & Biblioîhèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services services bibliographiques

395 Wellmgtm Street 395, rue Wellingtwi 0na~aON K1AON4 Mtavr;tW K 1 A W CaMda CaMda

The author has granted a non- L'auteur a accordé une Licence non exclusive Licence aliowing the exclusive permettant a la National Library of Canada to Bibliothèque nationde du Canada de reproduce, loan, distriiute or sel1 reproduire, prêter, distribuer ou copies of this thesis in microfom, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/^ de

reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts kom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission autorisation.

Dysregulation of Indian hedgehog - Parathyroid hormone related protein signatling

ris a mechanism of cartilaginous neoplasia

Degree of Doctor of Philosophy, 2001

Sevan Hopyan

Institute of Medical Science

University of Toronto

ABSTRACT

Enchondroma is a benign cartilage - forming tumour of bone that might be çaused

by dysregulation of growth plate signais. A parxnne feedback loop couples Indian

hedgehog (Mi). which stimulates growth plate chondrocyte prolifention. to Parathyroid

hormone related protein (PTHrP). which regulates chondrocyte differentiation. Here 1

show thnt these two signallin_o pathways were uncoupled and IHH was dysinhibiced in

explant cultures of human enchondromas and their malignant counterpms,

chondrosarcomns. A heterozygous variant ripe I PTHffTHrP receptor (PTHRI) was

discovered in the germline of one patient and as a somatic change in another with

enchondromatosis. In this condition, multiple enchondromas mny lead to deformity and

chondrosarcorna- The variant PTHRI suppressed Cyclic ndenosine monophosphate

baseline level in a dominant fashion and abolished Inositol triphosphate accumuiation in

vitro. Transgenic mice expressing the variant receptor under the cegulatory elements of

Tvpe II coilagen developed enchondrornatosis.

1 show that the transcription factor Gli2 was a positive regulator of ihh function in

the mutine ,mwth plate, while Gli3 was a negative regulator- Overexpression of Gli2

was found in human enchondromas, and was sufficient to cause enchondromatosis in

transgenic mice. A lack of Gli3 accelerated another condition of benign chondrocyte

neoplasia. synovial chondromatosis.

Patched I (PTCHL) . a Hedghehog receptor. formed a cornplex with PTHRL. and

this complex was required for effective accumuIation of PTHR 1 second rnessengers. The

variant ETHRI did not associate with PTCHl. and constitutively activated Hedgehog

sigalling in a mrinner that was likely dependent upon suppression of Protein kinase A.

Dysregulation of growth plate signais causes certain benign crinilage ttumours.

Agents thrit block Hedgehog signalling in particular. might be useful in preventing the

deleterious sequelrie of enchondromas.

My geatest thanks to God. I apologïse to al1 the people whose love and whose

needs 1 have neglected during my selfish drive to complete this thesis. Chief among

these are my sister Talar. my rnother Lucya and my father Takvor.

1 am indebted to rny supervisors. mentors and friends, Ben Alman and Jay

Wunder. Their enthusiasm, engagement and support in ;ood tirnes and bad were

unparalleled and moving I am wmteful to irene Andrulis and Bob Bell for their

unconditional support over many years, My heartfelt thanks IO Jme Aubin and Serin

Egan. the other members of my thesis advisory cornmittee, for scholarly exchange of

such scat value that it has changed me profoundl y and, dare 1 Say. irrevocabl y. Many

thanks to Chi-Chung Hui for GCiZ consuucts and G1iZ and Gli3 deficient mice as well as

for supportive discussions.

The number of non-principal investigators from whom I have leamed is great.

These people patiently taught me how to carry out many of my experiments. and helped

me, inevitably. to trouble-shoot Thanks especially to Nalan Gokgoz (for SSCP and

sequencing in particular), Minh To (for help with CAMP assays in particular). Kolja

Eppert, Monica Sauer (for cloning and protein work advice in particular), Raymond Poon

(for SSCP and sequencing). Chunying Yu (Cor fetai bone cultures. COLX

immunohistochernistry, Hedgehog sipdling assays and coimmunoprecipitation), Chris

Huggîns. Susan Done, Nona Ameson, Sherry Hendry, Lucy Collins. Tania Zadro. Teresa

Mastracci, Spyro Mousses, Catherine Li, Puvi Nadesan (for breeding ~ 1 i 2 ~ - and ~ 1 i 3 ~ -

mice), Alex Cheah, Sophia Cheung, Rong Mo. Mary Zhang, Vigitha Nadesan, Ti10

Kunath, Brian Cinina, Derinna Church, Ekaterina Hadjantonakis, Hassina Benchabane,

Pamela Hoodless, John Hudson, Linda Doughty, Gabriella Nagy, Lily Morikawa (for

histology), Michael Ho (for immunohistochemistry), Sandra Tondat and Lois Schwartz

(for pronuclear microinjection), MaIgosia Kowmacka (for ES cell media and feeder

cells), Aileen Davis (for statisticai help), Paul Reynolds. Mingyao Liu, Jim Dennis and

Janet Rossant for help, support and general love. Bless them all.

1 feel deep appreciation for the eager cooperation of patients and their families:

may they benefit a thousand times over in retum. For sharing precious reagents, 1 thank

U.4. Chun; (for PTHR lJ* ES cells), F. de Sauvage (Genentech - for SM0 cDNA), W.

Gaffield (for cyclopamine), C-C Hui (for Gli2 and PTCHI cDNAs and for Gli2 and

Gli3 detkient mice). T. Ingolia (Ontogeny - for Shh-N protein). H. Juppner and A.

Karaplis (for PTHRL cDNA), C, Tabin (for iHH and SHH cDNAs), and Y. Yamada

(for ColII promoter/enhancer cDNA). For human specirnens in addition to those

obtained from the Mt. Sinai Sarcoma Tissue Bank. 1 thank W. Cole. A. Gross. C.

Hutchison and J. Wright.

TABLE OF CONTENTS

ABSTRACT

ACKIOWLEDGEMENTS

LIST OF TABLES

LIST OF FIGURES

LIST OF ABBREVIATIONS

CHAPTER 1 Introduction

BACKGROUND

Development and neoplasia

Cartilage - forrning tumours

Bone development

Hedgehog signal trmsduction

PTHrP signal tnnsduction

Regulation of chondrocyte differentiation

HYPOTHES IS

RATION ALE

OBECTNES

FIGURES 1-4

CHAPTER 2 Uncoupled Indian hedgehog - Parathyroid hormone

related protein signaliing in cartilage tumours and a mutant

Type 1 FTiWTHrP receptor in enchondromatosis

SUlMiMhRY

INTRODUCTION

RESLJLTS

=e tumours Expression of MWPTHrP pathway members in catilac

Functional but uncoupled MWPTHrP signalling in

cartilage tumour explants

A mutant PTHRl in enchondromatosis

Mutant PTHRl signalling is impaired in vitro

Tnnsgenic mouse model of enchondromatosis

DISCUSSION

Enchondroma formation

Genetics of enchondromatosis

METHODS

TMLES 1-4

FIGURES 1-6

CHAPTER 3 Gli2 and GIi3 in growth plate regulation and pathology

SCTMtMARY

INTRODUCTION

RESLZTS

Growth plate proliferation is increased by GliZ and

diminished by G1i3

PTiirP stimulation decreases proliferation in addition to

delaying hypenrophic differentiation, and these responses are

diminished by GIi2 and au-mented by Gli3

vii

ColII-Gli2 mice develop enchondromatosis 59

Gli3 heterozygous mice develop accelerated synovial chondromatosis 60

DISCUSSION 6 1

Gli2 mediiites and GIi3 inhibits Ihh function in the growth plate 6 1

Evidence of functional antagonism between Gliî and PTHrP 6 1

Excessive Hedgehog signallins is sufficient to generate

cartilaginous neoplasia 63

METHODS 64

FIGURES 1-6 66

CHAPTER 1 Patched is a CO-receptor with Smoothened and PTHR1, two seven-

pass transmembrane proteins with opposite effects on CAMP accumulation 72

SUMiMARY 73

INTRODUCTION 74

RES üLTS 75

Coenprcssion of PTCHl with PTHRl augments CAMP accumularion

and limits Hedghog signal transduction 7 5

ETCHL and PTMRl fonn a complex 75

DISCUSSION 77

ETCH l-FiTRl association is required for effective accumulation

of c A i and iPj

Ce11 autonornous mode1 of Hedgehog - PTHrP interaction

m O D S

RGURES 1-2

CHAPTER 5 Conclusions and Future Directions

APPENDIX Expression of Osteocalcin and its transcriptional regdators

CBFAZ and MSX2 in osteoid forming tumours

S W Y

INTRODUCTION

0 tumours Osteoid - formin,

Osteoblast differentiation

RES üLTS

DISCUSSION

METHODS

TABLES 1-3

REFERENCES

LIST OF TABLES

C r n E R 2

1 Expression of IHH-PTHrP pathway members by RT-PCR

3 Expression level of GLII, GL12 and GU3 by semiquantitative RT-PCR

3 RT-PCR primer sequences

4 SSCP/sequencing primer sequences for PTHRl

APPENDIX

la RT-PCR primer sequences

Lb SSCP/sequencing primer sequences for CBFAl Runt domain

2 OC, CBFAI and MSX2 expression IeveIs by semiquantitative RT-PCR

3 Categorized of OC. CBFAI and MSX2 expression levels

LIST OF FiGüRES

1 Enchondroma location within a longbone

2 Growth plate chondrocyte differentiation

3 Hedgehog sipalling

4 PTHrP sigalling

CHAITER 2

1 MH-PTHrP signalling in human rumour explant cultures

2 In vitro characterisation of n variant (RISOC) PTHR 1 in enchondromütosis

3 Genotyping and transgene expression in ColII-RISUC PTHRI transgenic mice

4 Neonatal groowth plate phenotype of Cuiil-RISOC PTHRl mice

5 Enchondromas in CollI-RISUC PTHRI mice

6 Adolescent growth plate phentitype and TRAP staining of

CulII-Rl5UC PTHRl mice

CHAPTER 3

1 Growth plate phenotype of ColIl-Gli2 transgenic mice

2 Comparative differentiation and proliferation in explant culture of fetal limbs

from mice lackin_g GliZ or Gli3. CoIII-Gli2 and CulII-RISOC PTHRI mice

3 uCOLX - stained fetal limb expiants

3 &rdU - stained fetal limb explants

5 Enchondromas in ColIl-Gfi2 rnice

6 Synovial chondromatosis in ColIl-Cl2 mice

CHAPTER 4

1 PTHRl second messenger and Hedgehog sigalling assays with PTCHl/PTHRl

coexpression 80

2 Coimmunoprecipitation of PTCHl with PTHRl 82

3 Model of ceIl autonomous regulation of prolifention and differentiation by the

Hedgehog and PTHrP signallin; pathways 76

APPENDIX

RT-PCR of regulators of osteoblast differentiation in osteoid - forrning turnours 105

xii

LIST OF ABBREMATIONS

A - adenine

AML L - Acute myelogenous leukemia 1

M C - Adenomatous polyposis coli

AS - Asparagine s ynthetase

a - anti: alpha

Brnp - Bone morphogenic protein

$ Gal - Beta galactosidase

B2M - Beta-2-microglobulin

bp - base pairs

BrdU - bromodeoxyuridine

cm - centimetre

C -cytosine: cysteine (as in R ISOC)

CAMP - Cyclic ridenosine monophosphate

CBFAl - Core binding factor alpha 1

ci - cubitus intemiptus

ColiI - Type EI collagen

ColX - Type X colhgen

Cos2 - Costal 3.

CPM - counts per minute

CREB -CAMP response element binding protein

CS A - chondrosarcoma

Dhh - Desert hedgehog

DMEM - Dulbecco's modified eaglc medium

E - embryonic day

ECA - enchondroma

ES - embryonic stem

EW S - Ewing's sarcoma

EXT - Exostosis

M - fentomolar

FBS - fetal bovine serum

FD - fibrous dysplasia

Fgf - Fibroblast growth factor

Fu - Fused

g - p m

G - guanine

GDPIGTP - guanine diphosphate1 guanine triphosphate

GIi - Glioblastoma

GP - growth plate

'H - tritium

HA - hnemaglutinin

Hh - Hedgehog

Ihh - Indian hedgehoj

Ig - immonoglobuIin

Igf - Insulin-like growth factor

IP - immumoprecipitate

xiv

iPj - lnositol triphosphate

kD - kilo-Dalton

L - liter

M - mohr

pM - micromotar

mm - milimetre

mM - milimolar

MO - myositis ossificans

N - normal

-N - amino terminus

OB - osteoblast culture

OC - Osteocalcin

OSA - osteosarcorna

pM - picomolar

PBGD - PorphobiIinogen deaminase

P U - Protein kinase A

ptc - patched (Drosophiia)

PTCH 1 - Patched 1 (mammalian)

P?THrP - Pmthyroid hormone related protein

PTHRi -Type I PTWPTHrP receptor

R - arginine

RLU -relative luciferase units

RT-PCR - reverse transcription-polyrnense chain reaction

SDS - sodium dodecylsulphate

Shh - Sonic hedgehog

SM0 - Smoothened

SSCP - single strand conformrition polyrnorphism

Su(Fu) - Suppressor of fused

T - thymine

TRAP - tartrate resistant acid phosphaease

WT - wild type

Gene

snlo

Srno

S M 0

Protien

smo

Smo

SM0

xvi

Introduction

BACKGROUND

Development and neoplasia

Development is the most substantial bio1ogical undertaking of any organism. The

regulation of development is necessarily exquisite. If for no other reason. the shear

mapitude and complexity of the problem of developrnental regulation lends this

regulation susceptible to error. In pmicular. dysregulation of developmental mechanisms

might cause neoplasiri.

Neoplastic cells share many chancteristics with developing cells, These

characteristics include maintenance of an incompletely differentiated state. rapid

prolifention. mobility and invasive capability. A variety of tumours resemble the

developing stages of their tissue of origin. and this resemblance may be manifest at a

morpholo@cal, behaviourai or molecu1ar leveI. Understanding why neoplastic cells

resemble undifferentiated cells would facilitate thenpeutic ripproaches designed to

differentiate those cells. thereby causins them to Iose ckwacteristics which make them

deleterious to the host. without necessririly restoring al1 senetic controls'.

Evidence irnplicating the significance of developmentd regulators in neoplasia

has been accumulating over many years. Positive regulators of developmental sipalling

pathways may behave as oncogenes when they are ectopicalIy expressed or harbour

activating mutations. Exarnples of such positive regulators are WNT (breast cancet) and

@eutenin (agpssive fibromatosis3). AMLI (leukemid), PAX3 and PM7

(rhabdomyosarcoma5). EWS and Manic Fnnge (Ewing's sarcoma6:'), and Sonic

Hedgeliog (SHH), Smootlrened (SMO), GUI and Gli2 (basa1 ce11 carcinoma and

medul1oblristorn~-"). Nesacive developmentaf regulators may behave as tumour

suppressor genes when they harbour loss of function mutations. Such negative regulators

include M C (agressive fibromatosis'*, colon cancer15) and Parchedl (PTCHI: basal

cell carcinoma and med~lloblastoma'~"~ What has become increasingly clear is the

importance of the dysregulation of a given signalling pathway, rather than of a single

gene. For example. while deregulation of either pcatenin or M C independently may

>1&17 give rise to ag_gessive fibromatosis . both these genes encode members of the same

signalling pathway. Likewise, SHH, PTCH. SMO. and GLI encode members of another

pathway ['.

Of the three general types of cancer. carcinoma, hematopoietic neoplasia and

sarcoma. sarcoma is the least studied. Sarcomas and their benign precursors are tumours

of connective tissue presumably derived from some mesenchymal ce11 type. Skeletal

smomas rire most common in children and adolescents. whereas soft tissue sarcornrts are

more common in adults. Although less common chan ocher types of cancer in the senen1

population. sarcomas have an arguably -mater negative impact then most neoplastic

conditions. This impact is due to the large number of life years and reproductive

potentiril Iost due to mortdity at a young age as well as the often tremendous physical and

non-physical trauma caused by sarcomas and their treatment

Cartilage - forming tumours

There are a number of neoplasms with ceils that bear some resemblance to

chondrocytes and which produce a cartilriginous mauix. The various types of

canilaginous neoplasms are individually distinct based on histologie features, skeletd

location and their potenud for locdly destructive behaviour and for metastasis, usually to

the iung~'~". Cartilage tumours, with the exception of chondrosarcoma, occur most

comrnoniy in children and younj adults. The following tumours have recogisably

chondroid elements, but also have non-chondroid featutes to vxying degm: 1)

Chondroblristomü is a locally aggressive tumour occuring within the epiphysis of a bone.

These tumours c q a risk of metastasis of less thm five percent. 3) Chondromyxoid

fibroma is ri rehtively rare, benign metaphyseal tumour that is iocalIy destructive. 3)

Chordoma is felt to anse from remnant notochordal tissue, and occurs most commonly at

the base of the skull or in the sacrum. These tumours are iocally nggressive and carry a

less than fifteen percent rïsk of metastasis. 4) Osteochondroma. or exostosis, is a bony

outgrowth with P central medullary cavity which is continuous with chat of the host bone

and h a a well di fferentiated cartilage cap that is simiiar to growth plate tissue. These

lesions can cause mechrinical symptoms, and the cartilage cap rnay also undergo

rnüliyant ttrinsfomarion. most cornmonly to chondrosmoma and rarely to

osteosarcorna. Osteochondromas theinselves may be aneuploid2'. and the nsk of

malignanr change is less thrin one percent for a solitîry osreo~hondroma'~. but is hi$er

when muItiple osteochondromas are present. Hereditq mutiple exostoses is an

autosorna1 dominant condition with an incidence of about 1/50000. and. as are some

sporadic exostoses. is caused by loss of function mutations in one of the three EXTfamily

'1-3 genes- . EXTgenes are homologs of Drosophila r m r velu, which is necessary for long

range diffusion of hedgeehog liSmdx3 (see below).

Other benign tumours with erisily recognizable cartilagïnous differentiarion

inchde enchonciromas, which occur adjacent to the metaphyseal side of the gowth plate

(Fig. l), and penosteal chonciromas. which occur at the surface of a b~ne ' '~~ .

Enchondromas can occur as solitary lesions, or as multiple lesions in enchondromatosis

(OIlier's and Maffucci's diseases). An enchondroma rnay be completely asymptornauc

and be discovered as an incidentai finding on a radio_mph. CIinicai problems thar can be

caused by enchondromas include pain, skeIetd deformity, bony weakness leading to

'0:2017.18 pathological fmcture, and rnalignanr change to chondrosarcoma"'- ,The risk of

ckondrosarcoma arising from an enchondroma is about 25% in Ollier's disease and

vinually 100% in Maffucci's diseaseIg. Maffucci's disease is further distinct from

Ollier's disease because of additional phenotypic features such as soit tissue

hemangiomas and visceral and central nervous system tumours. The extent ofskeletal

involvement is variable in enchondrornatosis. and mriy include dysplasia that is not

directly attributable to enchondrorna~'~-". Enchondromatosis is rare. obvious inherïtance

of the condition is un~sual'~''' and no candidate genetic loci have k e n identified.

Enchondromas are composed of cells cytoIogically sirnilm to gowth plate chondrocyces.

possibly representing foci of incomplett endochondnl ossifi~ation'~'~. Enchondrornas

may disappear with tirne following skeleta1 maturity. possibly due to "gradua1 completion

of endochondnl ossification. Little is known about the molecular pathology of these

lesions. and there are no animal rnodels, Current treatments are exclusiveIy surgical. and

consist of tumour excision with bone defect reconstniction, fixation of pathological

fracture^'^ and osteotorny for limb realipnent3j.

Synovial chondromatosis is another variety of cartilaginous neoplasia. In this

condition. tissue resernbling *pwth plate cartilage &mws ectopically from the inner

surface of the synovid lining of a joint. The cmilage zmwths typicaiIy undergo central

ossification. malogous to endochondnl ossification. The cartilage in this case fXls the

joint cavity causing painful swelling and secondary oste~arthritis'~. Inheritance of

synovial chondromatosis is unusual, although a familial form of the condition has been

reportedjJ". The relative risk of malignant transformation to chondrosarcoma has been

estimated to be 5%'b'37. Current treatment consists of periodic surgical removal of the

ectopic cartilage and of the synovium to relieve symptoms and prevent the destruction of

articular

Chondrosarcoma is a malignanc y that occurs primarily in adul ts". Approxirnatel y

two thirds of chondrosarcornas are currently thought to anse de novo frorn otherwise

normal bone. and m l y from soft tissue. The remaining third of chondrosarcomas anse

from some precursor lesion. most commonly an enchondroma or an oste~chondroma'~.

A large number of cytogenetic denngements have been discovered in

chondro~arcomas'~. most notably at 9p?~'9. but none are universal. The only correlation

between rt cytogenetic change with s histologie subtype of chondrosarcoma is the

occurrence of a reciprocal t(9:22) trcinslocrition resulting in the fusion of EWS with

another gene. which generates a transcription factor with increased potency, in

extrikeletal myxoid c h o n d r o ~ a r c o m ~ ~ ~ ~ ' . Mice overexpressing c-Fos develop

chondrosrircomas and osteosarcomas*~ although expression of c-Fos has not been

demonstnted in human chondrosarcomaJ3. It has more recently been shown that c-Fos

expression is induced by Pxathyroid hormone related protein ( P T H ~ P ) ~ . Unlike in

osteosarcoma, Rerinoblasrornu (RB), p53 and 12q13-15 aItentions are not common in

chondmsar~om$~". nor is telomerase activity'8. Loss of hererozygosity for markers

linked to E;rTI and EX22 has been obsewed in sporadic chondrosarcomas, in addition to

those that anse in people with hereditary multiple exostose^"^^. Genetic changes that

have been observed in a chondrosarcoma but not in an enchondroma from the sarne

patient with enchondrornatosis include deletion of l p 11-3 1.2 in one casejO. and loss of

heterozygosity at 9p2 1 and l3q 14 together with overexpression of p53 in anothei".

Most histological subtypes of chondrosarcoma are relatively radiation and chemotherapy

resistant, and are treated with wide surgical excision alone". Major ablative amputations

or limb salvage procedures are often necessary, resulting in approximately 60% long-

term suwival from the diseaseI9.

Bone development

Development of a mammalian zygote proceeds by multiple rounds of ce11 division

to generate a morula of equipotent cells, Cavitation of the morula convens it into the

blastocyst. which has an eccentric inner mass of tells- The embryo proper is derived

from this inner m a s . which undergoes gastrulation in the third wcek of gestation in

humans. Gastrularion establishes the three distinct gerrn layers of endoderm. mesoderm

and ectoderm. Mesoderm is the germ layer thrit ail connective tissues are derived from.

Following the onset of neurulation at four weeks in humans, paraxial mesoderm

differentiates and sezments into 4344 pairs of somites in a cranial to caudal direction.

Instructed by secreted signais, in large part from the notochord, somites differentiate into

sclerotorne venuomedially and dermornyotome dorsolatedly- The axial skeleton is

derived from sclerotorne. Appendicular bones and tendons. including the shoulder and

pelvic girdles. are derived from lateral plate mesoderrn. CeIls from lateral plate

rnesoderm migrate to the prospective sites of the limb buds. In contrast, ceIls destined to

form rnusc1es. nerves and bluod vessels are derived from somitic mesoderm, which is

medial to lateral plate mesodem. and invade the limb buds secondarily to join the

growing ske~eton~'~~.

The early limb bud consists ofa prolifenting population of mesodermal cells

covered by ectoderm, Development of the limb bud is regulated by four regional

organizing centres "n. These include 1 ) the mesodermal progress zone from which the

skeletal elements are derived, 2) the mesodermal zone of polarizing activity which

primarily establishes anterior - posterior pattern polarity, 3) the apical ectodcrmal ridge

which is primarily responsible for Iimb outgrowth, and 4) the surrounding ectoderm

which primarily establishes dorsal - ventrtil pattern polarity. These regional orgrinizing

centres are CO-dependent. and together they coordinrtte outgrowth of the limb dong the

three orthogonal axes. The basic pattern of the entire skeleton is cornplete by the end of

organogenesis. which takes place between weeks four and eight of gestation in the

humün.

Most of the bones of the vertebnte skeleton. with the exception of the craniofrtcial

bones and part of the clrtvicte. are tnnsformed fmm mesodermally - derived mesenchyme

to bone by the process of endochondnl ossifi~ation'~'~. In this process. condensed

mesenchyme which prefigures the primitive shape of a bone first diffe~ntiates to

cartilage. Ossification begins in the centre of the cartilage template and proceeds toward

either end, trailing behind the receding zone of cartilage known as the growth plate. In

mammols, a secondary centre of ossification forms at the two extreme ends of the

cartilage template. The growth plate is responsible for longitudinal growth, and shnnks in

height until it disappem following adolescence in human bones, but rernains present in

the long bones of other mammals induding rodents. Chondrocytes in the growth plate

differentiate toward the p n m q centre of ossification (Fig. 2). Resting chondrocytes first

enter a prolifemtive phase, followed by a post-mitotic hypertrophic phase at the end of

which they mainly undergo programmed cell death. The cartilage matrix produced by

growth plate chondrocytes changes in composition from predominantly Type U cohgen

(Colti) in the proliferitive zone to predominantly Type X collagen (ColX) in the

hypertrophic zone. where it also calcifies. This calcitied cartilage is removed by

ch~ndroclasts~':~~, and forms the scaffold on top of which primary bony tnbeculae are

laid by osteoblasts which amve by way of invading blood vessels. An understanding of

current models of the molecular regulation of chondrocyte differentiation will be

hcilitated by a synopsis of two key signal trinsduction pathways.

Hedgehog signal transduction

The importance of Hedgehog signalling as well as the identity and function of key

Hedzehog signal transduction mcdiators was first established by work in Drosophila.

where it contributes to the segmentation of embryos and the patterning of imaginai-disk

our_mowth "':". Hedgehog signalling has since been discovered CO be essential for the

regulation of vital vertebrate embryonic processes as well ris for the development of

many organ systems. These include differentiation of visceral endoderm. the

establishment of left - nght asymrnet$'. somite patteminebJ and differentiat~on~"~.

0'1 central nervous systern patterning and differentiati~n~'"~'. spermatogenesis .

specification of hematopoietic and endothelid ~el ls '~, pancreatic3 and intestinal

1 .la75 developmen? . haïr cycle regulation76. tooth development7, limb patterning and

o~tgrowth'~. and regulation of c h o n d r o ~ ~ t e ' ~ ' ~ ~ and osteoblast differentiationsos'-

Hedgchog signalling has been co-opted to regulate a variety of celluiar functions, and

dysregulation of this signalling pathway is known to cause disorders of pattern and of

proliferation in humans. Patterning anomalies involve the central nervoüs system,

viscera and limbs. and include holoprosencephalys2, midline faciai anomalies. defects of

the kidneys. hean. and genitalss3, syndactyly and po1ydactylyw'85. Neoplasms are caused

by excessive Hedgehog signalling, and include sporadic and familial basal cell

carcin~rna '~ '~~, medull~blastoma~~ and rhabdomyosarcom$O.

The Hedgehog signal transduction pathway is conserved between species. There

are three homologs of Drosophiln hrdgrlrog in rnammals: Sonic hedgehog (Sirit), Mian

iredgeliog (Illh) and Drsen iledgehog (~hh)""". Shh is the rnost wide1y expressed and

best studied of these lisands. The humnn IHH gene is located on chromosome bands

2q33-35 'j3. It shares high sequence similarity with SHH"'. is processed in a similar

manner '', and either protein can substitute for the function of the the?'"^. fih is

expressed by differentiating extraembryonic end~derm"'~'. kidneyqm. gut". growth plate

~hondroc~tes'~ and osteoblasts'M. The role of Dhh is the most resrricted of the Hedgehog

100.101 Iigands. Dhh regulates spermatogenesis and organization of the perineurium .

Hedgehog ligands are processed post-translationally by autocleavage of a 45 kD

precursor, releasing a 19 kD amino (N) terminal signalling domain and a 26 kD carboxy

terminal domain. The N-terminal sigalling domain is funher modified by covalent

addition of a cholesterol moiety by the carboxy terminal domain which acts as an

intramolecular cholesterol transfeme9'. N-terminal hedgehog protein can be secretedlO<

1 O 3 and can signal at short as well as long range . The cholesterol rnoiety may alIow the

ligand to be tethered to ceIl surfaces. and regulate the distance over which it acts,

although it is not yet clear whether cholesterol modification restncts or enhances

diffusion of hedgehogt8. A twelve-pass transrnernbnne protein with a sterol sensing

domain called dispatched has been identified in Drosophila which likely displaces

hedgehog from the lipid bilayer to allow effective signallinglM. Patched (see below), a

distant relative of dispatched 'O5, and Hedgehog interacting protein'06 contribute to the

spacial restriction of Hedgehog ligands. Exthout velu is a glycosyl transferase that

107;108 induces the expression of ce11 surface heparan sulphate , and is essentiai for long

range diffusion of the hedgehog ligandtw.

Hedgehog signal transduction is best understood at the genetic level (Fig. 3).

Work in Drosophila has shown that ilil and prc have opposite functions in regulating

expression of wingless and clrcapentaplegic. Mutants of purchrd (ptc). which encodes a

twelve pass transmembrane protein, phenocopy iiedgehog (irii)lprc double mutants,

indicating that in the absence of prc, iiii is not nece~siuy"~. Mutants of snioori~ened

(smo), which encodes a seven-pass transrnembnne protein with homology to G protein-

coupled receptors. display similar phenotypes to iih mutants. Snio is required for al1

Hedgehog signalhg in Drosophila. When embryos lack both mro and ptc. they

phenocopy smo single mutants. indicating that prc is genetically upstrearn of srm. In the

absence of both prc and hh. snio activity leads to constitutive activation of Hedgehog

tqe ts . In the absence of Iih, ptc represses smo activity. This repression is relieved by

LI 1-1 I j the presence of irh. thereby allowing activation of Hedgehog targets by snto . Ir was

shown in vertebrate cells that labelled Hh protein binds Ptchl (vertebnte ptc) -

expressing cells with high affinity, and that Hh and PtchL, but not Hh and Smo, c m be

Il-kl15 coimmunoprecipitated and crosslinked . How this binding is communicated to Srno

is not clear. A direct rnechanisrn is suggested by biochernical studies showing that

overexpressed Smo and PtchL form a cornplex'". An indirect mechanism is also likely at

work, as it was shown that ptc destabilizes smo in the absence of hh, and that hh binding

to ptc removes ptc from the celt surface while causing phosphorylation, stabilization and

accumulation of smo at the cell surface'L6. It was shown indirectly in a frog melanophore

assay that SM0 may acrivate Gcq and inhibit CAMP synthesis'17. A second mammalian

Ptch gene, Ptch2. has been isoIated which Iacks a carboxy-terminal domain and is

expressed in the testis and skin. Unlike Prchl, Prcld expression is not restricted to Hh

113-120 target cells. but is coexpressed by Shh-expressing cells . Mutants of PtcIi2 have not

yet been described.

So fa. al1 Hedgehog signalling appears to activate zinc finger-containing

transcriptional effectors of the ci/Gli family, In Drosophila. cubitus interruptus (ci) is

constitutively cleaved in the absence of hedgehog sigrilling, generating a transcriptional

repressor that binds to. and inactivates, hedgehoz t q e t senes. Hedzehog signalling

inhibits this proteolytic cleavnge by a phosphorylation - dependent process. As a result.

full-length ci protein activates tmnscriptional t~~ets"';"'. There are three vertebrate

homologues of ci. known as the GIi (after glioblastoma) famiiy of transcription factors.

GliZ and GIi3 both have an N-terminal repressor domain and a C-terminal activator

domain. whereas Gli 1 has onIy an activator domainI3. The activator and repressor

capabilities of GIi2 and Gli3 are context - dependent. [n a n t neural stem ce11 cell line

(MNS70). full length GIii and GIi2 can activate whereas Gli3 cm repress a Hedgehog

responsive reporter %ne. Expression of N-terminai truncated foms of either Gli2 or Gli3

can ectopically activate Shh-target genes in the dorsai cenml nervous ~ ~ s t e r n ' ~ . In the

limb bud mesenchyme, Glil is upregulated whereas Gli3 is downregulated in response to

~hh'?". In the presence of ectopic Ihh in the anterior limb bud of the Doublefoot mouse

mutant, Gli2 is downregu1ated9', in contnst to what might be expected from the in vitro

studies mentioned above. In flies, the activity of ci is modulated by its interaction with a

microtubule - associated complex containing sevenI other proteins. These include

costal2, which negatively regulates ci nuclear impon but is required for ci

13;126 activation , fused, a proiein-serinelthreonine kinase and an activator"', and

suppressor of fused. a repressor"! Mouse Suppressor of fused has been shown to bind to

al1 three Gli proteins. and inhibits Gli 1 action by preventing nuclear impon'29. Protein

130-.131 kinase A (PKA) has been regarded as a conserved inhibitor of Hedgehog sigalling .

Indeed, expression of a dominant negative fom of PKA mimicks eciopic Hedgehog

expression1". and phosphorylation of ci by PKA is required for proteolytic cleavage of

ci'". However. there is contnsting data from sepante groups of investigators with

regard to the effect of PKA on ci activation. Wang, G. et al showed that P M inhibits the

full len~th activator form of ci':" whereas Wang, Q.T. et al showed that ci activation

requires phosphoryiation by PKA':~.

PTHrP signal transduction

PTHrP and its receptor, the Tvpe 1 PTKfPTHrP rrcepror (PrlirI) are expressed in

the pregastrulation e r n b ~ ~ o [ ~ ~ . in extraembryonic tissuesE6 and in tissues derived from al1

three germ laYers1". PTHrP functions c m be divided into four ~ategones"~. First,

PTHrP stimulates transepicheIiai calcium cransport in a varïety of tissues. Second, PTHrP

is a smooth muscle relaxant in the vasculru systern and in a broad range of viscen.

Third. PlThP is a regdator of prolifention, differentiation and apoptosis in many ceIl

types. Finally, PTHrP regulates a number of pre- and post-implantation developmental

processes. In most sites, F i R P and its receptor are cwrdinateiy expressed in adjacent

cells, irnplying that these proteins are involved in pancrine signdling p t t ~ w a ~ s ' ~ ~ . For

example, PTHrP is expressed by the perichondrium, whereas Pthrl is present on a subset

79;tU) of proliferating ,gowth plate chondrocytes . In exuaskeletal sites PTHrP is

predominantly expressed in the surface epithelium of developing organs, and Pthrl

resides on the adjacent mesenchymai cel1s'j9-

PTHrP is the product o fa single gene, which is highly sirnilx in organization and

sequence CO the PTH gene'41. (Its functions. howcver, are more similnr to those of

developmental factors'".) Through altemarive splicing the multiexonic PTHrP gene

oives rîse to t hree initial translation produc ts. These products undergo extensive - postrrinslationril processing and gîve rise to ri frimily of mature secretory f o n s of the

protein. each with its own physiologicül functions and its own receptoror receptors. The

arninu-terminal secretory form, PTHrP (1-36). is the rnost relevant form with regard to

chondrocyte di fferenciation. and acts through Pthr 1 lJ'.

Two mammaiian PTWPTHrP receptors have been isolated. PTHRl is the

clasicai receptor. and binds PTH and ETHrP with equd nffinity'S3. FTHR2 binds PTH

with hi$ affinity, but noc P T H ~ P ' ~ . PTHR3, which has so far been isolated from

zebnfish, binds PTHrP. but nor PTH"'. PTHRL is a seven-pass rnnrrnembnne protein

143~146 that is linked to heterotrimeric G proteins . E have focused attention on PTHRl

becriuse it is the key mediator of paracrine MH-PïHrP signalling in the growth phte. A

family of related proteins includes receptors for Secretin, Growth hormone-releasing

hormone, Vasoactive intestinal polypeptide, Type 1 gastric-inhibitory polypeptide,

Glucagon, Glucagon-like peptide 1, Corticotropin-releasing factor, and Pituitary

adenylate cyclase activating peptide 1 IJ7.

G proteins are membrane - associated, and are composed of an u subunit that, in

the inactive state of an associated receptor, hydrolyses guanine triphosphate (GTP) to

guanine diphosphate (GDP), and is loosely bound to a tightly associated dimer made of B

and y subunits. Upon binding ligand, the receptor acts upon the G proteins. causing GDP

to be replaced with GTP, As a result, the a subunit carrying GTP dissociates from the

two other subunits. Either the monomer or the dimer then acts upon a target effector

protein that is often also membrane associated, which in tum reacts with targets in the

cytoplasm. generating second messengers. Various G a proteins define different G

protein trimers (Gui, Ga,,, GG, Gu,, and G g ) , each of which regulates a

14%149 distinctive set of downstream signalling pathways . PTHRI most strongiy

IJ3:146 associates with G c and G q . Activation of G q stimulates Adenylate cyclrtse.

which generates the second messenger Cyclic adenosine monophosphnte (CAMP) (Fig.

4). Activation of G q stimulates Phospholipase C to generate Inositol triphosphate [iiJ;).

PTHR1 hrts been shown to weakly associate with Gcq, which inhibits Adenylate

c y c l a ~ e ' ~ . Cyclic A i i binds the regulatory R subunit of P M . causing relerise of the

catalytic C subunit. Some of these C subunits then diffuse into the nucleus, where they

phosphorylate CAMP response element binding protein (CREE). Fhosphorylated CREB

c m then bind to the response element CRE. which is found on genes whose transcription

is induced by CAMP'^"^^.

PTHRl has also been shown to signal via altemate mechani~ms'~~. These include

activation of the iMAP kinase pathway15', andan intracrine mechanism involving

nucleolar targeting of PTHrp15'. Intemalization of the P'HrP/PTHRI complex has been

['?154 dernonstrated , but how this phenomenon relates to the intncellular signalling

pathways has yet to be shown. E'THRl is aiso known to eiicit differing and even opposite

cellular responses depending on whether it is stimulated continuousIy or intemittentlyL5'.

Regulation of chondrocyte differentiation

Growth plate chondrocyte proliferation and differentiation are tightly regulated.

Of the molecular sigals that regulate these processes, Ihh and PTHrP have been show

to play major rotes (Fig. 2). Ihh is expressed by prehypenrophic and hypertrophic

chondr~c~tes"'~'. PTHrP is primarily expressed by perichondrial cells surrounding the

growth plate, in a region near the anicular surface 79:s I:LSG . The Iikely Ihh receptor Ptch i .

initially thought to be expressed exclusiveiy in the perichondnum79. has been shown to be

expressed also by chondrocytes in the proliferative zone in mices0.

Although they are intimately related, the individual functions of Ihh and of PTHrP

are Iargely distinct in the growth plate. In mice lacking the secreted ligand PTHrP or its

receptor Pthr 1. hypenrophic differentiation of proliferating chondrocytes is dnmritically

accderated. indicating that PTHrP sigalling is responsible for delaying differentiation of

156-161 proliferating chondrocytes . It has been shown that the ability of PTHR1 to induce

CAMP accumulation by activatins Gu, is responsible for chondrocyte difFerentiation

de1ayU. In conmt. the abiIity of Pthrl to cause PI accumulation by activating G q is

associated with accelerated hypertrophic differentiation. This was shown in a knock-in

study where mutation of Pthrl that seIectively inactivated its ability to stimulate Ger,

resulted in delay of hypemophic differentiationib'. Therefore. Pthrl has a dual role in

regulating chondrocyte differentiation, rtlthough its ability to dehy hyperuophy seems to

predominate, Mice lacking Ihh have very short Iimbs and a profound defect in

chondmcyte pmliferationsO. Alrhough hypertrophie difFe~ntiation is also accelented in

these mice, it is associated with a loss of PTHrP in the periuticular perichondrium,

indicating that PTHrP is downsueam of lhhmo. In fact, IhhPTHrP doubte homozygous

mutant mice phenocopy lhh single mutant mice. demonstrriting that il111 is tequired for

P7HrP functioni6'. UpreguIation of P T W expression in rhe periarticular

penchondrium by Ihh is probably indirect, since PTCHI is not expressed in thnt region

of the perichodrium. Ihh driven chondrocyte proiiferation is Iargely independent of

PTHrP. since expression of a constitutively active PTHRl in l i r j ~ ~ ~ mice could not rescue

the defect in profifentiontb3. Therefore. the mitogenic function of Ihh is likely mediûted

directly by the Hedgehog receptor complex expressed by prolifenting chondrocytes.

PTHrP downregulates I M , likel y by both direct and indirect means. Delay of

chondrocyte differentiation by PTHrP may decrease the number of prehypenrophic and

hypenrophic cells. and thereby indirectly downregulate Ilth. A direct mechanism is

implied by the finding that inhibition of lhlr expression by FTHrP in virro does noc

require protein synthesislU.

Many other signais contribute to p w t h plate regulation, For example,

chondrocyte prolifention is stimulated by Insulin-like p w t h factor 1 ( ~ ~ f - p . basic

Fibroblast growth facror ( ~ g f j ' ~ ~ and CREB'~'. Tnnsforming &pwth factor betaLb8 as

well as Delta1 - activated Notch sig~aIlin$~ delay hypemophic differentiation. whereas

Core binding factor alpha li70, ~gf-li7l, T h p i d h~rmone'~', Connective tissue growth

factor17' and C R E B ' ~ ~ promote this process. There are conflicting reports as to the effect

174: 175 of Retinoic acid on chondrocyte hypertrophy . Bone morphogenic proteins (Bmp) 2

and 4 may mediate the induction of PïiirP by Ihh by means of the Bmp receptor

Bmp-6, which contributes to hypertrophic differentiation, may partially mediate the

inhibition of IHH by P T H ~ P ' ~ ~ . Sox-9 likeiy panially mediares the effects of ETHrP with

regard to delay of chondrocyte h y p e r t r ~ p h ~ ' ~ ~ . Activation of the Fgf receptor 3 inhibits

expression of ~ h l i ' ~ ~ . Matrix metalloproteinase 9IGelatinase B is expressed by

chondroclasts and regulates apoptosis of hypertrophic chondrocytes. removal of cartilage

at the chondro-osseous junction and. tosether with Vascular endothelial growth factor,

- 59:60 reguiates growth plate angiogenesis . Galectin 3 is important for chondrocyte survivlil

and may help coordinate chondrocyte death with vascuiar invasion'80.

HYPOTHESIS

Cartilaginous neoplasia is due. at lest in pan, to dysregulation of the MH-PTHrP

sigalling pathway.

RATIONALE

i've chosen to focus on the roie of the iHH-PTHrP pathway in pmicular in

cartilaginous neoplasia for three main reasons. Fïrst, as noted above, certain cartilage

tumours, inciuding enchondromas and chondrosarcomas, share cytological and matrix

features with gowth plates, aibeit in an unorganised fashion. Morphological similarity

naturally implies some d e p e of molecular sirnilarity, and these moiecules rnight include

regulators of ce11 behaviour. I find it difficult to conceive of enchondromas and

chondrosarcornas as not k ing responsive to MH and PTHrP, the most potent regulators

of chondrocyte proliieration and differentiation discovered to date. Second, from a

therapeutic perspective, whether activation of the MH-PTHrP pathway plays a causal

role or is merery a phenomenon secondary to some other factor causing expression of a

chondrocyte phenotype in canilage tumours might be IargeIy irrelevant. Differentiation

of neoplastic chondrocytes with agents that interfere with LHH and PTHrP signülling

might be possible in either case. Third, blockade of developmentally important sigalling

pathways is attractive because of the relatively selective effect such blockade would have

beyond the stage in an individual when that pathway is no Ionger as important. For

exmple. one particular agent which specifically blocks Hedgehog sipalling.

[S:18-I cyclopamine , is known CO be harmless to adult ewes who eat the substance, while it

is toxic to their embryos. This agent and others like it might differentiate canilage

tumours with minimal side effects.

OBJECTIVES

To investigrite the validity of the hypothesis. the following questions will be

addressed:

1) 1s the MH-PTHrP signalling pathway expressed and active in cartilage - formin$

tumours?

3) If SO, is activation of IHH-PTHrP signalling merely o secondary phenomenon in

cartilage tumours, or can dysregulation of this pathway initiate neoplasia?

3) Cm knowledge of how iHH-ETHrP signalling is aitered in pathologicai conditions

contribute to the understanding of normal growth plate deveIopment?

Figure I

growth

I enchondroma

epiphysis

metaphysis

diaphysis

Enchondromas m most commoniy fouad adjacent to the metaphyseal side of a gcowth plate. and may be connected to the growth piate by a core of cartilage. Enchondromas might be caused by a focal failure of endochondral ossification.

Figure 2

Proliferative

Preh y pertrop hic

Hypertrophie

Programmed Cell Death

P T H r P

I h h

Growth plate chondrocytes differentiate in a columnar fashion. toward the centre of ossification (dowaward). A fixed nurnber of reserve cells fust enter a proliferative state, foiiowed by a pst-mitotic state of hyperuophy, at the end of which they undergo programmed ceii death. Chondroclasts resorb the calcified matnx at the chondro-osseous junction. Blood vessels invade the distai p w t h plate. ailowing osteoblasts to access the caicified matrix and lay new bone.

ihh is secreted by prehypertropbic and hypertrophie chondrocytes. Ihh induces proliferation of chondrocytes as well as secretion of a second secreted sipnd, PTHrP, from the penarticuiar perichondnum. PTHrP signals its receptor on prolifenting chondrocytes to delay hypemophic differentiation.

Figure 3

Cos2

Fu

Su(Fu) Ptch

Gli

Hedgehog signal is conserved across species. Bindîng of a hedgehog ligand (Hh) with Ptch relieves the otherwise constitutive inhibition of Ptch on Smo. Smo then induces expression of Hedgehog target genes by way of the Gli family of transcription factors.

J PKA

Stimulation of PTHR1 by PTHrP acts on membrane associated G proteins to repIace GDP with GTP, allowing the a subunit to dissociate from the Dy dimer. Gas (pictured here) stimulates adenyIate cyclase to produce CAMP, which in tum reIeases the catalytic subunit of PKA, allowing it to diffuse into the nucleus and phosphoryIate CREB. Phosphorylated CREB then binds to response elements on genes whose transcription is induced by CAMP.

Uncoupleci lndian hedgehog - Parathyroid hormone related protein signalling in

cartilage tumurs and a mutant Type 1 PTHPI'HrP receptor in enchondromatosis

Hopyan. S.*, Gokgoz, - N., Poon. R., Bell. R.S.. Cote. W.G.. Andrulis, I L . , Wunder. J.S.,

Alman. B.A.. Submicted for publication

*Performed al1 procedures other than SSCP, sequencing, pronuclex microinjection and

immunohistochemistry.

SUbiMARY

Enchondromatosis is a condition in which multiple enchondromas predispose

patients to developing skeletal de formity and c hondrosarcoma. 1 investigated whether

dysregulation of Indian hedgehog - Parathyroid hormone related protein signailing (EH-

PTHrP), which regulates chondrocyte proliferation and differentiation, contributes to the

genesis of enchondrornas. Here 1 show that IHHPTHrP signalling is functional but

dysregulated in enchondromas and chondrosarcomas. A mutant Type 1 PTWPTHrP

receptor (PTHRI) was discovered in human enchondromatosis that signals abnonnally in

vitro and causes enchondmmatosis in transgenic mice.

INTRODUCTION

Enchondromas are common benign cartilage ttumours of bone. They can occur as

solitary lesions, or as multiple lesions in enchondromatosis (Ollier's and Maffucci's

diseases). Clinicd probkms caused by enchondromas include skeletal deformity and the

potencial for midilant change to chondr~sarcorna'":'"~. The extent of skeletal

invoIvement is variable in enchondromatosis, and may include dysplasia that is not

directly attributable to enchondr~mas~~. Enchondmmatasis is me, obvious inheritance of

the condition is unusual, and no candidate loci have been identified. Enchondrornas are

usually in close proximity to, or in continuity with, growth plate cartilage. ConsequentIy.

they may result from abnonnal regulation of proIiferdtion and terminal differentiation of

chondrocytes in the adjoining growth plate, In normal growth plates. differentiation of

proliferative chondrocytes to post-mitotic hypertrophic chondrocytes is regulated in pan

by a tightly coupled signalling relay involving Parathyroid hormone related protein

(ETHrP) and Indian hedgehog (IHH) 79:SO: 156: 163 . PTHrP de Iays the hypenrophic

differentiation of prolifenting chondmcytes. while iHH promotes chondrocyte

proliferation. I speculated that dysregutation of EW PTHrP signdling contiibuies to the

genesis of enchondromas.

RESüLTS

Expression of MIUPTHrP pathway rnembers in cartilage tumours

In order to test for expression of key IHH/PTHrP sigalling pathway rnembers, I

utilized reverse cnnscrÏption - poIymerase chah reaction (RT-PCR) maiysis. 1 found

that IHH and PTHrP were both expressed in 4/4 human enchondroma and 23/33

chondrosarcoma specimens, but less consistently in human articular cartilage (IHH 014,

PTHrP 214). cortical bone (IHH 2/11, PTHrP 1111 1) and other benign cartilage tumours

including chondrob~astomas'~ (IHH Y7, PTHrP 37) and mature osteochondromas"

(IHH 012, PTHrP 2/2). 1 therefore tested for expression of PTHRI, the MH receptor

Patciied 1, and the Hedgehog responsive transcription factors GLII, GU2 und GU3 in

enchondromas and chondrosarcomas, and found that they were al1 consistently expressed

in these iesions (Table 1). Semiquantitative RT-PCR analysis revealed that the ratio of

expression of (GLII +GU2)IGL13 was higher in enchondroma ( 1.5, n d ) and

chondrosarcoma specimens ( 1.5, n= 1 1) than in normal articular cartilage (0.34, n 4 ) and

cortical bone (0.40. n=l 1) (Table 2). suggesting that cells in these pathologie tissues are

uansducing the MH signal, since GU1 and CL12 are genenlly upregulated and CL13 is

'*.I81 downre_oulated in response to Hedgehog ligandt' .

Functional but uncoupled IHHmrHrP signalling in cartilage tumour explants

To test whether the FT'HrP and MH signalling pathways are functional in

cartilage tumours. short term primary organ cultures of enchondromas from two patients

and chondrosarcomas from ten patients were established. Treatment of neoplastic

cultures for four days with ~ o - ~ M KHrP (1-34) resulted in delayed hypertrophic

differentiation ris assessed by decreased expression of Tvpe X collagrn (COWC). an

exclusive marker of hypertrophic c h o n d r ~ c ~ t e s ' ~ ~ (p3.O 1, Fig. la). FïHrP treatment

had no effect on neoplastic chondrocyte prolifention. as measured by tntiated thymidine

('H-T) uptake (p3.53). Conversely, treatment of cultures with 10 @ml recombinant

Sonic hedgehog (Shh-N. which is functionaIly redundant with ihh in ~ulture'~) increased

'H-T uptake (p=0.02, Fig. 1 b) without effecting COLX expression level (p=0.39).

Treatment of cultures with agents that specifically block Hedgehog sigalling. namety

I8':LiU LO-'M cyclopamine (Fig Ic) and 10 pgml neutralizing anti-pan-Hh-N

5 ~ l a n t i b o d ~ ' ~ ~ partially decreased 'H-T uptake (n=6. p=0.07 for cyclopamine and n=E,

p=û. 1 1 for anti-pan-Hh-N antibody) but had no effect on COLX expression (p=0.32,

p=0.39, respectiveIy). These findings are consistent with the known consequences of

&O163 and p~~~pI61:186 signaIIing on normal chondrocyte differentiation and

proliferation, respectively. In the growth plate, these processes are CO-regulated because

iHH induces PTHrP expression. and PTHrP in turn feeds back negritively on iHH

79:156 expression . I did not detect downregulation of 1HH following treatment with PTHrP

in any of the turnour cultures. Downregutation of IHH was. however. observed in 4/4

growth plate cultures (Fig. Ld). This feedback mechanism is therefore not operative in

enchondromas and chondrosarcomas.

A mutant WHR1 in enchondromatosis

A mutant PïKRl could account for the lack of downregulation of IHH by FJXrP

in enchondroma and chondrosarcoma- To test for this possibility. single stnnd

conformation polymorphism analysis and manual sequencing was undertaken to screen

cartilage tumour specirnens for PTHRl sequence alterations. A heterozygous cytosine to

thymine change. resulting in the substitution of iuginine with cysteine in the extracelluiar

domain (R150C) of the PTHR1 in enchondroma specimens from two of six patients with

enchondromatosis (OIlierTs disease), was discovered (Fig. 2a). Both of these patients

were male, with mild to rnodente disease seventy. In one of these patients, the variant

RISOC PTHRI aiIele was cmied in the _oermIine, and was inherited from the hther who

has mild spondyloepiphysed dysplasia, but no evidence of enchondromas on

ndiographs, simiIar to a scenario reponed previously29. In the other patient. the RI5OC

PTHRI allele represented a somatic change, since normal bone adjacent to the

enchondroma specimen did not contain the variant allele. It is possible that this second

patient is a mosaic carier of the RISOC PTHRl allele, but 1 was not able to test for this

possibility since enchondroma tissue from only one site was available for study. The

RlSOC change was not found in DNA samples from 100 unaffected individuals, nor in 50

spomdic chondrosarcoma specimens,

Mutant PTHR1 signalling is impaired in vitro

The signalling ability of the PTHRI has k e n most extensively studied in the

context of its G protein - linked abiiity to induce Cyclic adenosine monosphosphate

(cAiiP) and Inositol triphosphate (P3) accumulation upon ligand binding 143: t17-180 . The

1'10- 19-4 functionül relevance of the arginine 150 residue is not known . Cyclic AMP and IP;

assays were performed using COS-7 cells and embryonic stem (ES) cells lacking both

copies of the native Pthrl gene, respectively. Wild rype (Wi) and Rl5OC PTHRl cDNAs

were generated by site - directed mutrigenesis and tnnsiently trinsfected for analysis.

Other PTHRI mutants, known to cause skeletai dysplasias with phenotypes distinct from

that of enchondromatosis. include CAMP activatin; mutations in Jansen metaphyseal

1S7:188 chondrodysplasia , and a CAMI? Ioss of function mutation in Blomstnnd

1 8 9 1% chondrodyspiasia . These mutants were also generated and tested in the CAMP

assays as intemal controls. Surpnsingly, expression of RISOC PTHRI in COS-7 ceils

lowered the basal level of CAMP in cornpaison to empty vector or WT PTHRl -

transfected ceIls (Fig 2b. d)- RING PTHRl - transfected cells could still accumuhte

CAMP in response to ligand in proportion to WT PTHRI - transfected cells. Imponantly.

RljOC PTHRl abolished the ability of transfected cells to accumulate IP; in response to

ligand (Fig. 2c). When increasing arnounts of the RIjOC PTHRl variant was

cotrmsfected with a fixed arnount of WT P T H R l , the variant dernonstrated a dominant

negative effect on CAMP level (Fig. 2e). 1 next tested the expression level and

subcellular localization of c-Myc-tagged WT and R150C PTHRI proteins by Western

andysis and confocal immunofluorescence. With these tests, 1 found no gross change in

mutant receptor expression or localization (Fig. 20.

Transgenic mouse mode1 of enchondromatosis

To evaluate the effect of the RL50C P ' l mutation in vivo, 1 used the rnurine

196: 1 O7 Tvpr II collugrn (ColIl) promoter and enhancer to drive expression of IVT PTHRI

or RITOC PTHRl in the growth plates of trmsgenic rnice (Fig. 3d). Mice expressing WT

PTHRI in two sepante founder lines did not display any rnorphological abnormalities

compared to wild-type littermates. At binh. growth plates of mice from three sepamte

founder lines expressing R 150C PTHR 1 were of similar ovenll size. but had shoner

hypertrophic zones. than those of rnice expressing WT PTHRl (Fig. Ja, b). This

difference was -matest in growth plates which contribute the most to longitudinal

orowth. such as the proximal humerus compared to the distal humerus. and was *

confirmed by immunostaining against COLX to highIight the hypertrophic zone (Fig. 4c.

d). Beyond eight weeks of age, features consistent with human enchondromatosis were

evident in two of these R150C PTHRl founder lines. Paraffin exnbedded sections stained

with safranin-O to highlight cartilage revealed cellular cartilage islands. or rem, in the

metaphyses of multiple long bones (Fig. 5), including the proximal humerus (Fig. jbe),

distal femur (Fig. 5f) and tibia. CoIumns of cartilage commonly connected the rests to

the adjoining gowth plate, as is often found in human enchondromatosis (Fig. jb, c).

Regions of varying cellularity, with prolifenting cells and occasional binucleate lacunae,

amnged in lobular patterns in a hyaline cartilage matnx that was partially or completely

surrounded by bony matrix (Fig. 5d. e), similar to human ench~ndromas'~, were

obse~ed. Another similuity to human enchondromatosis was the persistance of these

lesions well into adulthood (Fig. 5c, e). The paucity of hypertrophie cells in the growth

plates of R 150C ETHRL mice persisted beyond adolescence (Fig. 6ri. b), and may have

resulted in the inclusion of transitionat celis. which were still competent to proliferrite.

within the trabecular cores of cartilage (evident as thin. red longitudinal streaks in Fi:.

5a. b) to :ive rise to metaphyseal enchondromas. Funher. the origin of the

enchondromas was likely due to abnormal prolifemtion/differentiation rather than to a

deiect of resorption. since chondroclascs, cells which resorb calcified cartilage at the

cartilage - bone interface and are distinguished by their ability to retain an acid

phosphatase stain following tanrate challenge (TRAP stain)". were found in equal

numbers in WT PTHRl and R150C PTHR1 mice (Fig. 6c, d).

DISCUSSION

Enchondroma formation

The results implicate abnorrnal growth plate development in the formation of

enchondromas. In particular, some cases of enchondromritosis are Iikely caused by a

heterozygous, dominant mutant PTHRI. The cysteine residue in position 150 might fom

new or alternate disulphide linkages with other cysteines that are present in the

exmceilular arnino terminus or in the extracellular loops between the tnnsmembrane

portions of PTI-Rl1"'. In so doing, the RlSOC change might alter the structure of

PTHRL in a way that changes its signalling characteristics.

The R150C PTHRl variant may be involved in the development of

enchondromatosis by altering chondrocyte differentiation and prolifention. Prevention

of IP3 accumulation in response to ligand is consistent with delayed chondrocyte

hypertr~ph~'~'. Baseline CAMP level, which has the opposite effect as P;, was

suppressed by the RLSOC variant, possibly due to suengthened association with Gai.

However, accumulation of CAMP was still observed in response to PTHrP. The

invacellular balance of second messengers is therefore tipped in favour of CAMP over IPj

in the presence of ligand. which is Iikely responsible for the more distal transition from

proliferation to hypenrophy seen in RISOC PTHRI mice. Cyclic AMP. through

activation of protein kinase A (PKA). is a conserved inhibitor of Hedgehog signalling1j2.

Therefore, ovenctive IHH sigalling might be caused directly by the R150C PTHRI

variant, Ieading to inappropriate chondrocyte proliferation. Both delay of growth plate

chondrocyte differentiation and maintainance of prolifentive ability are conceivribly

necessary for enchondroma formation. In RISOC PTHRl trrinsgenic mice. proliferating

chondrocytes were abnormally close to the chondro-osseus junction. increasing the

likelihood that some of them would be tnpped in cartilage cores that normally extend

into the metaphysis, where they could give rise to enchondromas.

Genetics of enchondromatosis

Different PTHRl mutations cause distinct phenotypes. which underlines the

importance and complexity of this protein. Delay of chondrocyte hypertrophy cm also

be achieved by PTHRI mutations which augment CAMP level in the absence of

ligand163; 187 . However, these mutations are not consistent with the enchondromatosis

phenotype. Elevation of CAMP IeveI causes systemic hypercalcernia and

hypophosphatemia which are Sound in Jansen's metaphyseal chondr~d~s~las ia '~~ , but not

in enchondromatosis'". The R150C mutation rilters siyalling in a manner that is

apparencly specific to enchondromatosis.

It is possible that enchondrornatosis is more commonl y familial than previousl y

thought. Variable penetrance or anticipation of the condition mriy rnask inheritance, as in

the case reported here of hther - to - son transmission of the RI5OC PTHRI allele. In

the other case reported above, the R150C PTHRl change was somatic. Mutations that

cause enchondromatosis might anse as somatic mutations during embryogenesis, and

subsequenriy be distributed in a mosaic fashion in the hosr. The distribution of the

mutant allele mighc determine the extent and s~verity of the phenotype. if the mutant

dleie is distributed outside of the musculokeIetd system. germline transmission rnitht

Occur-

Cases of enchondrornatosis without a demonstrable FTHRl mutation. ris weli as

solitary enchondromas, likely anse from similar molecular mechanisms. except that the

primary gene defect may involve other members of the ü-IHiPïHrP signalling pathway,

Active Hedgehog sipdling, in pmiculm. is likely an important component in the genesis

and maintenance of enchondromas. Evidence for this notion is found in the observations

that 1) IHH could not be dowreylated in the tumour explant cuItures, 2) tumour

proliferation wris dependent upon Hed$ehog sigaiiing, and 3) the R150C PTHR1

mutant suppressed CAMP levef, and therefore inhibited activation of the Hedgehog -

inhibitory Pm. Togethet with these findings, observation of elevated levels of GU1 and

GL12 and depressed levels of GLI3 expression in the cartilage tumours suggest that the

former are positive regulators while the latter is a negative regulator of Hedgehog signal

transduction in chondrocytes. Whether this suggestion is true and whether excessive

Hedghehog sigalling is sufficient to generate enchondromas are testable hypotheses.

The persistence of growth plate tissue in the form of enchondrornas beyond

adolescence, when growth plate tissue has normally disappeared in humans, may allow

accumulation of secondary genetic events which cause chondrosarcoma to anse within a.

preexisting enchondroma. One way a chondrosarcoma might anse is through a mutation

that establishes a clonal population of chondrocyte stem cells. These stem cells would.

by definition. be capable of self-replication as wdl as seneration of daughter cells chat

could differentiate as +mwth plate chondrocytes do. Thcsc latter cclls would then be

dependent upon MH and ETHrP. consistent with the findinps in the explrint cultures

described tibove.

METHODS

Tissue specimens

Tissues were obtained from the Nationai Cancer Institute of Canada Sarcoma

Tissue Bank housed at Mount Sinai Hospital, the Mount Sinai Hospital Bone Bank and

from The Hospita1 for Sick Children. Consent was obtained for each specimen accordin:

to each institutionS policies.

RNA extraction

Approximately 50 mg of each specimen was crushed under Iiquid nitrogen and

total RNA extracted using 4mL of an extraction reagent (Trizol Reagent. Gibco BRL,

Grand Island, New York) according to the manufacturer's directions. The specimens in

Trizol were initially homogenized following crushing and centrifuged at 12000 X ; for

10 min. at d C to remove matrix. The concentration and purity of the RNA samples were

determined using spectroscopy and electrophoresis to visualize 18s and 28s rRNA bands.

Semi-quantitative RT-PCR

Two hundred nanograms of RNA template were reverse transcribed into cDNA

per 8 pl reaction using cDNA synthesis reagents (Perkin Elmer/Roche. Bnnchburg, New

Jersey). One microlitre of cDNA from the above RT reaction was used for each PCR

reaction. Primer sequences are listed in Table 3. Eich primer pair was designed to flank

at least one intron. in order to prevent and distinguish amplification of contaminating

genomic DNA. Genes of interest were matched to intemal control genes that had similar - PCR kinetics in order to allow near 1: 1 expression level compared to the gene of interest

in the linear range of amplification under optimized PCR conditions.

PCR reactioris were carried out in a DNA thermal cycler (Gene Arnp PCR System

9600. Perkin Elmer. Branchburg. New Jersey) at an anncaling temperature of 5 8 ' ~ . and

the linex range of amplification wris determined over a range of cycles for each primer

paii6. PCR pmducts were electrophoresed on 11% polyacrylamide gels and stained with

ethidium bromide. Positive and negritive controls for each primer pair. including a no-

tempIate water conuol cmied over from the RT step, were run on each gel. The gels

were photo-mphed and negritives used for densitometry (ImûgeQuant Software v3.0,

Molecular Dynamics, Sunnyvale, California). Relative levels of band intensity of the

oene of interest were compared to the interna1 control house keeping gene within the "

linear range of amplification, and normalized to a positive control sample on each gel.

Western blot

Protein was extracted from tissue samples by homogenization in lysis buffer (1%

SDS, 10 mM/L TrisIHCl, pH 7A), followed by 15 seconds in a boiling water bath and 5

minutes of centrifugation (13 000 X g)- Sample proteins were electrophoresed on an

SDS-polyacrylamide gel. tnnsferred to ri polyvinylidene difluoride membrane and

stained to verify successful tnnsfer. Western blot was performed using SEI '". a

monocIonal antibody to the amino terminus of rat Sonic hedgehog developed by TM.

Jessel. Columbia University and obtained from the Developmental Studies Hybridomri

Bank. University of Iowa (Western blot confirmed binding to human MH. and note that

SHH transcnpts were undetectable -data not shown), and the anti-c-Myc 9ELO

monoclonal was obtained from Santa Cruz Biotechnology, Inc. Hybridization was

canïed out ovemight at 4°C and detected using an ami-mouse IgG - horseradish

peroxidase secondary antibody and chemiluminescence.

Explant culture

Ptimary tissue specimens were cut into 1-3mm thick sections and maintained in

Dulbecco's modified eagle medium (DMEM) with high giucose. as described

elsewhereIq9 for gowth plate explants, for four days. Cultures were treated with either

0.1% FBS (agonist control). 10 @ml Shh-N (a gïft from Ontogeny Inc.), LOO-~M FTHrP

(1-34, Bachem), 1 0 0 ~ ~ cyclopamine (a $fi from William Gaffield - cornpared to

treatment with equal volume ethanol carrier), or 10 pg/rnl neutralizing monoclonal anti-

Shh-N antibody (5El) (compared to treatrnent with 10 pgml anti-rnouse IgG).

Statistical cornparisons were made using the Wilcoxon Signed Ranks Test. which

is a nonpararnetric test suitable for paired data. A parmetric test could not be used

because it could not be assurned thztt each of the human turnour specirnens were identical

prior to the start of the experiments.

Tritiated thymidine assays

Explant specimens were incubated with 10 pCi/ml tritiated ('H) thymidine for the

final 20 hours of culture, then digested with 0.58 Pronase (Boehringer Mannheim. Laval,

Québec) for 1 hour. followed by 0. LX% Collagenase A (Boehringer Mannheim. Laval,

Québec) for 8 hours. Cells were counted and resuspended in 0.5 ml 1% SDS/O.ûûSM

EDTA. Nucleic acids were precipitated with 1 ml ice cold 35% trichloroacetic acid

(TCA). and pipetted onto 2.4 cm &ES microfiber filters (Whatman GFICI over a sucrion

apparatus. The filters wüshed twice with 3 mL 5% TCA. air dried and placed in

scintillation vials with 5 ml scintilIation cocktail. and counts per minute avenged over 5

minutes.

Mutational analysis

Single stmnd conformation polymorphism (SSCP) analysis was used to screen for

mutations. as previously describedm. in the PTHRl gene. Genomic DNA was extracted

from specirnens using a kit (Qiagen) and used as ternplate for PCR amplification of

fi-agments containing an exon and its adjacent intronic boundaries. Primer sequences are

listed in Table 4. The "P-ATP incorponted PCR product was denatured and

electrophoresed on a native polyacrylarnide ge1 containing 10% glycerol. Band sh ih

were further analyzed using a sequencing kit (Thermo Sequenase Sequencing Kit,

Amersharn Life Science, Cleveland, Ohio).

PTHRl functiond analysis

Non-PCR based site-directed mutagenesis (Site-Directed Mutagenesis Kit,

Clontech) of the W PTHRI cDNA in the pcDNAl vector was used to generare the

RITOC and other published mutations of the PTHRI gene 187-189:189:19%.195 . The mutant

contructs were verified by autornated sequencing. COS7 cells were plated at ri density of

5 X 10~cells/ml and transfected the following day with the aid of a trmsfection reagent

(Fugene 6, Roche). Fony hours following transfection, cells were serum starved for two

hours and then treated with either PTHrP (1-34) in DMEM containing 0.1% FBS carrier

or carrier alone for 15 minutes for ciWP rissriys and for 40 minutes for IP; assays.

Cyclic A i assays were performed using an "enzymeimmunorissay" kit (RPN 225.

Amersharn), and iP3 assays were performed using a tritiated IP; assay system (TRK LOOO.

Amersharn). Transfection efficiency, which was typically over 40% as assessed by

cotransfection with green fluorescent protein cDNA. wtis controlled by cotransfection of

cells with fi galacrosidase cDNA and normalization of assay results to B gdactosidase

activity measured at 415h (fi Galactosidase Assay Kit. Stratasene). Immunofluorescent

subcellular localization was carried out using carboxy terminal c-Myc-tagged versions of

the wild type and variant PTHRI and anti-c-Myc 9EIO antibody (Santa Cruz

Biotechnology, Inc.) following ce11 membrane permeabilization with 0.2% Triton Xlûû

for 3 minutes at room temperature. (Tag$ng the carboxy terminus of F ï I R I has been

shown not to interfere with protein expression, ligand binding or sigdling ~ ~ a b i l i t ~ ' ~ ) .

Transgenic mice

The WTPTHRI or RITOC PTHRl constnicts were cloned by blunt 1i;ation into

the NotI sites of the mouse ColII promoter and enhancer - containing construct pKN185

(a giti from Y. Yarnada) and linearized by digestion with NdeI and HindIlI prior to

microinjection into the pmnucteus of fertilized ICR eggs. Genomic DNA was extracted

from triils with the aid of an extraction kit (Qiarnp, Qiagen) and screened by Southern

blot for integation of the uansgene. Heterozygous founder (Fig 3a) and FI seneration

(Fig. 3b) mice were used for phenotypic mdysis.

For Southern blot, approximately 5pg of DNA was digested with EcoRI and

BamHi. electrophoresed through a 0.7% agarose gel and depurinated in 0.W HCI for 10

min, folIowed by a distilled water wrtsh. Capiilary transfer of DNA to a nylon membrane

(2. Roche) was camed out ovemight using 0 .W NriOH. The nylon membrane was

wrished briefiy in 78 sodium chtoride/sodiurn citrate soiution (SSC). air dned and

crosslinked with ultraviolet light. Radioactive probe was senented by PCR amplification

of -20ng transgene remplace using 3.3phI "P-labelled d c r P (without any cold dCTP)

with the following prïmers: fonvard (b Globin F): CTCTGCTAACCATGTTCATG. and

reverse (SevS): GCAGGAAGATCTGïTCCTC: mpticon=223bp. The probe was

purified using a Sephadex G-30 coIumn (hershm). and quantified by scintillation

counter assessrnent of sequentid eluant samptes. The nylon membrane was

prehybridised with ülrmhyb ( h b i o n ) at 43' for Lh. rotath. in an oven, FolIowed by

addition of lob counts of radioactive probe per mi hybndisation solution ovemi$& at 42'.

The membrane was then washed twice in 2% SSC. 0.1% SDS. and twice in 0.1% SSC,

0.1% SDS at 42' followed by exposure to a phosphor screen ovemight Bands were

visualised using a phosphorimager apparatus. Southem Hot genotyping was often

complernented with PCR genotyping (Fig. 3c), using the following primers: forward

(Sev4): GGGTTCCTCAACGGCTCCT, and reverse (Sev7):

CAGGTCCTCTTCGCTAATC; amplicon= 133 bp.

Immunohistochemistry

Anti-c-Myc: Five micron 4% paraformaldehyde Fixed, panffin embedded

sections were dewaxed and hydrated, and epitope exposed using heat induced epitope

retneval. Sections were blocked for endogenous peroxidiise and biotin, incubated for 1

hour at room temperature with primary antibody (9ELO anti-c-Myc). detecred with a

mouse -on - mouse immunostaining kit (Vector Labs), and visualized using

diaminobenzidene.

Anti-COU: Depanffinized sections were incubated with 0.3% HG2 / MeOH rit

room temperature (RT) for 40min. to quench endogenous peroxidase. then rreated with

0.1% pepsin in O.SM acetic acid at 37" C for 2h and demasked wirh îmg/ml

hyaluronidase PH5 at 37" C for 30min. Samples were preblocked with 3% BSA, 1%

BM blocking solution ( IOmM maleic acid. lSmM NaCl in H20). 20mM MgCl2? 0.3%

Tween-30 in PBS overnight at 4°C. Samples were incubated with a 1:JO dilution of

anti-human recombinant COLX (Quanett Immunodiagnostikri Biotechnoiogie GMBH)

ovemight at a°C in a hurnid chamber, then washed with PBS and incubated with

biotinylated anti-rnouse IgG (1:300 dilution) for lh at RT. Srimples were incubated

with ABC solution for 4Smin. at RT for colour subtnction and colour was developed

using 0-lM Tris pH 7.5,0.5mg/mI DAB, 0.03% H20Î-

TRAP (tartrate resistant acid phosphatase$ staining

A TRAP staining kit (Sigma) was used on paraffin embedded sections to identify

chondroclasts.

TABLE 1: The fraction of a given specirnen type for which transcripts were detectable by

RT-PCR. SHH = Sonic Iiedgelrog, IHH = Indian hedgehog, PTCH = Patclied. PTHrP =

Paratliyoid Iiontione related prorein. PTHRl = Tvpe 1 PTHffTHrP receptor

SHH

Growth Plate 014

ArticuIar Cartilage 014

Cortical Bone 011 1

Osteochondroma (mature) 012

Chondroblzistoma 017

Enchondromri 014

Chondrosarcoma 0123

Osteosarcorna 0110

IHH PTCH

4 414

014 414

2/11 11/11

012 Y2

Y7 717

414 414

22/23 23/23

0110 lO/lO

TABLE 2: Average expression ratios. by semiquantitative RT-PCR. of the GU gene of

interest over the internai housekeeping gene Asparugine sptrlietuse (AS) in nomai and

lesional tissues.

GUl/ASa GLIXASa GLI3/ASc (GLIl + GLI2) l GU3

normal bone (n= 1 1 ) +

micular cartilage ( n 4 )

enchondromas ( n 4 ) +

chondrosarcornas (n= 1 1)

0.60

0.74

0.60

0.93

3.23

1.11

0.37

1 -5

TABLE 3: Forward (F) and Reverse (R) RT-PCR Primer Sequences

lHH=Indian hedgeliog, SHH=Sonic iwdgeI~og, PTCH=Patched, CLi=Glioblasronra

PTHrP=Pnrariyoid Irornione reiared proiein, PTHRI =Type 1 PTH/PTHrP recrptor.

COLX= Tvpe X collagen, AS=Asparag ine synrlierase, PBGD = Porphobiliriogen deuniinase.

p?M=Beta-2-niicrogloblilin

Primer pairs used were IHH/ASu, SHH/ASu, PTCMASb. GWI/ASa. GLIZ/I\Sa, GL13/ASc.

PTHrP/PBGD, PTHRUPBGD, and COWhM.

GENE SEQUENCE (5' - 3') AMPLICON GENBANK

PTCH

GLII

CL12

GU3

PTHrP

I H f i F CGGCGCCTCATGACCCA

R GACTTGACGGAGCAATGCA

SHH F GCGGACAGGCTGATGACT

R TGTGCCrrGGACTCGTAGT

F GGTCATGGTTACATGGACC

R TGCTGTCITGACTGTGCC

F GAGAAGCCACACAAGTGCA

R AGGGAGCTTACATACATACG

F GGAGTACGACACCCAGGA

R CCCTCGAACGTGCAmG

F CmCTGAGTCCTCACAGA

R GTTGITCCITCGGGCTGTT

F GTGACACACGGCAGCAGT

R CTCGWAlTGCAGAAACAG

ACCESSION

U S 5 17

L3851S

U59.164

X07384

AB007298

M57609

M57293

c o u

ASu

ASb

A Sc

PBGD

F TTCCTGCTGAGCTACGCG

R TGCGATCAGATGGTGAAGG

F CCATCTCCAGGAACTCCC

R C T r G C T a C C T m A c r G C

F ACATTGMGCACTCCCGCGAC

R AGAGTGGCAGCAACCAAGCT

F GAGGCITCTGAGGGAACTCT

R TTTCTGGTGGCAGAGACAAG

F ACATTGAAGCACTCCCGCGAC

R CCTGAG GTTGlTCïTCAC AG

F TCACTCCTTGAAGGACCTG

R GCTGTTGCTAGGATGATGG

F ACCCCC-4CTGAAAAAGATGA

R GGAGACAGCACTCAAAGTAG 267

45

TABLE 4: Fonvard (F) and Reverse (R) SSCP Primer Sequences for al1 PTERl exons

LF: CAGCCTGACGCAGCTCT

LR: GTAACCCGCAGCCTGTC

IF: GAAATTGGGATGTCTiTGGG

IR: TCCCTCACCTGACCTCATA

3F: GïTGTAGCACAGCTGACAG

3R:TCCAAGCCCATGCCAGCT

4F: CTT GACTCTCCCTTGGTAT

-IR: AGGTCCATGGGAGGTCAG

SF: AGTGCCTCGAGACCTCC

SR: GGAGTGAArnATcTGGTCA

6F: TCïT.4CCCTGATGCTCTCC

6R: GCATCAGACCTCGGCCA

TF: ACCCACGGTCATGTCGC

7R: GAGGCCAGTGGCGGCA

8F: TGACTTCCCGGAGGCAG

SR: CTCTCCCTGTCACCCAC

9F:GAATGACCïT'GTGGACAGC

9R: CTCACATGCTTCCTGGAAG

IOF: TCAACAGCTAATGTCAGACC

IOR: GCAGGCGCACATCCCA

L IF: GACAGAGCAGAGCCTATG

1 1 R: ATACATGGTACCAGGACTC

13F: GfiGTGGCG'ITGGCCCT

12R: GGTGTCTCGAATCAGGC

13F: CTTCCAGGTGCGCAGTG

13R: CTTCCAGGTGCGCAGTG

IdlF: GGAGACACACCTGACTG

14- 1 R: ITGGCATGGCCAGGCAG

14F:TACTGCCCAflGCCACCA

14-ZR: TGGTCCATCTGTCC ATCCA

CSA EC A

26 28 26 28 cycles

Ciethanol ~cyciopamine G ~ G ~ CSAI $= gH

I

CSA ECA 0 . l I F B S 1 FlTkP

Chondrosarcoma (CSA, n=lO) and enchondrorna (ECA, n=2) explant cultures. a, Downregulation of Type X collagen (0 compared to internd control p2- microglobulin (B2M) by semiquantitative RT-PCR foliowing ueatment with LO' 'M ETHrP (1-34) compared to mamient with carrier aione (pû.01, n= 12 - Wilcoxon Signed Ra& Test). b, hcreased incorporation of uitiated thymidine (3H-T) foiiowing treatrnent with 10 pg/d recombinant Shh-N protein (p-d.02, n=12 - Wilcoxon Signed Ranh Test). c, Decreased incorporation of 3H-T foUowhg treatment with 104M cyclopamine dissolved in e thanol (~=û.07, n=6 (2 enchondroma and 4 chondrosarcorna specimens) - Wilcoxon Signed Ranks Test). d, Semiquantitative RT-PCR assay showing IHH is downregulated compared to internai control Asparagine synthetase (AS) in growth plate (GP) explants (n=4), but not in chondrosarcoma (CSA) (nor in enchoadroma - data not shown) expianrs, following treatment for four days with lO-'M PTHrP ( 1-34).

Figure 2

a normal enchondromatosis

- -

A C G T A C G T

00.1% FBS + PTHrP

~ C D N A ~ WT RI SOC 1 0 0 M PTHrP

Figure 2 cont.

FER1 mutation and functional aaalysis. a, C to T change and resulting amino acid substitution. b, Cyciic AMP "enzymeùnmunoassay" (Amersham) in COS-7 ceiis cornparhg RLSOC variant with empty vector (pcDNAl), wild type (WT) FEIRI, other known loss of function (P132L) and activating mutants (H223R, T410P) and a polymorphisrn we identified (E546K)! in the presence of 10dM FiWP (1-34) compared to carrier done, c, IP, 3H-T assay (Amersham) in ES ceus which lack the native PWRl gene. d, Cyctic AMP accumulation in response to increasing P'i'HrP dose. Assays were performed three times in dupiicate with DNA from two separate bacteriai colonies. e, Dominant negative effect of MJOC PTHRl over WT PTHRl with regard to CAMP Ievel. f, Confocai irnmunofluorescent subceUular Localization of carboxy terminal c-Myc-tagged WT and RISOC PTHRI in COS-7 ceIis using 9E10 antibody foilowing cetlulac permeabilization.

Figure 3

a Founder

.. - --- - . . .. . . "-, .

Southem blot of founder (a) and F1 generation (b) mouse genomic DNA. c, PCR genotyping ofF1 generation mice. d, Ami-c-Myc immunohistochemistry reveaied expression of the c-Myc tagged RlSOC (and W- not shown) PTHRI transgene (brown colour) under the conml of Col11 regdatory elements in growth plate chondrocytes, but not in the surrounding penchondrium where CoUI is not expresse4 nor in a negative control growth plate (e).

Figure 4

Newbom mouse growth plates. Trichrome stained WT PTHRl (a) and R150C PTHR1 (b) rransgenic proximai humerai growth pIates at identical magnif~cation demonstrated a difference in the height of the hypemophic zone (between dark lines). Anti-COLX immunostaining of WT PTHR1 (c) and R150C PTHRl (d) transgenic distal femorai growth plates highlighted the relatively shon Liypemphic zone (dark brown coIour and arrows) in R 150C PTHRl mice.

Murine enchondromas were evident in R150C PTHRl (b-arrows, d), but not WT PTHRl (a), proximal humerii at 8.5 weeks of age as muItipIe metaphyseal cartilage tests. These rests persisted to at least 38 weeks of age (c,e), occurred in multiple boues including the dista1 feumur (20 weeks old - f), and consisted of islands of chondrocytes in a hyaline cartiiage matrix, similar to human enchondromas.

Figure 6

A paucity of hypenrophic chondrocytes in RISOC PTHRl growth plates (b) persisted relative to WT PTHRl growth plates (a) and may have allowed proliferating chondrocytes to have been Ieft behind in the metaphysis to form enchonciromas. Tartrate resistant acid phosphatase staining specificaiiy highlighted equal numbers of chondroclasts (red- bmwn colour) at the cartilage-bone interface in WT PTHRl (c) and RISOC ETHRl (d) boues, suggesting the metaphyseal cartiiage rests were not caussd by a resorption defect.

Gli2 and Glij in growth plate regulation and pathology

SUMMARY

The Hedgehog responsive transcription factors Gli2 and Gli3 have specific and

redundant functions dunng limb patterning'sl, but their roles in later growth plate

regulation have not been elucidated. Growth plate chondrocyte prolifention is regulated

10163 in large part by Indian hedgehog (Ihh) signalhg . Here 1 show how the likely thh

transcription factors Gli2 and Gli3 contribute to the regulation of chondrocyte

prolifention. as well as how these molecuIes modify regulation of differentiation by

PTHrP. Based on relative growth plate size and BrdU uptake of prolifentive zone

chondrocytes in long bone cultures ti-om mice Iacking Gli2 and Gli3 and from those

overexpressing G1iZ under the regulatory elernents of T y r II collagrn (Culll). 1 show

that chondrocyte proliferition is positiveIy reguiated by Gli3 and negatively regulated by

Gli3. PTHrP treatment of organ cultures inhibited prolifention in addition to having

delayed differentiation of chondrocytes. Both of these effects of PTHrP were dampened

by GliZ and au-mented by Gli3. Dysregulation of Hedgehog signalling in fetal bone

explants of mice which overexpress GiiZ or lack Gli3 is analagous to that caused by the

mutant PTWPTHrP receptor (PTHR 1) which causes enchondromatosis in humans

(Chapter 2). These mice developed chondrocyte neoplasia equivalent to the human

conditions enchondromacosis and synovial chondromatosis. respectively.

mTRODUCTION

Growth plate chondrocyte proliferation and differentiation are regulated in part by

[ndian hedgehog (Ihh) and Pmthyroid hormone related pmtein (PTHrP). Expression of

PTHrP. which results in delay of hypertrophie differentiation of chondrocytes, is

dependent upon lhhl". Ihh - driven chondrocyte prolifention, however. is l q e l y

independent of PTH~P'~'. Therefore, the mitognic function of Ihh is likely mediated

directly by the Hedgehog sipalling pathwriy in prolifenting chondrocytes. The GLi

family of transcriptional regulators likely mediates Ihh function in the growth plate.

The skeletal phenotype of mice lacking either of the Hedgehog transcription

Y0:Ifil factors GliZ or Gli3 is less severe than that of Ililt nul1 mice . while Glil single

hornozygous mutant mice are apparently normal20'. ~ l i ? mice drvelop short long

bones. while ~1i3"- mice display regional. mainly antenor ribnormalities such as

shonened or deleted tibias and preaxial polydactyly. Some of these phenotges are

exacerbated in double mutants. indicating that Gli2 and GIi3 have redundant as welI as

specific functions dunng limb patterning'sl. These functions may be attributable to the

patterns of expression of Gli2 and Gli3 and to the distinct biochemical propenies of the

different Gli proteins. Gli2 and Gli3 both have an N-terminal repressor domain and a C-

terminal activator domain. whereas Gli l has only an activator dornaini3. The activator

7;. 12.4 and repressor capabilities of Gli2 and Gli3 are context - dependentg5"-. . Whether

GliZ and G1i3 are positive or negative regulators of prolifention and differentiation in the

p w t h plate remains to be elucidated.

Ectopic tissue resembling _mwth plate cartilage is found in cenain benign

cartilage tumours. These include enchondroma~"~~ and synoviai chondrornat~s is~~~.

Such cartilage tumours mighr be caused by inappropriate expression or regdation of

-mwth plate signalling proteins. 1 recently shawed that Hedgehog sigalling is

ovetrictive in enchondromas relative to nomal mesenchymai tissues (Chapter 2).

Whether excessive Hedgehog signalling is sufficient CO cause cartilage tumours is

unclear. Here E show that Gli? acts as an activator and Gli3 as a repressorof [hh function

in the rnurine growth plare, Overexpression of Gli3 or a lltck of Gli3 led to eccopic

proliferation of cartilqe in mice. resernbling rhe human conditions enchondrornatosis

and synovial c hondromatosis, respective1 y.

RESULTS

Growth phte pmliferation is increased by Gii2 and diminished by Gli3

Relative srowth plate size and proiiferative zone bromodeoxyuridine (BrdU)

uptrikr: were cornpared becween mice with genetic alterations of GIi2 and Gli3. Ovenll

mwth pIate heizht at EL65 was most notably altered in bones from ~ l i ? mice. where - the distal femonl srowth plate was 17% shorter (1.28 +/- 0.8mm. n d ) than in WT mice

( 1.54 +1- O. 1 Smm. nd). AI1 long bones of GE" mice were previously reported to be

shorter than those of WT mice'''. and I measured a difference of 10% in fernonl lcngth

between these animals ( ~ l i Z ' - = 3.0 +/- 0,OZmrn. n=4 vs. WT = 3.35 +/- 0 . l h m . n d ) .

At E 16.5. there was no alteration of p w t h plate height in ~ l i 2 ~ " ' . ~li3-'-, ~ l i 3 ~ ' mice.

nor in urinsgenic mice overexpressing Gli2 under the regdatory elernenrs of mouse Dpe

II collagen (Colll). Offspnng of chree out of four founder mice expressin; the Colll-GliZ

tmssene dicl. however. deveIop growth plates that were longer than those of WT

littemates by four weeks of age. Proximal tibid p w t h plate height. for example. was

49% tdler in Culif-Gli2 mimals (380 cl- 1 0 ~ m , n=3 vs. 255 +/- 25, n=2 in WT mimals -

Fia. 1c.d). Both the prolifentive zone and the hypertrophic zone were -sater in size,

which was most clearly demonstrated by anti-COLX immunostaining to highlight the

hypertrophic zone (Fis. 1e.f). This growth plate height difference between ColII-Gli'

and WT mice diminished as the ~ o w t h plate thinned throughout adolescrnce, and wris

barely perceptable by eight weeks of age (Fig. Lg,h). Al! long bones were also taller in

these tmqenics. Tibias of ColII-Gl2 mice were 11% longer (14.4 +/ OLCrnm, n=3) than

those of WT littermates f 13 +/- 0.3 mm, n=3).

Lower extremity long bones of EL65 rnice were dissected free of soft tissues and

maintained in cul ture for seven days. BrdU was added to the media for the final 20 hours

of culture More the limbs were fixed. pnraffin embedded. sectioned and stained to detect

BrdU. The proportion of proliferative zone BrdU positive cells in the growth plates of

these rnice was decrerised by ri total Irick of GIiZ and increased in mice overexpressing

Gii2 compared to WT littermates. BrdU uptake was also increased by a total Iack of Gli3

(Fig. 2b. 4).

PTHrP stimulation decreases prolireration in addition to delaying hypertrophic

differentiation, and these responses are diminished by Gli2 and augmented by Gli3

Right and left Iimbs from a single animal were maintained in organ cultures

described above and underwent daily changes of media containing either 10% PTHrP

(1-34) or 0.1% bovine semm alburnin (BSA). Immunostaining of sections with anti-

COLX mtibody to highlight the hypertrophic zone of the growth plate reveded h t

KHrP treatment of cuitures resulted in diminution of this zone. as anticipated (Fïg. 2a.

3). Surprisingly, W P treatment of cultures also resulted in a small decrease in the

proponion of BrdU positive cells (Fig. '>b, 4).

The extent to which PïHrP treatment of organ cultures altered differentiation and

prolifention varied according to genotype. Growth plates from Gli2-'- mice were the

most sensitive to PTHrP with respect to hypertrophie differentiation delay and decrease

in BrdU uptake compared to growth plates from Gli2"- and ~ l i 2 ~ ' mice (Fig. 2.1, b, 3.4).

Consistent with these findings, these changes were smaller in CollI-Gii2 transgenic

growth plates compared to those of WT littermates (Fig, 2a. b, 3,4). Sensitivity to

PTHrP was also diminished in Gli3-"growth plates compared to growth plates from

~l i3~ ' - and Gli3"" mice (Fig. '>a, b, 3.4).

Collf-Gli2 mice develop enchondromatosis

Since tnnsgenic mice expressing ColII-Rl5OC PTHRl develop

enchondrornatosis. and since the growth plates from these mice behaved similar to those

of Colll-Gli2 mice in explant culture, 1 investigated whether enchondrornas arose in

Colll-GliZ transgenic mice. Histological sections beyond twelve weeks of age of long

bones were stained with safranin-O to highlight cartilage, Islands. or rem. of cartilaze

on the rnetaphyseal side of the growth plate were evident in three founder lines

expressing the ColIl-Gli3 transgene, but not in WT littermates (Fig. 5). The distal fernur

(Fig Sa, c). tibia, proximal humerus (Fig. 50 and distd radius were the most common

locations of the lesions, which persisted well into adulthood (Fig. 5f). These rests of

cartilage were similar to growth plate tissue (Fig. jb, d, f), and similar to those found in

the ColII-RIjOC PTHRl uansgenic mouse mode1 of enchondromatosis. In order ro

determine whether these cartilage rests resulted from a multifocd lack of resorption of

the growth plate rather than from dtered growth plate proliferation and differentiation,

sections were stained with acid phosphatase and challenged with tartrate to highlight

chondroclasts. Chondroclasts are responsible for resorption of the cdcified cartilage

matrix at the bone -cartilage jun~tion"'~. and were found in equnl nurnben in both

ColII-G1i2 and WT littemate bones (Fig. Sg, h).

GIi3 heterozygous mice develop accelerated synovial chondromatosis

Given the similuity in response to FTHrP of fetaI long bone cultures from mice

overexpressing Gii2 and those Lacking Gli3, I investigated whether Gli3"- rnice develop

skeIetaI pathoIogy. Although no enchondromas were discovered in ~li3"'- micc. the

males in particuIar developed cmilaginous neoplasia similx to the human condition

synovial chundromatosis in their knees. which is the mosr comrnon site in humans.

Ectopic caniIage containing chondrocytes cytologically simiIar to growth plate

chondrocytes was seen extending directly off the synoviaI Iining of the joint (Fig. 6c).

Binucielir Iacunae and occrisional mitotic figures indicated active proliferation within the

lesion (Fig. 6e). This ecropic cartilage was sometimes seprirated from the synovial lining

and present as Ioose bodies. and was apparently eroding into the articular surfaces of the

femur and tibia, simiIar to lace stage human synoviai chondromatosis. Infiammatory ceils

were not seen in or mund the joint Iining. Calcification within the ectopic cartilage was

seen as speckled opacities on radiographs mg. ba, b). Synovial chondrornatosis occurred

predominrintly in males. and was often biIaterri1. The occurrence of synoviai

chondromatosis increased with age in al1 genotypes exltmined (WT. G&-, ~ii3*'- and

G I ~ ~ ~ - / G I ~ S ' ~ , but was significantiy accelented in males lacking one Gli3 allele at 13

months of age (p=0.012 for ~ l i p - a n d p9.047 for GliT"'/G!i3+'-, Fisher's Exact Test)

(Fi;. 6f). By 21 months, the incidence of synovial chondromatosis in ~li.3'" mice was

diminished by the additionai lack of a single Gli2 allele (Fig. 60.

DISCUSSION

Gli2 mediates and Gli3 inhibits Ihh function in the growth plate

Ihh is largely responsible for growth plate chondrocyte prolifention and induction

of FTHrP expression, which in tum delays hypertrophie differentiation YO:YL:163 . In this

study. the absence of GliZ was associated with a decrease in prolifention and an increase

in hyperrrophic differentiation, while overexpression of Gli2 and a lack of Gli3 were

associated with the opposite results. Therefore GliZ is a positive mediator and GIi3 is a

negative regulator of Ihh function in the proliferative zone of the growth plate.

Evidence of functional antagonisrn between Gli2 and ETTHrP

It is known that, in addition to signalling the perichondriurn to induce PTHrP

expression. Ihh likely sigals proliferating chondrocytesYO. Therefore, a subset of Ihh and

PTHrP target cells are likely the same. The nature of the interaction between these

signals in the proliferative zone has not been elucidated. 1 found that chondrocyte

proliferation, which is positively regulated by Glil, was inhibited by exogenous PTHrP in

orgrin culture. This finding is in contmst to previous reports which showed that a lack of

PTH~P"~, as well as exogenous PTHrP in explant cultures'", did not alter the

prolifention of growth plate chondrocytes. My cultures were treated with a saturating

dose of ETHrP for a prolonged period of seven days, and the tesult may be due to at Ieast

two factors. Prolonged superphysiolo~c PTHrP treatrnent of cultures may have

decremd the number of hypettrophic chondrocytes, a source of the mitogenic ihh si@,

enough to alter proliferation and thereby indirectly antagonize ihh function. However,

Ihh is also expressed by a population of prehypenrophic chondrocytes, which is probabty

expanded by PTHrP treatment, and it is difficult to estimate the relative importance of the

sources of Ihh in inducing chondrocyte proliferation. An altemate possibility is that

PTHrP directly antagonizes Ihh function in cells of the proliferative zone. Consistent

with this possibility is the fact that protein kinase A (PKA), which is activated by PTHrP

signal tnnsd~ct ion '~~, is a conserved inhibitor of Hedgehog signalling'".

Antagonism by Glil on PTHrP function was also evident. The de-gree to which

PTHrP inhibited both proliferation and differentiation was inversely relsited to GliZ levei

and directly related to Gli3 level, which shows that Gli2 and Gli3 have distinct d e s as

mediators in the interaction between Ihh and PTHrP sigalling, although the effects of

Gli3 were weak in this regard. These findings sugest that Glil antagonizes PTHrP

function because the presence of a higher number of GliZ alleles was associated with a

diminished abiiity of PTHrP to inhibit proli feration and di fferentiation. An altemate

possibility is that endogenous PTHrP levels were raised with increasing numbers of Gli2

alleles. consequently delaying differentiation with greater effectiveness and thereby

minimizing the effect of exogenous PTHrP. A problem with this altemative possibility is

that endogenous PïHrP produced by the perichondrium may diffuse into the media

without reaching the prolifentive zone in a concentration sufficient to explain the resuIt-

Interestingly, others have shown that Hedgehog stimulation of growth plate chondrocytes

exens a pro-hypertrophie differentiation effect in vitro'". One can explain these findings

and my data by invoking a mode1 whereby Ihh - stimulated Gli3 antagonizes PKA -

stimulated delay of chondrocyte differentiation.

Excessive Hedgehog signalling is suficient to generate cartilaginous neoplasia

As previously noted, enchondromatosis, a human condition where multiple benisg

cartilage tumours form on the metaphyseal side of the growth plate as a result of

abnormal -mwth plate development, is caused in some cases by a mutant PTHR1

(Chapter 2). Other genetic alterations that cause similar biochemical derangement would

also be expected to cause benign cartilage tumours. 1 found that overexpression of Gli2

in the growth plate has effects on the regulation of proliferation and differentiation

simiiar to those caused by the enchondromatosis mutant PTHRI. and does in frict resuIt in

multiple enchondromas in mice. Mice lacking a single Gli3 allele did not develop

enchondromas, despite the functional approximation between ~1i3"- and Colll-Gii:!

sowth plates in culture. Rather. these mice developed a different condition of benign - chondrocyte neoplasia which was sirnilar to human synovial chondromatosis. The

difference in location and in the time of appeannce of this condition suggests that Giil

and Gli3 do not simply perform opposite functions. and reflects the regional imponance

of Gli3, especially around the perianicular perichondrium of the proximal tibia.

Consistent with this hypothesis. lack of a single Gli2 allele did not prevent the mer, 'lence

of synovial chondromatosis in ~1i3"' rnice, at les t until 13 months of age. In both

~l i3+" and ColIl-Gli? mice. an alteration of Hedgehog signalling resulted in ectopic

chondrocyte prolifention, which underlines the importance of tight regulation of growth

plate s igak.

METHODS

Embryonic limb organ culture

Hind limbs from embryonic day (E) 16.5 mice were dissected free of soft tissues,

leaving the perichondrium incact. These Iimbs were maintained in culture in 24 well

plates, one Iimb per weI1, for 3 days for prolifention assays and for 7 days for

differentiation assays. with daily changes of 3 0 pl media which just covered the

explants. Right and left limbs fmm each mouse were compared as treatment ( 1 0 " ~

PTHrP (1-34) in DMEM + 0.1 % bovine senim albumin (BSA)) or control (DMEM +

0.1% BSA) groups.

Sample variance is provided in Figure 2 in the form of error bars. Calculation of

Type 1 error (false positive) was not canied out because the low and varying sample

number would have rendered accunte assesment of the complimentary and equally

important Type II error (false negaiive) impossible.

COLX immunohistochemistry

Anti-COLX irnmunohistochemistry was cmied out 3s described in Chüpter 2.

BrdU uptake analysis

Explant cultures were treated with LOOpM bromodeoxyuridine (BrdU) for the

final 20h of culture. Limbs were then Fixed in 10% formalin, pmffin embedded and

sectioned. BrdU uptake was detected by anti-Brdu imrn;rnohistochemist (similar to

C O U above).

Generation of ColII-Gfi2 transgenic mire

Full length 5'Jag-tagsed murine Gli? cDNA was cloned by blunt Iisation into the

Noti sites of the murine ColII promoter and enhmcer - contiüning consmct

p ~ ~ 1 8 5 ' ~ L 9 7 and Iinearized by digestion with NdeI and HindüI prior to microinjection

into a pronucleus of Fertilized K R eggs. Genomic DNA was extncred from tails and

screened by Southern blot for inte,mtion of the transsene. Southern bIot was carried out

as described in Chapter 3, except that genomic DNA was digested with EcoR1. Primer

sequences for generation of the radioactive Southern bloc probe as well as for PCR

genotypin; were: forward (0 Globin F): CTCTGCTAACCATGTTCATG, and reverse

(mGli2R): CAGAGGACAGGCCMTTTC: amplicon=33jbp.

Radiology and histology

Mice were killed and radiopphed usin; a Faxitron X-ray machine usins JSV for

a 33 second exposure on Kodak X-OMAT film. Limbs were dissected free of soft rissucs

and fixed in 4% paraformaIdehyde. panffin embedded. scctioned and stained with

safranin-O CO highlight cartilage.

66 Figure 1

Anti-Bag immunohistochemistry reveded expression of the fïag-tagged Gii2 transgene (brown colour) under the conuol of ColII regdatory elements in p w t h pIate chondrocytes (b), but not in a negative conmt growth plate (a). P r o W tibia1 growth plates of WT (c,e,g) and transgenic COU-Gli2 (dfh) Iittermates show that overexpression of GI2 resuIted in a posmatal increase in total growth plate height (safranin-0 red), whicti was most noticeabIe at 4 weeks of age (c,d). Anti- COLX staining highlighted the fact that both the proliferative zone (flat, blue nudei, between da& lines) and the hypemophic zone (brown matnx) were Mer in ColII-Gli2 mice (f) compared to WT mice (e). The differeace in height between the p w t h plates virtuaily normalized by 8.5 weeks (g,h).

Giii+i+, Gli24- GGIP-1- Gii i4- Gli3-1- \NT Colli- Collt Cdli- G l i i + n=3 n=2 n 4 n=2 n=3 Giii VVT R150C

n=6 n=2 PTHRl PTHRl n S n=2

Growth plate chondrocyte differentiation and proliferation in short tenu expiant cultures of skeletonised lower extremities fiom E16.5 mice. a, Height of the hypertrophie zone, as measured by the number of ceil diametes counted longitudinaiiy within the anti-COLX - stained area, b, Proportion of proHerative zone chondrocytes that have taken up BrdU. ColII- craasgenic mice (the 4 groups on the far right in a and b) were assessed by comparkon to separate controls since they are of a genetic backgouad distinct h m the remahder, and were investigated at a Iater time.

Figure 4

0.1% BSA PTHrP

Anti-BrdU - stained sections of distal femorai gowth plates following expiant culture of El6.S skeIetonized limbs (surnmarized in Fig- 2b). Cornpared to WT mice (a), fewer proHerative zone chondrocytes (benveen dark lines) have taken up Brriü (brown stain) in mice Iacking Gli2 (c). The opposite was true of mice which lacked Gli3 (e) or which overexpressed Gli2 (g). PTHrP decreased the number of BrdU positive ceiis in most samples. Sensitivity to PTHrP in this regard was greatest in the absence of Gli2 (d vs. c) and less in the absence of G1i3 (f vs. e) or in the presence of excess Gli2 (h vs. g).

Figure 5 70

Enchondromas in ColII-G12 mice. Ectopic rests of c d a g e (arrows) are evident in CdIZ-GliZ, shown here in the distai femurs at 13 weeks (a, b) and 16 weeks (c, d) and in the proximal humerus at 43 weeks (0, but not wild type (e) mice. The rests are similar to growth plate tissue and to human enchondromas (b, d, f). TRAP stainiag highlights cbondroclasts (red-brown, arrows) in equal numbers at the chondrodsseous junçtion in wild type (g) and ColII-Gli2 mice (h).

Figure 6

- no. mo.

. --

Synovial chondrornarosis in G l W mice. a, Calcification in the knee of a Clip-, but not WT (b), mouse at 13 months. c, Massive ectopic cartilage (arrows) within the knee joint of a 13 mon& old (313'- moue is contiguous primarily with the synovial lining, but aisu witti the articdar surfaces of the femur (F) and the tibia (T) where it is associated with destructive asteoarthritis. d. A knee joint h m a 23 month old wild type mouse demonstrates no ectopic cartilage. e, The ectopic ceUs resemble gmwth plate chondrocytes. f, The incidence of synovial chondmmatosis increases with aoe. and is hiehest in Gli.3 heceromeotes at ail time wints.

Patched is a CO-receptor with Smoothened and PTHR1, two seven-pass

transmembrane proteins with opposite effects on CAMP accumulation

Note: The data contained in this chapter are preliminary. The mode1 proposed is

intended as a template for future experiments.

SrnrnIARY

The Indian hedgehog (Ihh) and Parathyroid hormone related protein (PTHrP)

signalling püthways are thought to be CO-regulated by means of a paracrine signalling

relriy". It has k e n shown chat the Ihh receptor Patchedl (Ptch 1). in addition to being

expressed in the perichondriurn, is coexpressed with the Type 1 PTWPTHrP receptor

(Pthrl) in the proliferative zone of the growth plateYu. Here I investigate the potential cell

autonomous relationship between the Hedgehog and PTHrP sipalling pathways. 1 show

that coexpression of PTCHI with W. but not with the R150C enchondromatosis mutant,

PTHR 1 augments accumul;ition of intraceIIular CAMP and of IP,. R 150C PTHR 1

constitutively activates Hedgehog sigalling in vitro, in a manner that is likely dependent

upon suppression of PKA. PTCH1 foms a complex with WT. but not R150C. PTHRI.

Therefore. augmented accumulation of c h i and IP3 is dependent upon the abiiity of

PTHRl to form a complex with PTCHI. In the proliferative zone. where PTCH1 and

PTHR1 are both expressed, the interaction of these signalling pathways rnay instruct the

appropriate proliferation and differentiation readout of a given chondrocyte.

INTRODUCTION

The curent rnodel of interaction between the MH and PTHrP sigalling

pathways in the growth plate involves a paracnne feedback loop. MH, expressed by

prehypertrophic and hypertrophie chondrocytes, induces expression of PTHrP by the

79:31:163 penarticular perichondriurn, which in tum downregulates IHH The PTHrP

receptor. Pthrl, is expressed by chondrocytes within the prolifentive zone of the growth

p~ate156:160161 . The likely Ihh transmernbrane receptor Ptch 1. in addition to k ing

expressed by the perichondnum7g, has k e n shown to be expressed also by chondrocytes

in the prolifentive zone in miceY0. Therefore. a subset of Ihh and PTHrP target celis are

likely the same.

Since Prchl and Pthrl rire both expressed in the proliferative zone of the gowth

plate. it is conceivable that they moduiate prolifeeration and differentiation in a ce11

autonornous rnanner. Transmembrane proteins. including G protein - coupled seven-pass

transmembrane receptors, often fom heterodimers thrit rnodify the sipalling ability of

203-205 either protein . PtchI is a twelve p a s tmsrnembrane receptor which can form a

cornplex. at least when overexpressed. with Smoothened (Srno). a seven pass

tnnsmembnne protein115. Hedgehog ligand binding to Ptch 1 relieves the othenvise

constitutive inhibition of Ptchl on Smo. which can then activate downstream sigalling

eVentSIS:l'3 - It was shown indirectly in a frog melanophore assay that SM0 likely inhibits

CAMP synthesis. and therefore rnay activate GG"'. PTHRl is also a seven pass

tnnsrnernbnne protein that couples with heterotnrneric G proteins'". Unlike SMO.

ETHRI rnosr strongly activates GG~, and G%'". 1 show here that coexpression of Ptclil

and WT PTHRI au-ments accumulation of CAMP and of iP3. but Iimits Hh signai

transduction, These modifications of sigal transduction intensity are dependent on the

ability of m l to forrn a complex with PTCHI, since coexpression of PTCHI with the

Rl5OC mutant ETHRL. which does not complex with PTCHl, is not associated with

altered signal tnnsduction.

RESULTS

Coexpression of PTCH1 with PTHR1 augments CAMP accumulation and limits

Hedgehog signal transduction

In order to test the functional consequences of PTCHl-PTHR t coexpression. I

performed CAMP. IP3 and Hedgehoz signalling rissays. Consistent with prior assay

results. CAMP baçeline level was suppressed (Fis. 13) and IP3 riccumulation in response

to PTHrP was nbolished by the R150C PTHR L mutant (Fig. Lb). Cyclic AMP and IP;

sccumulation were both au-mented in s dose - dependent frishion when increasing

amounts of PTCHl cDNA was cotransfected with WT PTHRI (Fig, la. b).

Cotmnsfection of PTCHI with R130C PTHRI. however. was not associated with

augmented accumdation of cAiMP nor of iP3 (Fig. la. b). Hedgehog signal transduction

was rneasured in HeLa cells by cotmnsfection of a Hedgehog - responsive Gli2-

Lirc,@ue reporter constmct. Luciferase activity was increased in cells tnnskcred with

R I j K PTHRl compared to cells transfected with WT P'IHRI, when PTCHI and SM0

were also mnsfected (Fig lc).

FTCH1 and FïHRi form a complex

Given chat PTHRL is a seven-pass uansmembme protein simïlar to SMO. a

prorein known to bind PTCHl, and that E'TKRI is coexpressed with PTCHI in the

proMerative zone of the growth plate, it is possible that PTHRl and PTCHl form a

cornplex in that zone, To test for this possibility, c-Myc - tagged PTHRl and

haernophilus influenza henraglritinin (HA) - tagged PTCHl cDNA constructs were

cotransfected into COS-7 cells, and irnrnunoprecipitation performed with anti-c-Myc

monoclonal antibody two days following transfection. Westem blot analysis of the

irnrnunoprecipitate was performed using anti-HA rnonoclonril antibody. An

approximatel y 18OkD band was consistent1 y seen in lanes representing PTHRl -PTCHI

cotransfected celIs. but not in singly transfected controls (Fig. 24. This L80kD band was

also seen when protein frorn PTCHl - transfected cells was irnmunoprecipitated using

anti-PTCHI antibody and probed on Westem blot with anti-HA antibody (Fig. 2a).

Reverse order coirnmunoprecipitation using anti-HA rnonoclonal antibody followed by

Westem blot with anti-c-Myc polyclonal antibody was weakly positive in PTHRI-

PTCHI cotransfected cells (Fig. 2b).

Ernbryonic stem (ES) cells lacking both copies of the native Prhrl sene'" were

used to test whether endogenous Ptch L could be coirnmunoprecipitated with exogenous

PTHRL. Immunoprecipitation perfotmed using anti-c-Myc monoclonal üntibody

followed by Western blot using anti-Ptchl polyclonal antibody agiriin revealed a unique

-180kD band in lanes from PTHRl transfected cells (Fig. 2c).

Western blot band intensity of coimmunoprecipitates from either ceIl type was not

affected by treatment of cultures with L O ~ M PTHrP (1-34) for 15 minutes prior to protein

harvest (Fig. 2a, c). Band intensity was diminished, however. when COS-7 cells were

ueated with 5p@nl Shh-N protein for 15 minutes prior to harvest (Fig. 2a).

Coimmunoprecipitation did not occur in either ce11 type when RIjOC PTHRI was

transfected in place of WT PTHRI (Fis. 2a, c).

DISCUSSION

The development of several organs, including skin, breast and growth plate, is

706-1w. The dependent on both the Hedgehog and PTHrP sigalling p a t h ~ a ~ s ' ~ ~ . -

interaction between these signalling pathways has only been investigared in the growth

plate. Ihh upregulates PTHrP expression in the periarticular perichondrium, probribly by

an indirect route since PTCHI is not expressed in that resion of the penchod~urnYO.

PTHrP in turn downregulates Ihli, likely by both direct and indirect means. Delay of

chondrocyte differentiation by KHrP decreases the number of hypertrophie cells, and

thereby may indirectly downregulate 1hlz7! A direct mechanism is implied by the finding

that inhibition of Ihh expression by PTHrP in vitro dws not require protein synthesi~"~.

The potential intracellular interaction between the Hzdgehog and PTHrP sigalling

pathways in tissues where PTCHl and PTHRI are both expressed. such as the

proliferative zone of the growth plate. is addressed here.

PTCHI-PïHRl association is required for effective accumulation of CAMP and IP3

The signalling assays demonstnted that PTHRl signal transduction. via the

second messengers cAiW and IPj, was positively influenced by cotmnsfection of cells

with PTCHl. This induction of c A i i and P3 accumulation was associated with the

ability of ETCH1 to complex with PTHRL, because RLSOC PTHRL, which did not

coimmunoprecipitate with PTCHI, was not able to au-ornent induction of these second

messengers. Protein kinase A (PKA), which is activated by CAMP. is an inhibitor of

Hedgehog signal transd~ction'~'. Therefore, the ability of PTCH1 to increase CAMP

accumulation in association with PTHRL is a role consistent with its inhibitory function

with regard to Hedgehog signalling. When cotnnsfected with PTCHl and SMO, RljOC

PTHRl constitutiveIy stimulated Gli2-Litc'rast, activity. The R 150C FTHR 1 mutant

therefore tips the balance of signais in favour of Hedgehog siyalling, likely by limiting

PKA activation. A model of intrricellular interaction between the Hedgehog and PTHrP

signalling pathways is found in Figure 3. Promotion of differentiation by Gli7 was

demonstrited in the previous chapter.

Cell autonomous model of Hedgehog-PTHrP interaction

An alternate mechanism to that proposed by earlier authors of IHH-FTHrP

signalling in the growth plate is possible with this model, The model in Figure 3 predicts

that the appropriate proliferation and differentiation readout of a given chondrocyte will

depend on its location in the growth plate, As a chondrocyte progresses through the

arowth plate. it will receive changing amounts of the [HH and PTHrP ligands. whose C

signal transduction pathways will compte intracellularly. This competition wilI

determine the relative strength of the proliferation and differentiation - delay instmctions.

In the proximal prolikrative zone. the concentntion of PTHrP will be high (since PTfIrP

is expressed by the perimicular penchondrium) and that of MH will be Iow (since IXH

is expressed by the prehypertrophic and hypertrophie zones), therefore differentiation

will be strongl y delayed. In the distal proliferative zone, ETHrP concentration will be

low and that of iHH will be hi$, dlowing the chondrocytes to become hypertrophie. Of

course, this is an overly simplistic model, and does not take into account the effect of the

feedback loop between iHH and EWhP at the transcriptionai level.

The RlSOC PTHRl enchondromatosis mutation alters the normal balance of

signals by preventing association of PTHRl with PTCHl. The mutant abolishes

accumulation of IP; and au-ments Hedghehog signal transduction in a manner that is

iikely ceIl autonomous and dependent upon dominant suppression of PK.4. Together.

these effects cause excessive chondrocyte differentiation delay and proliferation, which

are conceivably necessary for the genesis of enchonciromas.

METHODS

Cyclic AiMP, IP3 and Hedgehog signaliing assays

COS7 cells were plrited at a density of 5 X [O" cells/ml in DMEM + 10% fetd

bovine serum (FBS) and transfected the following day with the aid of a transfection

reagent (Fugene 6. Roche). Cells were transfected with some combination of Lpg erich of

W ï PTHRI. RI5OC PTHRI. PTCHI. or SMO cDNA. plus OApg ~galacrosidusr plus

pcDNA3 to a total of 4Apg. One day foIIowing transfection. cells from each 60mm plate

were split to three wells of ri six weIl plate. Two days following tnnsiection. cells were

serum starved for two hours and then treated with either PTHrP (1-34) in DMEM

containing O. 1% FBS carrier. 5pg Shh-N in carrier or carrier alone for 15 minutes for

CAMP assays and for 40 minutes for iP; assays. Ice cold 65% ethanol was used to lyse

COS-7 cells, and cAiW assays were performed using an "enzyrneimmunoassay" kit

(RPN 225, Amersham). For IP; assays. the same transfection protocol was used in

Pthr I ~ - ES cells plated on O, 1% gelatin in DMEM containing 15% FBS and leukemic

inhibitory factor (LE). Lysates obtained by incubation of cells with ice cold 75%

trichloroacetic acid were subjected to a tritiated [P3 assay (TRK 1000, Amersham).

Transfection efficiency was typicaily over 40% in COS-7 cells and 10%-20% in ES cells

as msessed by cotmnsfection with green fl uocescent protein cDNA in a separate set of

experiments. For Hedgehog signalhg assays, O.6pg of a Hedgehog - responsive G E -

Liicfrruse reporter sene was cotransfected into HeLa cells, and luciferase activity was

determined from lysates. Al1 assays were performed at Ieast twice in duplicate, and

resuits were controiled for transfection efficiency by assessrnent of P-Galnctosidase

activity using a colourimeuic assay (Stratagene) two days following cotransfection of

0 . 4 ~ ~ ~Gnicictosiduse cDNA.

Coimmunoprecipitation and Western blot

COS-7 cellç or ~rlzrl-" ES cells were placed as above. this time in 1OOrnm wells,

and transfected with equal amounts of each cDNA of interest to a total of 6.Oyg. PTHRl

constructs were tagged with c-Mye. and PTCH w u tagged with HA. The ES cells were

differentinted to early ernbryoid bodies (in which there is active Hedgehog signalling

'""'} by removing LIF from the medium and initiating suspension culture. Two days

following transfection. cells were trecited as above and lysed with lm1 extraction buffer

(50mM Tris. pH7.5. 1 SOmiM NaCI. t mM EDTA. 0.58 M'-LU) containing protease

inhibitor) on ice. Lysates were incubated with 2pg/ml of immunoprecipitation antibody

(mi-HA monoclonal. anti-c-Myc monoclonal or anti-PTCH polyc[onal antibody (Santa

Cruz)) at 4°C overnight on ti rotating apparatus. Samples were then incubaced with 30p1

Protein G PLusProtein A Agarose Suspension (Oncogene) for one hour at 4OC on a

rotatins apparatus to bind proteins complexed with the antibody. The agarose suspension

was washed three Urnes with extraction buffer (containing only 0.1% NP-LCO), and the

bound proteins were elured with simple buffer containing reducing PmercaptoethanoI in

a boiling water bath. Sarnples were nin on ,-dient polyacrylamide gels and transferred

to nitrocellulose for Western Mot. Blocking wûs canied out with 5% milk at 4"'

overnight on a rocking apparntus. Blots were incubnted with primary antibodies diluted

LI500 in 3% BSAITBS-T at 4*C ovemight on a rocking apparatus. Secondary IgG

antibodies (Jackson Immuno) included anti-mouse (diluted 115000). anti-rabbit ( 11 10

000) or anti-goat (1/1000), and detection was canied out using a cherniluminescent

horseradish peroxidase system (ECL).

Figure 1

a

O 0.1 O h FBS

PMrP

pcDNA3 W RlSOC PTCH SM0 PTCH+ PTCH+ PTCH+ PTCH+ PTHRl PTHRt SM0 SMO+ ShfOt SMO+

WT R l j K WT PTHR t PTHR 1 PTHR 1 +

R150C PTHRl

0 0.1% FBS I Shh-N

a, Intracellular CAMP and b, EP, accumulation was augmented in a dose - dependent manner when increasing amounts of PTCHI cDNA were cotransfected with WT, but not RijUC, PTHRi. c, RlSOC P T . 1 constitutively activated a Hedgehog-responsive Gli2-Luciferase reporter in COS-7 cells when cotransfected with Hedgehog receptor-complex proteins PTCHI and SMO.

Figure 3

a COS-7 cells Western ctHA

COS-7 cells

WT RLSOC PTHRI PTHR L

Western K - M y c

C

Pthrl" ES celis IP ac-M~C, Western PPTCH 1

WT PTHR 1

R 1SOC PTHR 1

a. Immunoprecitation from COS-7 celi lysate of HA-tagged PTCHl with anti-PTCHI polycIona1 antibody was detected on Western bfot using anti-HA monoclonai antibody as an -180kD band. HA-tagged ETCHL coimmunoprecipitated with c-iMyc - tagged WT FTHRI using anti-c-Myc monoclonai antibody. and was detected on Western blot probed with rinti-HA antibody as a band of the same size ris that obtained specificaIly for PTCHL. Treatment of COS-7 cclls with 10dM ETHrP did not alter coimmunoprecipitation of WT Fï?R 1 with PTCH 1 compared with treatment with O. L% FBS cimier alone. Treatment of cultures with S@ml Shh-N. however, diminished WT PTHRl - ETCH1 complex formation. Coimmunoprecipitation did not occur between R 150C PTHR L and PTCH 1. b, Reverser order immunoprecipitation of c-Myc - tqged WT PTHRl with HA- - tagged PTCH1 using mi-HA antibody followed by Western blot with anti-c- Myc antibody revealed a faint band the srime size as a c-Myc - tagged WT PTHRl l ysis - specific band. c, Endogenous PTCHl was coimmunoprecipitared with exogenous WT, but not R150C. PTHRl in embryoid bodies derived from Pthr14 ES cells.

Figure 3

IHH

4 I

Gli2 1- PKA d

proliferation differentiation

Mode1 of cell-autonomous Hedgehog - RHrP signalling, The intempted arrows are novei propositions based on the data in chapters three (arrows fmm GU) and four (membrane mow).

Conclusions and Future Directions

The data were discussed in detail in the proceeding chapters. 1 had asked whether

cartilaginous neoplasia is due, at l e s t in part, to dysregulation of the IHH-PTHrP

pathway. 1 found that dysregulation of that pathway is common in enchondromas. and

that it can be causal. The foIlowing section is intended to integrate the relevance of al1

the findings.

1) Cenes that regulate skeletal development are eapressed in skeletal tumours.

In both cardage and osteoid - forming tumours (Appendix), genes that regulate

the differentiation of normal celis that produce the same matrix are expressed. This

expression, however. is not universai, nor is it necessrtrily predictive with regard to

tumour phenotype, tn osteoid - foming tumours. the level of expression of markers of

normal osteoblast differentiation did not necessarily predict the neoplastic cellular

phenotype (Appendix). Xlthough expression of the IHHIPTHrP pathway members

studied was most complete and consistent in tumours that are microscopically most

similar to nomal cartilage. namely enchondromas and chondrosarcornas. a distinction

between these two tumour types was not evident despite the difference in invasive and

metastatic potential between them. Function of the JHH and PTHrP pathways in these

two tumour types was also simikir. Given the frequent and v;lried cytogenetic changes

described in chondrosarcorna. it is likely that other pathways operate independent of MH

and PWrP sigalIin_o to give chondmsarcoma its maligant propenies.

2) Dysregulation of developmentd signailing pathways can initiate skeletal

neoplasia.

I've shown that a developmentally important protein may initiate neoplasia if it

operates out of context with its CO-regulators. Obviously, an imbalance in the function of

an entire pathway, and not of any one particular signalling member, is the key to

pathogenesis. This concept is supported by the finding that overexpression of Gli2,

which h a bbiochemical and functiona1 similarities with the effects of the

enchondromatosis Rl5OC PTHRl mutant. was demonstrated in human cartilage tumours.

and caused the same pathological condition as RISOC PTHRl in mice.

3) Knowledge of how signals are altered in pathological processes can contribute to

the understanding of normal development.

Studying Hedgehog signalling in cartilage tumour explants led to a prediction

regarding normal growth plate function. It was felt from early studies in the chick that

Ihh was responsible for both chondrocyte prolifention and differentiation delayi9. In the

rumour explant cultures described in Chapter Two. under circurnstances in which the

feedbrick loop between MH and PMrP was not readily demonstnble. it was clear rhat

Hedgehog stimulation increased prolifention. but did not affect differentiation. in later

murine studies that isolated the role ihh frorn thnt of PTHrP. it was shown that indeed Ehh

is exclusively responsible for inducing chondrocyte proliferation. and that its effects on

SiF.163 differentiation are secondary to the induction of E"THrP expression .

Further insight into the regulation of chondrocyte behaviour was made possible by

investigation of the mechanisrn of pathogenesis initiated by the R150C PTHR1 mutant.

Dernonstntion of how PTCHl and PTHRl rnight cooperatively regdate proliferative

zone chondrocyte proliferation and differentiation was aided by the avaiIabiIity of a

ETHEU variant, namely the Rl5OC mutant, that did not associate with PTCHl.

Understanding pathologie mechanisms, therefore, can faditate understanding normai

function.

FUTURE DIRECTIONS

A number of experirnents can be propsed to funher understand skeletnl

pathology and developrnent by building on the data presented in the proceeding chapters.

Even the partial understanding of cartilage tumour pathogenesis gleaned by the

proceeding chapters may be useful in devising therapy. Delivery of pharmacologîc

agents that block Hedgehog signalling (cyclopamine. tripann~l'~'") andor PTHrP

sipalling (PTH (7-34)'14), might be a practical approrich to iittempting reduction in

cartilage tumour proliferation and to overcoming blocks to completion of differentiation.

In the skeletally mature person with enchondromatosis for example. systemic ueatment

would ideally cause differentiation CO noma1 bone of al1 enchondromas. and therefore

prevent the apperuance of chondrosrircoma. In a skeletally immature person. systemic

delivery of agents that intertère with chondrocyte behaviour might stunt growth, therefore

local delivery of these asents to sites where deformity from an enchondroma is likely

might be more appropriate.

The anima1 models of enchondromrttosis and of synovial chondromatosis cm be

of use in testing potentially therapeuric agents. The endpoint of such tests might be more

definitive with animais that exhibit a more severe phenotype than the ones described in

Chapters Two and Three. Crossing Colif-RISOC PTHRI mice with ColII-Giî2 mice, or

with mice that lack known tumoursuppressors such as Ptclrl, might result in a more

severe phenotype that would aid the assessment of thenpeutic triais. Another way of

testing the therapeutic value of agents such as cyclopamine is to treat mice implanted

with human cartilage turnour xenopfts.

in addition to the R150C PTHRI variant, it's likely that there are other genetic

altenitions that cause enchondromatosis. [t is possible that these other altentions disrupt

IHH-PTHrP signalIing in a manner similar to RIjOC PTHRl- identification of these

mutations would be useful in more precisely defining the pathogenesis of cmiiage

tumours by establishing genotype-phenotype correlations, in devising therapy, and

possibly in better understanding normal growth plate regulation.

Synovial chondromatosis in ~li3'" rnice is the only naturally occumn, O murine

crtrtilaginous neoplasia i'm awxe of. It may be worthwhile determining whether there is

loss of heterozygosity of Gli3 in those Iesions, and wherher the same phenornenon takes

place in human synovial chondromritosis.

The possibility chat PTCHI forms a cornplex with PTHRl that alters intricellular

sisgalling has profound implications with regard to the regulation of development of the

early embryo as well as of a number of tissues. The evidence presented in Chapter Four

is preliminuy, and requires demosmtion of PTCH1-PTHRl association at endognous

levels of expression of both proteins. CurrentIy, a good FïHRl antibody is not availûbie

to carry this out. There rire other merhods of determining whether a physical interaction

between the two proteins exists. Scable expression to near physiologie levels of

fluorophore conjugated PTCHl and PïHRI can be followed by analysis by fluourescent

resonance enersy m s f e r (FRET) analysis. Excitation of the fluorophore with the higher

wavelength by transfer of energy from the fiuomphore with the Iower, but overlapping

wavelength would imply the proteins are withing 100 Angstroms of one another, and are

therefore bound to each other. A less definitive altemative is to express GST-fusion

proteins, purify GST-bound proteins. and identiiy them by Western blot. The ar, =unent

that effective accumulation OCCAMP requires association of PTHRl with PTCHL might

be strengthened by demonstration of a p a t e r affinity between PTCHL and activating

PTHRl mutants. narnely H23R andT41OP. The physiologie significance of PTCHL-

PTHR1 association was partially assessed at ri biochemical level in Chapter Four, but

needs to be determined in organs in vivo. The best way of doing this is to generate rnice

in which PTCHI and PTKRL do not form a complex and compare the development of

various organs with those of mice in whicti PTCHL does associate with PTHRI. To

generate such mice, the RL5OC mutation can be knoctd into wild type ES cells,

assuming of course that this mutation truly disrupts Pthrl's ability to complex with

PTCHI. Altematively, ES cells that Iack both copies of endognous Prhrl can be

electroporated with either RljOC PTHRI or WT PTHRI, and aggregated with wild type

morulas. Another way of ssessing the functional interaction of PTHRl with PTCH is to

express PTHRI in Drosophiln, which don't have endogenous PTHRI. The mode1 of

Chapter 4 (Fig. 3) predicts that Hedgehog sigalling will be attenuated in areas where

PTHRl is activated. What's more, it cm be determined whether there is cell-autonomous

interaction in cells expressing the PTHRI transgene that alters subcellular localisation of

smo and ptc proteins.

The concept that normai developmental signais may be dysregulated to cause

neoplasia can be funher exploited to investitate the pathogenesis of other skeletid

tumours. For exarnple, given the expression of neural markers in Ewing's sarcoma and

peripheral neuroectodermal tumo~rs"~, and the importance of Notch signalling in latenl

inhibition of identical differentiation of neuronsX6, a lack of Notch signalling that allows

expansion of a proliferating, stem ceIl - like population of neuroectodermal cells may be

important in the genesis of those turnours. Evidence of the importance of Notch

signalling in Ewing's sarcoma is alrertdy evident, since the EWS-FLI fusion transcription

factor found in Ewing's sarcoma has k e n shown to dysregulate Manie fnnge, a Notch

signal modifie?. Osteosarcorna consists of a population of proliferating mesenchymal

cells". which are possibly analagous to cells in embryonic pyaxial and IatenI plate

mesoderm and in the limb bud pro-pss zone which are capable of producing clirtilage

and bone under the influence of certain signais. Proliferation of these embryonic cells is

dependent, in part, upon Fibroblast growth factor and Shh si_gnals ""'" . It might be

worthwhile investigating whether sipais that regulate development of embryonic

mesoderm are reactivated in osteosarcomn. Giant cell tumour, sometirnes referred to as

osteoclast~ma'~, might be caused by ovenctivation of signalling pathways that induce

monocytes to form osteoclasts. These include the P ' ' P t 2 0 and ~steoprote~enn"'

pathways. A somatic activating mutation of PTHRL, for example. could conceivably

give nse to giant cell tumour. Vascular tumours, such as hemangiomk

perihemangiocytoma, angiosarcorna, aneurysmal bone cyst and telansjectatic

osteosarcorna might be caused in part by an excess of sipals that regulate angiogenesis,

such as Vascular endotheliai ~ o w t h factor-A, which requires tight regulation of

expression Ievel for normal functionx- Chordoma, a turnour felt to be derived frorn

remnant notochordal tissue", might be maintained by signais elabonted by the

embryonic notochord. Given the partial chondroid differentiation of this tumour,

notochordal signals that induce panxial mesoderrn to differentiate into cartilaginous

a 3 7 4 sclerotome, namely SHH and PAXlPAX9 . may be active in chordoma.

Dysregulation of developmental sipals may be a relatively common mechünism of

pathogenesis of skeletal tumours, especially those that anse in children.

APPENDIX

Expression of osteocalcin and its transcriptional regulators CBFAl and MSX2 in

osteoid forming tumours

Kopyan, S.*. Gokgoz, N., Bell, R.S., Andnilis, I.L., Alman. B.A.. Wunder, I.S. Jo~ïntul

of Onhopaedic Research 17,633-638 (1999).

*Perfonned al1 procedures other than SSCP and sequencing.

SUhiMARY

Osteosarcorna, fibrous dysplrtsia and myositis ossificans contain osteoid producing

cells that are not rnorphologïcally typical osteoblasts. Nevertheless, these pathologie

cells may share differentiation steps with osteoblasts at the molecular level.

Osteocalcin (OC), a bone - specific extracellular matrix protein, is a marker of

mature osteoblasts. OC is upregulated by the transcription Factor CBFAL, which is

responsible For commitment to the osteoblastic Iineage, and downregulated by MSXî, a

Homeobox containing transcription factor expressed dunng the early proliferative phase

of osteoblast differentiation.

Semi-quantitative RT-PCR anaiysis was used to compare expression Ievels of OC,

CBFAl and MSX2 in 34 osteosarcorna. 5 fibrous dysplasia and 5 myositis ossificans

specirnens and 7 normal cortical bone samples. Despite normal or elevated levels of

CBFAI expression in rnost specimens, osteocalcin expression was low or undetectable in

rnost osteosarcornas (2334) and myositis ossificans (415) cases. No CBFM DNA

binding dornain mutations were identified by single strand conformation polymorphism

and sequencing CO account for this observation. However. a high level of iMSX3

expression was demonstrated in these Iesions, which may be inhibitory to OC

tmscription. The presence of moderate levels of OC found in fibrous dysplasia rnay

contribute to the charactenstic disconnected appemnce of trabeculae in that entity. since

OC is a negative regulator of bone formation.

WI'RODUCTION

Osteoid - forrning tumours

Osteoid is the extracellular matrix of bone that is normally produced by mature

osteoblasts. CelIs found in some patholo$c processes such as osteosarcoma, fibrous

dysplasia and myositis ossificans also produce osteoid. Osteosarcoma is a skeleral

maligancy in which, by definition, osteoid is produced directly by rnalignant spindle

ce11.s~. Some degree of morphologie osteoblastic differentiation can be seen in cases of

osteosarcoma, but pathologists have noted the cellular pleomorphism of this

116.777

tumour '- . Althoush it might be natunlly assumed that osteosxcoma cells are denved

from osteoblast progenitors, this notion has not been explicitly demonstrated in the

literature by a systematic effort to better define the origin and lineage of these cells from

human tumour specimens. Expression of some osteoblast - associated markers such ris

alkaline phospharase2's. osteonectinE9 and bone morphogenic proteins '30 has bren

demonstrated. pnmarily in cell lines. Fibrous dysplasia is a benign skeletal lesion

characterized by the production of disconnected trabeculae of woven bone directly from

spindle cells in a fibrous stroma backgroundz'. The term fibro-osseous rnetaplasia has

been used to refer to the bone production in fibrous dysplasia because of the absence of

histologically typical, cuboidal osteobiasts on the surface of the osteoid3'. Myositis

ossificans is a benign, usually self-limited ossifying lesion of muscle, often occurring

secondary to trauma, Proliferating undifferentiated mesenchymal cells infiltrate skeletal

muscle in the acute phase. This lesion exhibits a clear zona1 architecture within four

weeks. in which there is progressive osteoblastic differentiation accornpanied by

decreasing proliferation and increasing bone formation from the center toward the

periphery seen in cross-~ection'~. The peripheral cells thus mature first, and ossification

progresses from the periphery towards the center over tirne. In osteosarcorna and fibrous

dysplasia in particular, osteoid is fonned from cells that do not have the typical histologic

appearance of mature osteoblasts. Genes specific to normal osteoblastic differentiation

are likely expressed by these pathologie cells.

Osteoblast differentiation

Osteocalcin (OC) is a bone - specific extracellular matrix protein, whose

expression foIIows the prolifentive phase of osteoblastic differentiation3< It is

considered a rnarker of mature oste~blasts'~~. Among the Factors that regulate OC. the

transcription factors C B F A ~ ~ ' and M S X ~ ~ ~ " ~ ~ are known to upregulate and

downreguIate its expression. respectively.

CBFAl (Core binding factor alpha 1) is a molecular sigril that has recently been

s h o w to commit mesenchymal cells to the osteoblastic lineage. CBFM is a. ntnt -

domain containing gene which is expressed in cells of the mesenchymal adage of the

developing skeleton and continues to be expressed exclusively in those cells that

differentiate dong the osteoblastic lineage3". CBFAI homozygous nuIl mutant mice

disphy a complete lack of endochondnl and intmembnnous ossification.

demonstrating that CBFA1 is necessary for osteogenesis'jf. Forced expression of CBFAI

in nonosteoblastic cells demonstrates that CBFAl is sufficient for transcription of

osteoblast - associated genes such as Bone sialoprotein, ïype Icollagen and OC in

vitro-*. Heterozysous loss of function mutations of this gene are present in cleidomial

7 7 7 3 s

dysplasia- . and CBFAI may potentidly play an oncogenic mie by colIabonting with

the MYC o n ~ o ~ e n e ~ ~ , which has been found io be sporadically arnplified in

osteosar~ornas'~. CBFAL binds the OC prornoter and upregulates its transcription".

In cornparison, the product of the horneobox containing gene MSX2, which aiso

35-36 binds the OC promoter, strongly suppresses OC transcription . MSXZ is expressed

during the early proliferative phase of osteoblastic diffcrentiation, and is downregulated

prior to osteoblast mat~ration~'. Mutations of this gene are found in individuals with

cranios ynostosis"'.

Cells of pathologic processes chat forrn osteoid may be at lest paninlly

committed to the osteoblastic lineqe. but stritled at a stage of differentiation consistent

with their clinical behaviour. with more aggressive lesions at a less differentiated stage.

in this report. 1 investigate whether this comrnitment rnight be reflected in the pattern of

transcription of these osteoblast linelige-specific senes.

RESULTS

Analysis of samples by semiquantitative reverse transcription-polyrnense chain

reaction (RT-PCR) revealed that sirniIar levels (within 4X of each other) of OC. CBFAI

and MSX2, were expressed by most normal conical bone samples. The bone sample (BO)

chosen to compare relative pathologic specirnen expression levels was representative of

this group (Tables 7 and 3). The osteoblast cells h m short term culture expressed OC

and LW at a lower Ievel compared to the cortical samples (Table 3).

CBFAI was expressed by the osteosarcoma specimens at levels similar to cortical

bone in 28 of 34 cases (82%) (Table 3)- OC expression was decreased in 25 of 34 (74%)

osteosarcornas compared to normal bone (Table 3. Figure 1)- In nineteen of these

samples, OC transcripts were undetectable. In order to determine whether the lack of OC

expression was due to disniption of CBFAl transcriptional activity in osteosarcoma.

mutational analysis of the ntnr DNA binding domain of CBFAI was canïed out. No

mutations were found in the osteosarcorna specimens screened by single strand

conformation polymorphism analysis (SSCP) and sequencing. MSX2 expression levels in

the osteosarcorna specimens were rnosrly comparable to that of normal bone (29134 or

85%) (Table 3. Figure 1). The level of expression did not correlate with a specific

histologic subtype or -mde of osteosarcoma for any gene studied (data not shown).

In the fibrous dysplasia and rnyositis ossificans specimens, expression levels of

CBFAI and h2.W both were comparable to those of normal bone in 315 cases. and low in

Y5 cases. Expression levels of OC in myositis ossificans. like osteosarcoma. were ni1 in

most (415) cases. OC Ievels were never zero in fibrous dysplasia (Table 3. Figure 1).

Fibrous and ossified areris of one case of myositis ossificans (MO% and MOSb.

respectively) demonstrated similar levels of expression of al1 3 genes studied.

DISCUSSION

The osteoid forming lesions I studied expressed osteoblast linea~e specific genes

despite the frequent presence of Iess chan typicil osteoblasts. CBFAI expression was

seen in al1 entities at levels comparable CO normal bone. suggesting that they are at lem

partially committed to the osteoblastic Iineage. The ovenll lack of _pss variation in

CBFAI transcription levels between specimen t-vpes of differing histologic appearance

and clinical behaviour is compatible with evidence from munne in situ hybridization

studies demonstrating that CBFAl is transaibed in mesenchymd osteoblast precursors

and throughout osteoblastic differentiation".

In contrast, expression of osteocakin was variabie between specimen types.

Fibrous dysplasia specimens expressed OC in a11 specimens, and did so at Ievels

comparable to normal bone in 315 (60%) of cases. suggesting that fibrous dysplasiri cells

are fairIy differentiated along the osteoblastic lineage at the transcnptional level, despite

the lack of phenotypically recogizable osteoblasts"'. These cells may represent hybnd

mesenchymd cells thrit dilferentiate pmially dong both the fibroblmtic and osteoblastic

lineages.

The absence of OC transcripts in most cases of osteosarcoma and myositis

ossificans suggests that ossification in these lesions does not require OC. This finding is

consistent with the phenotype of the OC deficient mice genecrtted by Ducy et alfJS which

develop abundant bone. The lack of OC is associated with an increase in bone of

improved functional qunlity without affecting the number of osteoblasts, bone resorption

or mineralization. suggesting that OC is a negative regulator of bone formation. OC

expression in fibrous dysplasia. which on the whole was higher than in the other lesion

types (Table 3). may therefore contribute to inhibition of bone formation in that entity'j5.

In osteosrucoma, CBFAI was apparently unable to upregulate transcription of

osteocalcin. There may be sevenl masons for this observation. It might be explained by

the relative immaturity of osteosarcoma celis. That is. there may be other factors

expressed by these ceils rit their stage of differentiation which inhibit OC transcription.

Induction of osteocakin expression in an osteosarcoma celi line has been found by

Chandaret alB5 to be dependent on the presence of wild - type p53. which is mutated in

a subset of osteosarcornas'j6. Lack of OC transcription may be due ro the presence of

CBFAl dysfunction in osteosarcorni. However. no mutations of the DNA binding nlnr

domain were identified. To further investigate why OC transcription is inhibited in

osteosarcornas, expression of an inhibitor of OC transcription, MSX2(264,275), was

anal yzed.

MSX2 transcription is a marker of the early, proliferative phase of osteoblat

differentiation(256}, and therefore would be expected ro be expressed at a high level in

relatively undifferentiated cells, such as those of osteosarcoma, and at a lower level in

more differentiated cell types. However. such a stmightfonvard correlation wiis not

observed. Although the level of MSX2 tnnscripts was robust in most osteosarcorna

specimens, comparable levels were also observed in the fibrous dysplasia and myositis

ossificans cases analyzed (Table 3). possibly reflecting the actively proliferating subset of

cells in these lesions. Unexpectedly, a similar level of h W expression was obsewed in

cortical bone, where the ceils presumribfy are mature, This finding may be partially due

to residual rnicroscopic periosteum left behind after dissection. or may be evidence of a

prolifentin,o, less mature subpopulntion of osteoblasts within cortical bone.

Altematively, MSX2 may indeed be expressed by mature osteoblasts. contrary to in vitro

studies2".

The results suggest that osteoid - forming patholo@c cells are at least partially

committed to the osteoblastic lineage at the genetic leveI despite their lack of

characteristic histologïc features of osteoblasists- However, the data also indicate that

clinical behaviour cannot be predicted on the basis of differential expression of

tmscription factors (CBFAI and M S E ) that induce phenotypic haiImarks.

rnTHODS

S pecimens

Cryopresewed pathologie specimens were obtained from the Sürcoma Tissue

Bank at Mount Sinai Hospital, Toronto. Specimens in this tissue bank are cut from

within the substance of n tumour, and confinned histologïcally as being entirely lesional

tissue folIowing open biopsy or resection by a patholo_oist experienced in musculoskeIetal

neoplasia prior to being flash frozen in liquid nitrogen and stored at -70'~- I feel the

specimens are reasonably representative samples of each turnour, given the similar IeveIs

of expression of at least one gene (MDRI) from different locations within osteosarcomas

demonstnted eariier'? 1 studied 34 osteosarcoma. 5 fibrous dysplasia and 5

posttnumatic. intnmuscular myositis ossificans cases. The osteosarcoma specimens

were al1 high _mde central lesions. For one case of myositis ossificans. both rhe fibrous

and ossified portions of the same lesion were available for study.

Cortical bone samples were obtained intraoperatively from individuals ranging in

age from 10 to 72 years. Normal bone from fresh osteochondral allo_mft specimens.

residual fresh auto-mft and fracture cases were compiled as a panel of controls. The

lower extremity. pelvis and spine were the anatomical sources of the specimens.

Periosteum and endosteum were completely stripped leaving only cortical bone. The

samples were then processed in the same manner as the tumour samples.

A short term osteobtast culture was established from washed tnbecular bone

obtained intraoperatively. The osteoblast population was passaged once pnor to being

harvested and appropriate expression of the osteoblastic markers aikaline phosphatase,

OC and CBFAI demonstrated.

Semi-quantitative RT-PCR

RNA extraction RT-PCR was cmïed out as descnbed in Chapter 2. PCR

reactions were carried out in a DNA thermal cycter at an annealing temperature of 5 8 ' ~

for 21,23 and 25 cycIes for CBFAIIEM and OC/bM primer pairs and for 24,26 and 28

cycles for the MSXYAS primer pair to ensure cornparison within the linear range of

amplification'JS. PCR products were electropho~sed on 12% (CBFAUB2M and

MSXYAS) and L6% (OC / ,M) polyacryiamide gels and stained with ethidium bromide.

Positive and negative controis for each primer pair, including a no-template water control

carried over from the RT step, were run on each gel, A negative control lymphobllistic

cell Line was included for each pair of genes. Relative levels of band intensity of the gene

of interest were compared CO the interna1 controI house keeping gene and norrnalized to a

positive controi normal bone (BO) sample on each gel. Band intensities over 4X or under

JX chat of BO were sirbitrariIy designated high and low expression. respectively.

Intensities within 4X that of BO were designated as comparable levels of expression.

Mutational analysis

Single stnnd conformlicion polymorphism (SSCP) analysis"'3 was used to screen

for mutations in the ninr DNA binding domain of the CBFAI sene in the osteosarcorna

specimens. Genomic DNA was used as template for PCR amplification of fn, aments

containing an exon. and its adjacent intronic bounduÏes using primers @en in Table lb.

The "P-ATP incorponted PCR prcduct was denatured and electmphorered on a native

polyacrylamide geI containing 10% glycerol. Band s h i f ~ wcre fucther ruialyzed using a

sequencing kit (Thermo Sequenase Sequencinz Kit. Amersham Life Science, CleveIand,

Ohio).

TABLE la: PCR primer sequences (5' - 3')

CBFAI (Core binding factor 1)

oc (Osteocalcin)

AS (Asparagine synthetase)

SEQUENCE (5' - 3') AMPLICON GENBANK (BP) ACCESSION

F Cr\AGTAGCAAGGTTC.UCGA

R TGAGGCGGTCAGAGAACAA

F ATGAGAGCCCTCACACTCC

R GCCGTAGAAGCGCCGATA

F CAGAAGATGGAGCGGCGT

R GCTCTGCMTGGAGAGGTA

F KCCCCACTGAAAAAGATGA

R GGAGACAGCACTCAAAGTAG

F GAGGCTTCTGAGGGAACTCT

R TITCTGGTGGCAGAGACAAG

TABLE lb: CBFAl RUNT domain SSCP primer sequences (5' - 3')

CBFAl.1

CBFAI -2

CBFAI 3

F CACCAGCCGGCGCITCAG

R GTAGCCTCTTACCTTGAAGG

F CTGATGTGCCATTAITGCTG

R ATCAAAGGAGCCTAATGTGC

F TGTGTAATCATCAACACTGT

R GGTCTCTGGAAACTCTATT

TABLE 2: Expression ratios: Lesion of interest/ BO

Ratio of the expression level of a given gene in the lesion of interest over the reference bone standard, BO, OSA = Osteosarcorna. FD = Fibrous Dysplasia, MO = Myositis Ossificans. OB = Osteoblast cuIture.

OSA - 1

7 -

3

4

5

6

7

S

9

10

1 1

13

13

14

15

16

17

CBFAl

0.7 1

1 .jO

0.7 1

1 .iO

1.30

1.50

1 .O0

3 .O0

0.50

O

1.20

1-ro

2.20

0.77

5.80

O

0.20

TABLE 3: Categorisation of Expression Levels

Number of specimens of each type that have a level of expression of the gene of interest sigiîïcantly above, comparable to, or below that of the ceference bone standard, BO. OSA = Osteosarcoma, FD = Fibrous Dysplasia, MO = Myositis Ossificans, OB = Osteoblast culture.

NUMBER OF SPECMENS:

OSA ( n a ) FD (n=5) MO (n=5) BONE (n=7) OB (n=l)

Osteocalcin

wlin 4X 9 3 1 6 O

<4X 6 2 O 1 1

O 19 O 4 O O

CBF.41

wlin 4X 28 3 3 7 1

<4x 5 - 3 - 1 0 0

1 w 2

A X 1 O O O O

wlin 4X 29 5 5 7 O

REFERENCES

I I .

Sachs L: Regulators of normal development and tumor suppression. Int J Dev Bi01 1993,375 1-59.

Alexander CM, Reichsman F, Hinkes MT, Lincecum J, Becker KA, Cumberledge S, Bemfield M: Syndecan-1 is required for Wnt-L-induced mammary tumorigenesis in mice. Nat Genet 2000.25:329-332.

Tejpar S. Nollet F. Li C, Wunder JS, Michils G. Dal Cin P. Van Cutsem E, Bapat B, van Roy F, Cassiman JJ, Alman BA: Predominance of betii-catenin mutations and beta-catenin dysregulation in sporadic aggressive fibromatosis (desmoid tumor). Oncogene 1999, 18:66 15-6620.

Lo CF, Pisegna S, Diverio D: The M L 1 gene: a transcription factor involved in the pathogenesis of myeloid and lymphoid leukemias. Haemütologica 1997.82:363-370.

Kelly KM. Womer RB. Brin FG: PAX3-FKHR and PAX7-FKHR gene fusions in rhabdomyosarcoma. J Pediatr Hematol Oncol 1998,70:5 17-5 18.

Ginsberg JP. de Alava E. Ladanyi M. Wexler LH. Kovar H, Paulussen M. Zoubek A. Dockhom-Dworniczrik B. Juergens H, Wunder JS. Andrulis IL, Malik R. Sorensen PH. Womer RB, Barr FG: EWS-FUI and EWS-ERG gene fusions are associated with similar clinical phenotypes in EwingS sarcoma. J Clin Oncol 1999. 17:1S09-18 14.

May WA, Arvand A. Thompson AD, Braun BS, Wright M. Denny CT: EWSfFLII- induced manic fringe renders NM 3T3 cel1s tumorigenic. Nat Genet 1997, L7:495- 497.

Fan H. Oro AE. Scort MP, Khavari PA: Induction of basal ceil carcinoma features in tnnsgenic human skin expressing Sonic Hedgehog. Nat Med 1997,3:788-792.

Oro AE. Higgins LM, Hu 2, Bonifiis JM. Epstein EH, Scott MP: Basal cell carcinomas in mice overexpressing sonic hedgehog. Science 1997,176:S 17-82 1.

Dahmane N. Lee J, Robins P. Helter P, Alraba A: Activation of the transcription factor Gli 1 and the Sonic hedgehog s ip l l ing pathway in skin tumours. Nature 1997, 389:876-88 1.

Xie S. Murone M, Luoh SM, Ryan A. Gu Q, Zhang C, Bonifas JM, Lam CW, Aynes M, Goddard A, Rosenthal A, Epstein EH, de Sauvage FJ: Activating Smoothened mutations in sporadic basakell carcinoma. Nature 1998,39 l:90-92,

Grachtchouk M, Mo R, YU S. Zhang X, Sasaki H, Hui CC, Dlugosz AA: Basal ceIl carcinomas in mice overexpressing GIiî in skin. Nat Genet 2000,24:216-217.

Zurawel RH, Allen C, Chiappa S, Cato W, Biegel J, Cogen P. de Sauvage F, Raffel C: Analysis of PTCWSMOISHH pathway genes in medulloblastoma. Genes Chromosomes Cancer 2000,27:44-5 1.

Alman BA, Li C, Pajerski ME, Diaz-Cano S, Wolfe HJ: Increased beta-catenin protein and somatic APC mutations in sporadic aggressive fibromatoses (desmoid tumors). Am J Pathol 1997. 15 1:329-334.

van Es JH, Giles RH, Clevers HC: The many faces of the tumor suppressor gene APC. Exp Cell Res 200 1,264: 126-134.

Hahn H, Wicking C. Zaphiropoulous PG, Gailani MR, Shanley S, Chidambaram A. Vorechovsky 1, Holmberg E, Unden AB. Gillies S, Negus K. Smyth 1. Pressrnan C. Leffell DJ, Gerrard B, Goidstein Ai, Dean M, Toftgard R, Chenevix-Trench G. Wainwright B, Bale AE: Mutations of the hurnan homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Ceil 1996,85:84 1-85 1.

Li C. Bapat B. XIrnan BA: Adenomatous pol-vposis coli gene mutation alters prolifemtion through its beta-catenin-regulatory function in ag,gessive fibrornatosis (desmoid tumor). Am J Pathol 1998, 153:709-714.

LMcMahon AP: More surprises in the Hedgehog signalino pathway. Cell 1000. 100: 185-13s.

Levesque J. Marx R. BeIl RS, WunderJS. Krindel R. White L: Enchondroma. A Clinicul G i d e ro Pritwn, Bone Triniors. Philadelphia. Williams and Wilkins, 1998. pp. 98-107.

Huvos AG: Solitary and multiple osteochondrornas and enchondromas: jusxtacortical chondrornx maffuci's diserise. Bone rrtniors, diusnosis rrciatnienr andprognosis. Philadelphia, W B Saunders, 199 1, pp. 264-294.

Bovee N. Cleton-Jansen Aii, Wuyts W, Caethoven G, Tminiau AH, Bakker E, Van HUI W. Cornelisse CJ, Hogendoom PC: EXT-mutation analysis and Ioss of heterozygosity in spondic and hereditary osteochondrornas and secondary chondrosarcomas- Am J Hum Genet 1999,65:689-698.

Hecht ST, Hogue D, Strong LC, Hansen MF, Blanton SH, Wagner M: Hereditary multiple exostosis and chondrosarcoma: linkage to chromosome II and loss of heterozygositv for EXT-linked markers on chrornosornes II and S. Am J Hum Genet 1995,jo: 112J4 13 1.

Stickens D, Clines G, Burbee D, Ramos P, Thomas S. Hogue D, Hecht JT, Lovett M, Evans GA: The EXT2 multiple exostoses g n e defines a family of putative tumour suppressor genes. Nat Genet 1996, 1425-32-

Bellaiche Y, The 1, Pemmon N: Tout-velu is a DrosophiIa hornoiogue of the putative tumour suppressor EXT-1 and is needed for Hh diffusion. Nature 1998,394:85-88-

The 1, Bellaiche Y, Pemmon N: Hedgehog movement is regulated through tout velu- dependent synthesis of a heparan sulfate proteoglycan, Mol Ce11 1999,4:633-639,

Brien EW, Micra hl, Kerr R: Benign and maligant cartilage tumon of bone and joint: their anatomic and theoretical basis with an emphasis on radiology, pathology and clinical biology. 1. The intramedullary cartilage tumors. SkeletaI Radiol 1997, 26:325-353.

Spjut W, Dorfman HD, Fechner RE, Ackerman LV: Turnors of Cartilaginous Origin. Tumors of Bone and Cartilage. Washington, Armed Forces Institute of Pathology, 1983, pp. 73-1 10.

Schwam HS, Zimmerman NB, Simon MA. Wroble RR. MilLu €A. Bonfiglio M: The maIigmt potential of enchondromatosis. J Bone Joint Surg [Am 1 1987,69269-274.

Hdal F, Azouz EM: Generalized enchondromatosis in a boy with only platyspondyly in the father. Am J Med Genet 1991,38:588-592.

Zack P. Beighton P: Spondyloenchondromatosis: syndrornic identity and evolution of the phenotype. Am J Med Genet 1995,55:478182.

Schorr S. Legum C, Ochshom M: Spondyloench~ndrodysplasia~ Enchondromatomosis with severe platyspondyly in two brothers. Radiology 1976, L 1S:133-139.

Chang S. Pndos MD: Identical twins with Ollier's disease and intracranial gliomas: case report, Neurosurgery 1994.34:903-906.

Haddad FS, H q e r GD, Hill RA: Intraopentive arthropphy and the [lizarov technique. Role in the correction of paediatric deformity and leg lengthening. J Bone Joint Surg Br 1997,79:73 1-733.

Felbel J. Gresser U. Lohmoller G. Zollner N: Familial synoviaI chondromatosis combined with dwarfism. Hum Genet 1992,88:35 1-354.

Steinberg GG, Desai SS, Malhotra R, Hickler R: Familial synovial chondromatosis: brief report. J Bone Joint Surg Br 1989,71: LM-145.

Davis RI, Hamilton A, Biggart JD: Primary synovial chondromatosis: a ~Iinicopathologic review and assessment of maligant potential. Hum PathoI 1998. 29:683-688.

Wuisman PI, Noorda RI, Jutte PC: Chondrosarcoma secondary to synovial chondromatosis. Report of two cases and a review of the literacure. hrch Orthop Trauma Surg 1997, 1 16:307-3 11.

Swarts SJ, Neff .IR, Nelson MF Johansson S, Bridge JA: Chromosoma1 abnormaiities in Iow _mde chondrosarcoma and a review of the Iiterature. Cancer Genet Cytogenet L997.98:126-130.

Jagasia AA, Block JA, Qureshi A, Diaz MO, Nobori T, Gitelis S, tyer AP: Chromosome 9 related aberrations and deletions of the CDm-2 and MTS2 putative tumor suppressor genes in human chondrosarcomas. Cancer Lett 1996, 105:91-103.

Sciot R, Da1 Cin P, Fletcher C, Samson 1, Smith M, De Vos R, Van Damme B, Van den BH: t(9:22)(q22-3 1:qlI-12) is a consistent marker of extraskeletal myxoid chondrosarcoma: evaluation of three cases. Mod Pathol 1995. S:765-768.

Lsibelle Y, Bussieres J, Courjal F, Goldring MB: The EWSKEC fusion protein encoded by the t(9:22) chrornosomal translocation in human chondrosarcornas is a highly potent transcriptional activator. Oncogene 1999, L8:3303-3308.

Wang ZQ, Grigoriadis AE, Mohle-Steinlein U, Wagner EF. A novel target cell for c- fos-induced oncogenesis: development of chondrogenic tumours in embryonic stem cell chimem. EMBO J 199 1, 10:2437-2450.

Franchi A, Calzolari A, Zarnpi G: Immunohistochernical detection of c-fos and c-jun expression in osseous and canilaginous tumours of the skeleton. Virchows Arch 1998. 432~5 15-5 19.

Ionescu AM, Schwarz EM, Vinson C. Puzas JE. Rosier RN, Reynolds PR. OKeefe RI: PTHrP modulates chondrocye differentiation through AP-1 and CREB sigding. J Bioi Chem 200 1. .

Pompetti F, Rizzo P. Simon RM. Freidlin B. Mew DJ, P a s HI, Picci P. Levine AS, Carbone M: Oncogene altentions in prirnary, recurrent. and metastatic human bone tumors. J Cell Biochem 1996.63:37-50.

Kanoe H. Nakaiyama T, Murrikami H. Hosaka T, Yamamoto €3, Nakashima Y. Tsuboyama T. Nakamun T, Sasaki MS, Toguchida J: Amplification of the CDK4 oene in srircomas: tumor specificity and telationship with the RB gene mutation. 3

Anticancer Res 1998. 1823 17-23? 1.

Yamaguchi T, Toguchida J. Wadayama B. Kanoe H, Nakayama T, Ishizaki K, lkenaga M, Kotoun Y, Sasaki MS: Loss of heterozygosity and tumor suppressor g n e mutations in chondrosarcornas. Anticancer Res 1996, 16:2009-2015.

Bovee JV, van Den Broek W, Cleton-lansen AM, Hogendoom PC: Chondrosarcoma is not characterized by detectable telornerase activity. J Pathol2001, 193:354-360.

Raskind WH, Conrad EU, Chansky H, Matsushita M: Loss of heterozygosity in chondrosarcomas for markers Iinked to hereditary multiple exostoses loci on chromosomes 8 and 11. Am J Hum Genet 1995,56:1132-1139.

Ozisik YY, MeIoni AM, Spanier SS, Bush CH, Kingsley KL, Sandberg AA: Deietion lp in a low-_mde chondrosarcoma in a patient with Ollier disease. Cancer Genet Cytogenet 1998.105:128-133.

Bovee JV, van Roggen JF, Cleton-Jansen AM, Taminiau AH, van der Woude HJ, Hogendoorn PC: Malignant progression in multiple enchondromatosis (Ollier's disease): an autopsy-based molecular genetic study, Hum Pathol 2000,3 1: 1299-1303.

Zaleske DJ: Overview of Musculoskeletal Development. Skeletal Growth and Development. Edited by by Buckwalter JA, Ehrlich MG, Sandell W, Trippl SB. Rosemont, Amencan Academy of Onhopaedic Surgeons, 1998, pp. 5-16.

Tuan RS: Cellular and Molecular Regulation of Embryonic Skeletal Development. Skeletal Growth and Development, Edited by Buckwalter JA, Ehrlich MG, SandelI W. Trippel SB. Rosemont, Amencan Academy of Onhopaedic Surgeons, 1998, pp- 17- 36.

Tickle C: Molecular basis of limb development. Biochem Soc Trdns 1994,22:565- 569.

Duboule D: How to make a limb? Science 1994,266575-576.

Maden M: Developmental biology. The limb bud-Part two. Nature 1994.37 1560- 561.

Johnson RL, Tabin CJ: Molecular models for venebrate limb development. Cell 1997. 90:979-990.

Karsenty G: The genetic transformation of bone biology. Genes Dev 1999, 13:3037- 305 1.

Vu TH, Shipley JM, Bergers Ci. Berger SE. HeIms JA, Hanahan D. Shapiro SD, Senior RM, Werb Z: MMP-9lgelatinrise B is a key regulator of growth plate mgiogenesis and apoptosis of hypenrophic chondrocytes. CeIl 1998,93:411122.

Gerber HP, Vu TH. Ryan AM, Kowalski J, Werb 2, Ferrara N: VEGF couples hypemophic cmilage remodeling, ossification and angiogenesis during endochondrül bone formation [see comments]. Nat Med 1999.5623-628.

Tabatci T, Eaton S. Kornberg TB: The Drosophila hedgehog gene is expressed specifically in posterior compartment cells and is a target of engaiIed replation. Genes Dev 1992,62635-7645.

Lee JJ, von Kessler DP, Parks S. Beachy PA: Secretion and locaiized transcription suggest rt role in positional sigaling for products of the se-mentation gene hedgehog. Cell 1992.7 L:33-50.

Levin M, Johnson RL, Stem CD, Kuehn M, Tabin C: A molecular pathway determining left-right asyrnrnetry in chick embryogenesis. CeII 1995,82:803-8 14.

Johnson RL, Laufer E, Riddle RD, Tabin C: Ectopic expression of Sonic hedgehog alters dorsal-ventral patterning of somites. Ce11 1994,79: 1165-1 173.

Fan CM, Porter JA, Chiang C, Chang DT. Beachy PA, Tessier-hvigne M: Long- range sclerotome induction by sonic hedgehog: direct role of the arnino-termina1 cleavage product and modulation by the cyclic APUlP signalin; pathway. Ce11 1995. 8 1:457465.

Teillet M. Watanabe Y, Jeffs P, Duprez D, Lapointe F, Le Douarin NM: Sonic hedgehog is required for survival of both myogenic and chondrogenic somitic lineages. Development 1998, 1252019-2030.

Echelard Y, Epstein DJ, St Jacques B, Shen L, Mohler J, McMahon JA, McMahon AP: Sonic hedgehog, a mernber of a family of putative s i~a l i ng molecules, is implicated in the regulation of CNS polarity. Cell 1993,75: 1417-1430.

Roelink H. Augsbur~er A, Heemskerk J, Korzh V, Norlin S. Altaba A, Tanabe Y, Placzek M. Edlund T. Jessell TM, .: Floor plate and motor neuron induction by vhh-1, a venebrate homolog of hedgehog expressed by the notochord. Cell 1994,76:76 1-775.

Bitgood MJ, Shen L. McMahon AP: Senoli cell signaling by Desert hedgehog regulates the male germline. Curr Biol 1996,6:298-304.

Dyer MA, Fanington SM. Mohn D, Munday JR, Baron LW: Indian hedgehog activates hematopoiesis and vasculogenesis and can respecify prospective neurectodermal ce11 fate in the mouse embryo. Development 2001, 128: 1717-1730.

Litingtung Y. Lei L. Westphal H, Chiang C: Sonic hedgehog is essential to foregut development. Nat Genet 1998,20:58-6 1.

Motoyama J, Liu J. Mo R. Ding Q, Post M. Hui CC: Essential function of Gli2 and Gli3 in the formation of Iung, trachea and oesophagus. Nat Genet 1998.2054-57.

Hebrok LM. Kim SK, St Jacques B. McMahon AP, Melton DA: Regulation of pancreas development by hedgehog signaling. Development 2000. 127:4905-1913.

Bellusci S. Fuma Y, Rush MG, Henderson R. Winnier G, Hogan BL: Involvement of Sonic hedgehog (Shh) in mouse embryonic lung growth and rnorphogenesis. Development 1997.124:53-63.

Ramalho-Santos M, Melton DA, McMahon AP: Hedgehog sigals regoiate multiple aspects of gastrointestinal developrnent. Development 2000,127:2763-2772.

St Jacques B. Dassule HR, Karavanova 1, Botchkarev VA, Li J, Danielian PS, McMahon JA, Lewis PM, Paus R. McMahon AP: Sonic hedghog sipaling is essential for hair developrnent. Curr Biol 1998,8:1058-1068.

Hardcastle 2. Mo R, Hui CC, Sharpe ET: The Shh signalling pathway in tooth development: defects in Gli2 and Gli3 mutants- Development 1998, 125:2803-28 11.

Riddle RD, Johnson RL, Laufer E, Tabin C: Sonic hedgehog mediates the polarizing activity of the ZPA. Ce11 1993,75:1401-1416.

Vortkamp A, Lee K, Lanske B, Se"- GV. Kronenberg HM, Tabin CJ: Regulation of rate of cartilage differentiation by Indian hedgeho; and FiTi-related protein [see commencs]. Science 1996,273:6 13-622.

Se Jacques B. Hammerschmidt M, McMahon AP: Indian hedgehog sigaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation [published erratum appears in Genes Dev 1999 Oct 1;13(19):2617]. Genes Dev 1999. 132072-2086.

Chung üI, Schipani E, McMahon AP, Kronenberg HM: Indian hedgehog couples chondrogenesis to osteogenesis in endochondral bone development. 5 CIin Invest 200 1. 107:295-304.

RoessIer E, Belloni E, Gaudenz K, Jay P. Bena P. Scherer SW. Tsui LC, Muenke M: Mutations in the human Sonic Hedgehog gene cause holoprosencephaIy. Nat Genet 1996, 1413.57-360.

Kelley RI, Hennekam RC: The Smith-Lemli-Opitz syndrome. J Med Genet 2000. 37:32 1-335.

Kang S. Graham JM. Olney AH, Biesecker LG: GLI3 fnmeshift mutations cause autosoma1 dominant Pallister-Hall syndrome, Nnt Genet 1997, 15266-268.

Vortkarnp A. Gessler M. Grzeschik KH: GL13 zinc-finger gene intemipted by translocaîions in Greig syndrome families. Nature 199 1,352539-540.

Gailani MR. Stahle-Backdahl M, Leffell DJ, Glynn M, Zaphiropoulos PG. Pressrnan C. Unden AB. Dean M, Brash DE, Bale AE. Toftgard R: The role of the human homologue of Drosophila patched in sporadic basal ce11 carcinomas. Nat Genet 19%. 14:78-8 1.

Hahn H, Wicking C, Zaphiropoulous PG, Gailani MR, Shanley S. Chidmbanm A. Vorechovsky 1, Holmberg E, Unden AB, Gillies S. Negus K, Smyth 1. Pressrnan C, Leffell Dl. Gerrard B, Goldstein AM, Dean M, Toftgard R, Chenevix-Trench G, Wainwright B, Bale E: Mutations of the human homolog of Drosophila patched in the nevoid basal ce11 carcinoma syndrome. Ce11 1996.85:841-851.

Xie 3, Murone M. Luoh SM, Ryan A. Gu Q, Zhmg C, Bonifas IM. Lam CW, Hynes M. Goddard A. Rosenthal A, Epstein EH, de Sauvage FS: Activating Smoothened mutations in sporadic basal-ce11 carcinoma. Nature 1998,39 1:90-92.

89. Raffel C. Ienkins RB, Fredenck L, Hebnnk D, Alderete B, Fults D W, James CD: Spondic medulloblastomas contain PTCH mutations- Cancer Res 1997. 57:842-835.

Hahn H, Wojnowski L, Zimmer AM, Hall J, Miller G , Zimmer A: Rhabdomyosmornas and radiation hypersensitiviry in a mouse mode1 of Gorlin syndrome. Nat Med L998,4:6 19-62?.

Marigo V. Roberts DJ, Lee SM, Tsukurov O, Levi T, Gastier JM, Epstein DJ, Gilbert DJ, Copelrtnd NG, Seidman CE, .: Cloning, expression. and chromosomal location of SHH and MH: two human homologues of the DrosophiIa segment polarïty gene hedgehog. Genomics L995,28:U-5 1.

Fietz W. Concordet JP, Batbosa R, Johnson R, Knuss S. McMahon AP, Tabin C, Ingharn PW: The hedgehog gene family in Drosophila and vertebrate development. Dev Suppl 1994, 28:43-5 1 .

Leek JP, Moynihan TP, Anwar R. Bonthron DT, Markham AF, Lench NJ: Assignment of Indian hedgehog (IHH) to human chromosome bands 2q33->q35 by in situ hybridization. Cytogenet Ce11 Genet 1997,76:187-188.

Porter JA, Young KE, Beachy PA: Cholesterol modification of hedgehog signaling proteins in animal development. Science 1996.274255-359.

Yang Y. Guillot P. Boyd Y, Lyon MF. McMahon AP: Evidence that preaxial polydactyly in the Doublefoot mutant is due to ectopic Indiiin Hedgehog signaling. Development 1998. i 25:3 113-3 132.

Becker S. Wang Zi, Massey H. Arauz A, Labosky P. Harnmerschmidt M. Sc Jacques B. Bumcrot D. McMiihon A. Grabel L: A role for Indian hedgehog in extraembryonic endodem differentiation in F9 cells and the early mouse embryo. Dev Bi01 1997, 187:295-3 10.

Dyer MA, Fmïngion SM. Mohn D, Munday .IR, Baron MH: Indian hedgehog activates hematopoiesis and vrisculogenesis and can respecify prospective neurectodemaI ceIl fate in the mouse embryo. Development 200 1. 128: 1717-1730.

Valentini RF. Brookhiser WT, Park J. Yang T, Briggs J. Dressler G. Holzman LB: Post-mslational processing and rend expression of mouse Indian hedgehog. J Bi01 Chem 1997.272:8466-8473.

Murakami S. Nifuji A. Noda M: Expression of Indian hedgehog in osteoblasts and its posttranscriptional regulation by tnnsfonning hmwth factor-beta. Endocrinology 1997, 138: 1972-197s.

100. Bitgood MJ, Shen L. McMahon AP: Senoli ceIl signaiing by Desert hedgehog regulates the maIe _~ermline. Curr Biot 1996,6:298-304.

101. Parmantier E. Lynn B, Lawson D, Turmaine M, Namini SS, Chaiuabarti L, McMahon AP, Jessen KR. Mirsky R: Schwann celIderïved Desert tiedsehog conuols the development of penphed nerve sheaths. Neuron 1999,13:713-714.

Fietz MJ, Jacinto A, Taylor AM, Alexandre C, Ingham PW: Secretion of the amino- ierminal fragment of the hedgehog protein is necessary and sufficient for hedgehog sigalling in Drosophila. Curr Biot 1995,5:643-650.

Johnson RL, Tabin C: The long and short of hedgehog signding. Ce11 L995,S L:3 13- 3 16.

Burke R. Nellen D, Bellotto M. Hafen E, Senti KA. Dickson BJ. Basler K: Dispatched. a novel sterol-sensing domain protein dedicated to the release of cholesterol-rnodified hedgehog from sigaling cells. Cell 1999,99:803-815.

I*

hedgehog- Nature 1996,384: 129-134.

Chen Y, Struhl G: Dual roles for patched in sequestering and transducing Hedgehog. Cell 1996,87:553-563.

Chuang PT. McMahon AP: Venebnte Hedgehog sigalling modulated by induction of a Hedgehog-binding protein. Nature 1999,397:6 17-62 1.

The 1, Bellniche Y. Pemmon N: Hedgehog movement is regulated through tout velu- dependent synthesis of a heparan sulfate proteoglycan. Mol Cell 1999, J:633-639.

Senay C. Lind T, Muguruma K. Tone Y, Kitagawa H, Sugahara K. Lidholt K. Lindahl U. Kusche-Gullberg M: The EXTIEXT2 tumor suppressors: catalytic activities and mle in heparan sulfate biosynthesis. EME3O Rep 2000. 1:281-286.

Bellaiche Y. The 1, Perrimon N: Tout-velu is a Drosophila homologue of the putative tumour suppressor EXT-1 and is needed for Hh diffusion. Nature 1998.394:85-88.

In$am PW. Taylor AM, Nakano Y: Role of the Drosophila patched gene in positional signa1ling. Nature 199 1,353: 184-187.

van den HM, ingharn PW: smoothened encodes a receptor-like serpentine pmtein required for hedgehog signalling. Nature 1996.382547-551.

Alcedo I, Xyzenzon M. Von Ohlen T, Nol1 M. Hooper JE: The Drosophila smoothened gene encodes a seven-pass membrane protein, a putative receptor for the hedgehog signal. Ce11 1996,86221-232.

Chen Y, Stmhl G: Dual roles for patched in sequestenng and transducing Hedgeho;, CeIl 1996.87:553-563.

Marigo V, Davey RA, Zuo Y, Cunningham M, Tabin CJ: Biochernical evidence that patched is the Hedgehog receptor. Nature 1996,384: 176-179.

Stone DM, Hynes M, Armanini M. Swanson TA, Gu Q, Johnson RL, Scott MP, Pennica D. Goddard A. Phillips H, No11 M, Hooper JE, de Sauvage F, Rosenthai A: The tumour-suppressor gene patched encodes a candidate receptor for Sonic

116. Denef N, Neubuser D, Perez L, Cohen SM: Hedgehog induces opposite changes in turnover and subcellular localization of patched and smoothened. Ce11 2000, 10252 L- 53 1.

117. DeCarnp DL, Tnompson TM, de Sauvage FJ, Lemer MR: Smoothened activates Galphai-mediated signahg in frog melanophores. J Biol Chem 2000,275:26322- 26327.

118. Carpenter D, Stone DM. Bmsh J. Ryan A. Arrnanini hi, Fnntz G, Rosenthai A, de Sauvage FJ: Chmcterization of two patched receptors for the venebratt hedgehog protein family. Proc Nat1 Acad Sci U S A 1998,95:13630-13634.

119. Motoyama J. Takabatake T. Takeshima K. Hui C: Ptchî. a second mouse Patched gene is CO-expressed with Sonic hedgehog. Nat Genet 1998, 18:104-106.

120. Motoyama J. Heng H. Cnckower MA. Takabatake T. Takeshima K. Tsui LC. Hui C: Overlapping and non-overlappin; Ptch2 expression with Shh during mouse embryogenesis. Mech Dev 1998.78:S 1-M.

121. Aza-Blanc P. Ramirez-Weber FA. Laget h4P. Schwartz C. Komberg TB: Proreolysis that is inhibited by hedgehog targets Cubitus intemptus protein to the nucleus and convens it to a repressor. Cell 1997. 89: 1033-1053.

132. Aza-Blanc P. Komberg TB: Ci: a complex tnnsducer of the hedgehog s i p l . Trends Genet 1999. 15:45S462.

123. Sasaki H. Nishizaki Y, Hui C. Nakafuku M. Kondoh H: Regulation of Gli7 and Gli3 activities by an amino-teminal repression dornain: implication of GliZ and GIi3 as p r i m q mediators of Shh signaling. Development 1999. 126:3915-3924.

124. Marigo V. Johnson RL. Vonkamp .4. Tabin CJ: Sonic hedgehog differentially regulates expression of GLI and GLU during limb development. Dev Bi01 1996. 180:173-283.

125. Sisson JC. Ho KS, Suyama K. Scott W: Costal?. a novel kinesin-related protein in the Hedgehog sigaling pathway. Cell 1991.90235-245.

126- Wang QT. Holm_mn RA: Nuclear irnport of cubitus intenuptus is regulated by hedgehog via a mechanism distinct from Ci stabilization and Ci activation. Development 2000. 1273 13 1-3 139.

127. Robbins DJ, Nybakken KE. Kobayashi R. Sisson JC. Bishop JM, Therond PP: Hedgehog elicits signai transduction by means of a large complex contriining the kinesin-related protein costal2 Ceil L997.90225-234.

128. Monnier V. Dussillol F, Aives G. Lamour-lsnard C, Plessis A: Suppressor of fused links fused and Cubitus intemptus on the hedgehog signalling pathway. Curr Biol 1998,8553-586.

179. Ding Q, Fukarni S, Meng X, Nishizaki Y, Zhang X, Sasaki H. Dlugosz A. Nakafuku M. Hui C: Mouse suppressor of fused is a negative regulator of sonic hedgehog signaling and alters the subcellular distribution of Gli 1. Curr Biol 1999.9: 11 19-1 113.

130. Jiang J, Struhl G: Protein kinase A and hedgehog signaling in Drosophila Iimb development. Cell 1995,80:563-572.

13 1. Lepage TT. Cohen SM. Diaz-Benjumea FJ, Pxkhurst SM: Signal transduction by cAiW-dependent protein kinase A in Drosophila Limb pntterning. Nature 1995. 373:7 1 1-7 15.

132. Hammenchmidt M. Bitgood MJ, McMahon AP: Protein kinase A is a common negative regulator of Hedgehog signaling in the venebrate embryo. Genes Dev 1996. 10:647-658.

133. Price MA, Kalderon D: Proteolysis of cubitus intemptus in Drosophila requires phosphorylation by protein kinase A. Development 1999. l26:433 14339.

134. Wang G. Wang B. Jiang J: Protein kinase A antagonizes Hedgehog signaling by regulating both the rictivator and repressor forms of Cubitus intemptus, Genes Dev 1999. 13:282S-2837.

135. van de SA. Karperien M. Lowik CW. Juppner H. Segre GV. Abou-Srimra .+B. de b a t SW. Defize LH: Parathyroid hormone-related peptide as an endogenous inducer of parietal endoderm differentiation. J Cell Biol 1993. 120:235-243.

136. Beck F. Tucci J. Senior PV: Expression of parathyroid hormone-related protein mRNA by uterine tissues and extnembryonic membranes during gestation in rats. J Reprod Fenil 1993,99:343-351.

137. Krirperien M. Lanser P. de Laat SW. Boonstra J. Defize LH: Parathyroid hormone reIated peptide rnRNA expression during murine postirnplmtation development: evidence for involvement in multiple differentiation processes. Int J Dev Bi01 1996. 40599-60s.

138. Wysolmerski JJ. Stewart A F The physiology of panthymid hormone-related protein: an ernerging role as a developmental factor. Annu Rev Physiol 1998,60:43 1-60.:43 L- 460.

139. Lee K. Deeds JD, Se-gre GV: Expression of parathyroid hormone-related peptide and its receptor messenger ribonucleic acids during fetal deveIopment of rats. Endocrïnology 1995. 136:453363.

140. Lee K. Deeds JD. Bond AT. Juppner H. Abou-Sam AB, Seb= GV: In situ locaiization of E T W ï H i P receptor mRNA in the bone of fetaI and young nts. Bone 1993. l4:3$ 1-345-

141. Yang KH, Stewart The PTH-related protein gene and protein pmducts, Principles of Bone Biology. Edited by Bilezikian JP, Raisz L. and Rodan G. San Diego, Academic, 1996. pp. 337-376.

143. OrIoff JJ. Stewart AF: Pmthyroid hormone-rehted protein as a pmhormone: posttranslationai processin; and receptor interactions. Endocrine Reviews Monograph Series. Edited by Negro-Villar A. Bethesda. Endocrine Soc.. L995, pp. 207-210.

1-13. Juppner H. Abou-Sam AB. Freeman M, Kong ,W. Schipani E, Richards J, Kolakowski LF. fr.. Hock J. Ports JT, Jr., Kronenberg HM: A G protein-linked receptor for parathymid hormone and prtrarhyroid hormone-related peptide. Science 1991.254:1024-1026.

144. Rubin DA. Hellman P. Zon LI. Lobb CJ, Berswitz C. Juppner H: A G protein-coupIed receptor from zebnfish is activated by human parathymid hormone and not by human or teleost pmthyoid hormone-related peptide. Implications for the evolutionary conservation of calcium-regulating peptide hormones, J Biol Chern 1999.27433035- 230;12.

145. Rubin DA. Juppner H: Zebrrifish express the cornmon panthyroid hormone/parathyroid hormone-related peptide receptor (PTHIRI and a novel receptor ffTH3R) that is preferentially activated by rnamrnalitin and fugufish piirrtthyroid hormone-related peptide. I Bi01 Chem 1999.274:28 185-28 190.

146. Schwindinger WF. Fredericks J. Watkins L. Robinson H. Bathon JM. Pines M. Suva U. Levine MA: Coupling of the mWPTHrP receptor to multiple G-proteins. Direct demonstmion of receptor activation of Gs. Gqll 1. and Gi(1) by [alpha-32PIGTP- gamma-rizidorinilide photoafftnity labeling. Endocrine L99S. 820 1-209.

147. Iuppner H: Molecular cloning and chruacteriration of a pmthyroid hormone/panthyoid hormone-related peptide receptor: a member of an ancient fmiIy of G protein-coupled receprors. Curr Opin Nephroi Hypertens 1994.3:37 1-378.

14s. Lewin 0: S i p i transduction. Genes. New York. Oxford University Press. 2001. pp. 1053-1088.

L.19. FarfeI 2. Bourne HR. Iiri T: The expmding spectrum of G protein disecises. N Enz1 J Med 1999.3.10: 10 12- 1020.

150. Iobert AS. Leroy C. Butien D. Silve C: Pmthyroid hormone-induced calcium release from intracelIuiar stores in a human kidney ce11 Line in the absence of stimulation of cyciic adenosine 3'5'-monophosphate production. Endocrinology 1997, L385182- 5297-

Verheijen MEi. Defize LH: Pmhyroid hormone activates mitogen-activated protein kinase via a CAMP-rnediated pachway independent of Ras. I Biol Chem 1997. 2723423-3429.

152. Henderson JE, Amizuka N, Warshawsky H, Biasotto D, Lanske BM, Goluman D, Karapiis AC: Nucleolar localization of parathyroid hormone-related peptide enhances sumival of chondrocytes under conditions that promote apoptotic ce11 death. MOI Ce11 Bi01 1995. 15:40644075.

153. Watson PH, Fraher W, Hendy GN, Chung UI, Kisiel M, Natale BV, Hodsman AB: Nuclear localization of the type 1 PTWPTHrP receptor in rat tissues. J Bone Miner Res 2000, 15: 1033-1044,

154. Huang Z, Bambino T, Chen Y, Larneh J, Nissenson RA: RoIe of signal transduction in intemalization of the G protein-coupled receptor for parathyroid hormone (PTH) and PTH-related protein. Endocrinology 1999, 140: 1294-1300.

155, Ishizuya T, Yokose S. Hori M. Noda T, Suda T, Yoshiki S, Yamaguchi A: Panthymid hormone exens disparate effects on osteoblast differentiation depending on exposure tirne in rat osteoblastic cells. J Clin Invest L997,99:2961-2970.

156. Lanske B. Karaplis AC, Lee K, Luz A, Vonkamp A. Pirro A. Karperien M. Defize WK. Ho C. Mulligan RC, Abou-Samra AB, Juppner H, Segre GV, Kronenber_o HM: PTWPTHrP receptor in early development and Indian hedgehog-regulated bone orowth [see cornrnents]. Science 1996,273:663-666. -

157. Lee K. Lanske B. Karaplis AC, Deeds JD, Kohno H, Nissenson RA. Kronenberg HM. Segre GV: Parathyroid hormone-related peptide delays terminal differentiation of chondrocytes during endochondral bone development. Endocrinoiogy 1996. 1375 109- 51 18.

158. Amizuka N. Henderson SE. Hoshi K. Warshawsky H. Ozawa H, Goltzrnan D. Knraplis AC: Pro_prnrned ceIl death of chondrocytes and aberrant chondro~enesis in mice homozygous for parathyroid hormone-related peptide gene deletion. Endocrinology 1996. 137:5055-5067.

159. Amizuka N. Kanplis AC. Henderson JE, Warshawsky H, Lipman ML, Mrttsuki Y, Ejii S. Tanrika M, izumi N, Ozawa H. Goltzrnan D: Haploinsufficiency of pmthyroid hormone-related peptide (FTHrP) results in abnormal postnatal bone development. Dev Bi01 1996. 175: 166-176.

160. Lanske B, Amling M. Neff L. Guiducci J. Baron R, Kronenberg HM: Ablation of the PTHrP gene or the PTWPTHrP receptor gene leads to distinct abnormalities in bone developrnent. J Clin Invest 1999, lW399407.

161- Chung üi, Lanske B, Lee K. Li E, Kronenberg H: The parathyroid hormone/panthyroid hormone-related peptide receptor coordinates endochoncirai bone developrnent by directly conmlIing chondrocyte differentiation. Proc Nat1 Acad Sci U S A 1998.95: 1303O-l303j.

161. Guo, J. Chung U, Bringhurst R. F. Kronenberg H. M. FaiIure of phospholipase C activation via the m P receptor leads to abnormalities in embryonic bone

development. JJ3one Miner.Res. lS[Suppl 11, S 153-S L52.Xûû. Ref Type: Abstract

163. Karp SJ, Schipani E, Sr Jacques B. HunzeImm 5. Kmnenberg H, McMahon AP: Indian hedgehog coordinates endochondral bone growth and morpho=enesis via parathymid homone related-protein-dependent and -independent pathways. Developrnent 2000, 127543-548.

164. Yoshida E, Noshiro M. Kawamoto T, Tsutsumi S. Kuruta Y, Kato Y: Direct inhibition of indian hedgehog expression by pamhyroid hormone (pth)/pth-related peptide and up-regulation by retinoic acid in growth plate chondroc)..te cultures. Exp Cell Res 200 1,265:64-72.

165. Sims NA, Clement-Lacroix P. Da Ponte F, Bouaii Y, Binm N, Moriggl R, Goffin V, Coschigano K, Gaillard-Kelly M. Kopchick J, Baron R, KeIly PA: Bone homeostasis in growth hormone receptor-nuIl mice is restored by IGF-1 but independent of Stat5. J Clin Invest 2000, LOB: 1095-1 103.

166. Wroblewski J, Edwall-Arvidsson C: Inhibitory effects of basic tïbroblast ~ o w t h factor on chondrocyte differentiation. J Bone Miner Res 1995. 10:735-742.

167. Long F, Schipani E. Asahan H. Kronenberg H. Montminy M: The CREB family of activators is required for endochondral bone development. Development 200 1, 128541-550.

168. Yanz X. Chen L. Xu X. Li C. Huang C. Deng CX: Tgf-betdsrnad3 signais repress chondrocyre hypertrophic differentiation and are required for rnaintaining articular cartilage. J Ce11 Biol1001, 153:3546,

169. Crowe R, Zikherman J, Niswander L: Delta-1 negatively regulates the transition from prehypertrophic to hypertrophic chondrocytes during cartilage formation. Development L999. L 16:987-998.

170. Takeda S, Bonnamy JP, Owen MJ, Ducy P. Kmenty G: Continuous expression of Cbfal in nonhypertrophic chondrocytes uncovers its ability to induce hypertrophic chondrocyre differentiacion and partially rescues Cbfal-deficient mice. Genes Dev 2001, 1546748 1.

171. Wang J, Zhou J, Bondy CA: Igfl promotes longitudinal bone p w t h by insulin-like actions augmenting chondrocyte hypertrophy. FASEB J 1999, 13:1985-1990.

172. Robson H, Siebler T, Stevens DA, ShaIet SM, Williams GR: Thyroid hormone acts directly on growth plate chondrocytes to promote hypertrophic differentiation and inhibit cIanaI expansion and ceIl prolifention. Endocrindogy 3000, 141:3887-3897.

173, Nrtkmishi T, Nishida T, Shimo T, Kobayashi K, Kubo T, Tamatani T, Tezuka K, Tdcigawa M: Effects ofCTGFMcs24, a product of a hypertrophic chondrocyte-

specific gene, on the prolifention and differentiation of chondrocytes in culture. Endocrinology 2000, 14 l:W-??3.

174. De Luca F, Uyeda JA. Mericq V, Mancilla EE. Yanovski JA, Barnes LM, Zile Mi, Baron J: Retinoic acid is a potent regulator of growth plate chondrogenesis. Endocrinology 2000,14 1:346-353.

175. Pacifici M. Golden EB, Iwamoto M. Adams SL: Retinoic acid treatment induces type X collagen gene expression in cuItured chick chondrocytes. Exp Cell Res 199 1. 195:3846.

176. Pathi S, Rutenberg JB. Johnson RL, Vonkamp A: Interaction of Ihh and BMP/Noggin sipaling during cartilage differentintion. Dev Biol 1999.209:239-153.

177. Grimsmd CD, Romano PR, DSouza M. Puzas JE, Reynolds PR. Rosier RN, OKeefe RJ: BMP-6 is an autocrine stimulator of chondrocyte differentiation. J Bone Miner Res 1999, 14475482.

178. Huang W. Chung Ut. Kronenberg KM, de Crombmgghe B: The chondrogenic tnnscription factor Sox9 is a target of sipiiling by the pmthyroid hormone-related peptide in the growth plate of endochondral bones. Proc Nat1 Acad Sci U S A 200 1, 98: 160-165.

179. N i k i K. Colvin JS. Coffin JD. Ornitz DM: Repression of hedgehog sigaling and BMP4 expression in growth place cartilage by fibroblast p w t h factor receptor 3. Development 1998.125:49774988.

180. Colnot C, Sidhu SS. Balmain N. Poirier F: Uncoupling of chondrocyte death and vascular invasion in mouse gdectin 3 nul1 mutant bones. Dev Biol 200 1.229:203-2 14.

18 1. Mo R, Freer AM. Zinyk DL, Crackower MA. Michaud J. Heng HH. Chik KW. Shi LM. Tsui LC. Cheng SH. Joyner AL. Hui C: Specific and redundnnt functions of G1i2 and Gli3 zinc finger genes in skeletril patterning and development. Development 1997. 124: 113-123.

182. OKeefe RJ, Puzas JE. Loveys L. Hicks DG. Rosier RN: Analysis of type II and type X collagen synthesis in cultured growth plate chondrocytes by in situ hybridization: npid induction of type X collagen in culture. J Bone Miner Res 1994,9:17 13-1722.

153. Incardona JP, Gaffield W. Kapur RP. Roelink H: The tentogenic Ventrum alkaloid cyclopamine inhibits sonic hedgehog signal tmsduction. Development 1998, t 25:3553-3562.

184. Taipale J. Chen JK. Cooper MK, Wang B, Mann K. MiIenkovic L, Scott MP, Beachy PA: Effects of oncogenic mutations in Srnoothened and Patched cm be reversed by cyclopamine [see cornrnents]. Nature 1000,400: lûû5-lOû9.

Ericson J, Morton S, Kawakami A, Roelink H, Jessell TM: Two critical periods of Sonic Hedgehog sigaling required for the specification of motor neuron identity. Ce11 1996,87:66 1-673.

Serra R. Karaplis A, Sohn P: Parathyroid hormone-related peptide (E'THrP)-dependent and -independent effects of transforming growth factor beta (TGF-beta) on endochonhl bone formation. J Cell Bi01 1999. 145:783-794.

Schipani E, Langman CB, Parfitt AM, Jensen GS, Kikuchi S, Kooh SW, Cole WG, Juppner H: Constitutively activated receptors for panthyroid hormone and parathyroid hormone-related peptide in Jansen's metaphyseal chondrodysplasia [see comments]. N Engl J Med 1996,335:708-7 14.

Schipani E. Kmse K. Juppner H: A constitutively active mutant PTH-PTHrP receptor in Jansen-type metaphyseal chondrodysplasia. Science 1995,268:98-100.

Jobert AS, Zhang P, Couvineau A, Bonaventure J, Roume J, Le Merrer M. Silve C: Absence of functional receptors for parathyroid hormone and parathyroid hormone- related peptide in Blomstrand chondrodysplasia. J Clin Invest 1998. 102:3440.

Xie LY, Abou-Samra AB: Epitope tag mapping of the extracellular and cytoplasmic domains of the nt parathyroid hormone (PTH)/PTH-related peptide receptor. Endocrinology L998, 139:4563-4567.

Lee C, Gardella TJ, Abou-Samn AB, Nussbaum SR. S e p GV, Potts JT. Kronenberg HM, Juppner H: RoIe of the extracellular regions of the parathyroid hormone (PTH)/PTH-related peptide receptor in hormone binding. Endocrinolo$y 1994, 135: lJ88-l4%.

Piserchio A, Bisello A. Rosenblatt M, Chorev M, Mierke DF: Characterimion of parathyroid hormonelreceptor interactions: structure of the first extracellular Ioop. Biochemisuy 2000.39:S 153-8 160.

Rolz C, Pe!legrini M. Werke DF: Molecular characterization of the receptor-ligand complex for parathyroid hormone. Biochemistry 1999,38:6397-6405.

Pellegrhi M, Bisello A, Rosenblatt M, Chorev M, Mierke DF: Binding domain of human panthyroid hormone receptor: from conformation to function. Biachemistry 1998.37: 12737- 12743.

Karaplis AC, He B, Nguyen MT, Young ID. Semeraro D, Ozawa H, Arnizuka N: Inactivating mutation in the human panthyroid hormone receptor type 1 gene in B lomstrand chondrodysplasia [see comments]. Endocrinology 1998, 1395355-5258-

Bruggeman LA, Xie HX. Brown KS, Yamada Y: Developmental replation For collaeen il gene exuression in uansgenic mice- Teratoloov 1991,35:203-208.

Yamada Y, Miyashita T, Savagner P. Horton W, Brown KS. Abmczuk J, Xie HX, Kohno K, Botander M. Bmggeman L: Regulation of the collagen II gene in vitro and in tmnsgenic mice. Ann N Y Acad Sci 1990. S8O:8 1-7:8 1-87.

Mora CR, Marques L. Silva E, Fonseca M, Pinho M, Torcato M. Barbot J, Medina M, Maroteaux P: Symmetrîcal enchondromatosis of the hands and feet in two sisters. J Pediatr Orthop B 1997.6: L5-19,

Long F, Linsenmayer TF: Regulation of growth region cmiIage proliferation and differentiation by perichondrium. Development 1998. 125: i067-1073.

Hopyan S, Gokgoz N. Bell RS, Andrulis IL, Alman BA, Wunder JS: Expression of osteocalcin and its rranscriptional regulators core-binding factor alpha 1 and MSX2 in osteoid-forming turnours. J Orthop Res 1999, 17:633-638.

Park HL, Bai C, PIatt KA. Matise MP, Beeghly A, Hui CC, Nakashima M. Joyner AL: Mouse Glil mutants are viable but have defects in SHH sigaling in combination with a Gli2 mutation. DeveIopment 2000. 127: 1593-1605.

Stott NS. Chuonj CM: Dual action of sonic hedgehog on chondrocyte hypertrophy: retrovirus mediated ectopic sonic hedgehog expression in Iimb bud micrornass culture induces novel cartilage nodules that are positive for alkaiine phosphatase and type X coIIrigen* J CeII Sci 1997, 110:269 1 -27Oi.

Rocheville M, Lange DC. Kumar U. PateI SC. Patel RC. Patel YC: Receptors for dopamine and somatostatin: fomlttion of hetero-oligorners with enhanced functional activity. Science 7000.288: 154157.

Liu F. Wan Q. Prïstupa ZB. Yu XM. Wang YT. Niznik HB: Direct protein-procein coupling enables cross-talk between dopamine D5 and gamma-minobutyric acid A receptors. Nature 2000, $03274-280.

AbciAlla S. Lother H. Quitterer U: AT1-receptor heterodimers show enhanced G- protein activation and aliered receptor sequestration. Nature 2000,407:94-98.

Lewis MT, Ross S. Strickland PA, Sugnet CW. Jimenez E, Scott W. Daniel CW: Defects in mouse mammary gland development caused by conditionai haploinsufficiency of Patched-L. Development 1999, 126:j 18 1-5 193.

DIugosz A: The Hedgeho and the hair follicle: a growing relationship. J Clin Invest L999. lO4:851-853.

Foley J, Dann P, Hong J, Cospve J, Dreyer B. Rimm D, Dunbar M. PhiIbtick W. Wysolmerski J: Panthymid hormone-related protein maintains rnarnmary epitheIiaI fate and trigers nipple skin differentiation during embryonic breast development. Develoriment 2001, 1285 13-525.

209. Foley J, Longely BJ, Wysolrnerski JJ, Dreyer BE, Broadus AE, Philbrick WM: PTHrP regulates epidermal differentiation in adult rnice. J Invest Dermatol 1998, 1 11: 1 122- 1138.

210. Yoshida E, Noshiro M. Kawarnoto T, Tsutsumi S. Kunita Y. Kato Y: Direct inhibition of indian hedgehog expression by parathyroid hormone (pth)/pth-related peptide and up-regdation by retinoic acid in growth plate chondrocyte cultures. Exp Cell Res 200 L,265:64-72.

21 1. Maye P. Becker S, Kasameyer E, Byrd N, Grabel L: Indian hedgehog sipaling in extraembryonic endoderm and ectoderm differentiation in ES embryoid bodies. Mech Dev 2000.94:117-132.

2 12. Gnbel L, Becker S, Lock L. Maye P, Zanders T: Using EC and ES ceIl culture to study early developrnent: recent observations on Indian hedgeho; and Bmps, Int J Dev Bi01 1998.42:917-925.

213. Cooper MK. Porter JA, Young KE, Beachy PA: Teratogen-rnediated inhibition of target tissue response to Shh signaling [see comments]. Science 1998,180: 1603-1607.

214. Horiuchi N. Holick MF. Potts JT, Rosenblatt M: A parathyroid hormone inhibitor in vivo: design and biological evaluation of a hormone analog. Science 1983.220: 1053- 1055.

215. Lizard-Nacol S. Lizard G. Justrabo E. Turc-Carel C: Immunoiogic characterization of Ewinz's sarcoma using mesenchymal and neural markers. Am J Pathol 1989. 135:847- 855.

216. Rooke JE, Xu T: Positive and negative signals between interacting cells for establishing neural fate. Bioessays 1998,20:209-214,

217. May WA, Arvand A, Thornpson AD. Braun BS, Wright M. Denny a: EWS/FLIl- induced manic fringe renders Nüi 3T3 cells tumorigenic. Nat Genet 1997, 17:495- 497.

218. Ohuchi H. Nnkagawa T. Yamamoto A, Araga A, O h m T. Ishimaru Y, Yoshioka Fi, Kuwana T, Nohno T. Yarnasaki M, Itoh N, Noji S: The mesenchymai factor, FGF10. initiates and maintains the outgrowth of the chick limb bud through interaction with FGF8, an apical ectodermal factor. Development 1997, 1242235-2244.

219. Laufer E. Nelson CE, Johnson RL, Morgan BA, Tabin C: Sonic hedgehog and Fgf4 rict through a signaling cascade and feedback loop to integrate growth and patterning of the developing limb bud. Cell 1994,79:993-1003.

220. Nakchbandi [A, Weir EE, Insoga KL, Philbnck WM, Broadus AE: Pmthyroid hormone-related protein induces spontaneous osteoclast formation via a pancrine cascade. Proc Nat1 Acad Sci U S A 2000,97:7296-7300.

221. Kong YY, Yoshida H, Sarosi 1, Tan HL, Timrns E, Cappmlli C, Morony S, Oliveim- dos-Santos Ai, Van G, Itie A, Khoo W, Wakeham A, Dunstan CR, Lacey DL, Mak TW, Boyle WJ, Penninger SM: OPGL is a key regulator of osteoclastogenesis. lymphocyte development and lymph-node organogenesis. Nature 1999,397:3 12-323.

222. Miquerol L, Langille BL, Nagy A: Embryonic development is dismpted by modest increases in vascular endothelial growth factor gene expression. Development 2000, L27:394 1-3946,

223. Fan CM, Tessier-Lavigne M: Patterning of mammalian somites by surface ectoderm and notochord: evidence for sclerotome induction by a hedgehog homolog. Cell 1994, 79: 1 175- 1 186.

224. Peters H, Wilm B. Sakai N, Imai K, Maas R, Balling R: Pax1 and Pax9 synergistically regulate vertebrai colurnn development. Development 1999, 1265399-5408.

225. Huvos AG: Bone Tumors: Diagnosis, Treatment, and Prognosis. Philadelphia, W.B. Saunders, 1991.

226. Orita M. Iwahana H, Kanazawa H, Hayashi K. Sekiya T: Detection of polymorphisms of human DNA by gel electrophoresis as single-strand conformation poIymorphisrns. Proc Nad Acad Sci U S A 1989,86:1766-2770.

727. Otto F. Thomell AP. Crompton T, Denzel A. Gilmour KC. Rosewell IR. Stamp GW. Beddington RS, Mundlos S. Olsen BR, Selby PB, Owen MJ: Cbfat, a candidate gene for cleidocnnial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 1997,89:765-77 1.

228. Ali NN. Rowe J, Teich NM: Constitutive expression of non-bonellivedkidney alkaline phosphatase in human osteosarcorna cell lines, J Bone iMiner Res 1996, 11:512-520.

229. S e m M, Morini MC. Scotlandi K. Fisher LW, Zini N, Colombo MP, Carnpanricci M. Maraldi NM, Olivari S, Baldini N: Evaluation of osteonectin as a diagnostic marker of osteogenic bone tumors. Hum Pathol 1992,23:1326-133 1.

130. Shafritz AB. Kaplan FS: Differential expression of bone and cartilage related genes in fibrodysplasia ossificans progressiva, myositis ossificans tnurnatica, and osteogenic smoma. Clin Orthop 1998.46-52.

23 1. MirnJMGRH: Fibrous dysplasia, Bone Tumors: Clinical Radiologie. and Patliologic Correlations. Edited by Mirra J'MPPGRH. Philadelphia, Lea & Febiger, 1989, pp. 19L- 226.

232. Bortell R, Owen TA, Shalhoub V, Heinrichs A, Aronow MA, Rochene-Egly C, Lutz Y, Stein IL, Lian IB, Stein GS: Constitutive transcription of the osteocaicin %ne in osteosarcorna cells is reflected by altered protein-DNA interactions a& promoter regulatory elements. Proc Nat1 Acad Sci Li S A 1993,90:2300-2304.

233. Zhou H, Choong P, McCarthy R, Chou ST, Martin TJ, Ng KW: In situ hybrîdization to show sequential expression of osteoblast gene markers during bone formation in vivo. J Bone Miner Res 1994,9: 1489-1499.

234. Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G: Ost?/CbfaL: a transcriptional activator of osteoblast differentiation. Cell 1997,89:747-754.

235. Newberry EP, Boudreaux JM, Towler DA: Stimulus-selective inhibition of rat osteocalcin promoter induction and protein-DNA interactions by the homeodomain repressor Msx2. J Biol Chem 1997,272:29607-296 13.

136. Towler DA, Rutledge SJ, Rodan GA: Msx-YHox 8.1: a transcriptional regulator of the rat osteocalcin promoter. Mol Endocrinol 1994,8: 14841493.

237. Kornori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, Deguchi K, Shimizu Y, Bronson RT, Gao YH, Inada M, Sato M. Okamoto R, Kitamura Y, Yoshiki S, Kishimoto T: Targeted disruption of Cbfal results in a complete Iack of bone formation owing to maturational arrest of osteoblasts. Cell 1997.891755-765.

238. Mundlos S. Otto F, Mundlos C, Mulliken JB, Aylsworth AS, Albright S. Lindhout D, Cole WG, Henn W. Knoll JH, Owen MJ, Menelsmann R, Zabel BU, Olsen BR: Mutations involving the transcription factor CBFA1 cause cleidocnnial dysplasia, CeII 1997,89:773-779,

239. Stewart M. Teny A. Hu M, O'Hara M. Blyth K. Buter E. Cameron E. Onions DE. NeiI JC: Proviral insertions induce the expression of bone-specific isofoms of PEBP2aIphaA (CBFAL): evidence for a new mye collaborating oncogene. Proc N i d Acad Sci U S A 1997,94:8646-565 1.

240, Ladanyi M. Park CK, Lewis R, Jhanwar SC. Healey JH. Huvos AG: Sporadic amplification of the MYC gene in humnn osteosarcomas. D ia s Mol Pachol 1993. 3: 163-167.

241. Hoffmann HM. Catron KM. van Wijnen AJ, McCabe LR, Lian JB. Stein CS, Stein JL: Transcriptional control of the tissue-specific, deveIopmentally regufated osteocalcin gene requires a binding motif for the Msx family of homeodomain proteins. Proc Nat1 Acad Sci U S A 1994.9 1: 12887-1289 1.

142. labs EW, Muller U, Li X. Ma L, Luo W. Haworth IS, Klisak 1. Sparkes R. Warmrin ML, Mulliken JB, .: A mutation in the homeodomain of the human MSXl sene in a family affected with autosomd dominant craniosynostosis. Cell 1993,75:443-450.

243. Singer FR: Fibrous dysplasia of bone: the bone iesion unmasked. Am J PathoI 1997, ljl:l5ll-l5lj.

244. Ducy P. Desbois C, Boyce B. Pinero G, Story B, Dunstan C, Smith E. Bonadio J, GoIdstein S, Gundberg C, Bradley A, Karsenty G: Increased bone formation in osteocaicin-deficient mice. Nature 1996,382:4!8452.

245, Chandar N, CmpbelI P, Novak J, Smith M: Dependence of induction of osreocalcin gene expression on the presence of wild-type p53 in a murinc osteosarcoma ceIl line. Mol Cminog 1993,8299-305.

246. Toguchida J, Yamaguchi T. Dayton SH, Beauchamp RL, Herrera GE. Ishizaki K, Yrimamuro T, Meyers PA, Little J8, Sasaki MS, .: Prevrilence and spectrurn of gemline mutations of the p53 sene among patients with sarcoma. N Engl J Med 1992, 326: 1301-1308.

147. Lee PD, Noble-Tophnm SE. Bell RS, Andrulis IL: Quantitative andysis of multidrug resistance gene expression in hurnan osteosarcornas- Br J Cancer 1996,74: lOJ6-1050.

348- Noonan KE, Beck C, Kolzmayer TA, Chin JE, Wunder JS, AndruIis L. Gazdar M. WilImnn CL, Griffith B, Von Hoff DD, .: Quantitative malysis of MDR1 (muItidrug resistance) gene expression in human tumors by polymerase chah reriction. Proc Nat1 Acrid Sci U S A 1990,87:7 160-7 163.