JBJS | Growth/Differentiation Factor-5 (GDF-5) and Skeletal Development

The Journal of Bone & Joint Surgery, Volume 83, Issue 1 suppl 1

Scientific Article | March 01, 2001

Growth/Differentiation Factor-5 (GDF-5) and Skeletal Development

Paul Buxton, PhD; Christopher Edwards, BSc, MBBS, MRCP; Charles W. Archer, PhD; Philippa Francis-West, PhD

From Department of Craniofacial Development, King's College London, London, United Kingdom; Department of Rheumatology, Tan Tock Seng Hospital, Singapore; and School of Molecular and Medical Biosciences, University of Wales, Cardiff, United Kingdom
Philippa Francis-West, PhD Paul Buxton, PhD Department of Craniofacial Development, King's College London, Guy's Hospital, St. Thomas Street, London, SE1 9RT, U.K.
Christopher Edwards, BSc, MBBS, MRCP Department of Rheumatology, Tan Tock Seng Hospital, 11, Jalan Tock Seng, Singapore
Charles W. Archer, PhD School of Molecular and Medical Biosciences, University of Wales, Cardiff, CF1 3YF, U.K.
In support of their research or preparation of this manuscript, one or more of the authors received grants or outside funding from the Arthritis and Rheumatism Campaign. None of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.

The Journal of Bone & Joint Surgery. 2001; 83:S23-S30

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

Article

Figures

References

text A A A

Abstract

Background: Growth/differentiation factor-5 (GDF-5) has been shown to be essential for normal appendicular skeletal and joint development in humans and mice. In brachypod , a Gdf-5 gene mouse mutant, the defect is first apparent during early chondrogenesis, with the cartilage blastema already reduced in size by E12.5. This defect is associated with changes in the expression of cell surface molecules.

Methods: To understand further how GDF-5 controls cartilage formation, we first mapped the expression of the Gdf-5 gene during skeletal development (please note that the abbreviation for the gene is given in italics and the abbreviation for the protein expressed by the gene is given in capital letters). Subsequently, we over-expressed GDF-5 in the developing chick embryo using a replication competent retrovirus, RCAS(BP). We determined its effects on skeletal development by histological examination and its effects on early growth by autoradiography of proliferating cells. In addition, we examined the effect of GDF-5 on chondrogenic differentiation using micromass and single cell suspension cultures of limb mesenchymal cells.

Results: These studies show that the Gdf-5 gene is expressed in the early cartilage condensation, the perichondrium, and the joint interzone. Over-expression of GDF-5 in chick limb buds, during the condensation stage or later when the skeletal elements have formed, increased the size of the affected elements. In both cases, the increase in size was associated with an increase in cell number and, at later stages, this was correlated with an increase in S-phase cells. In vitro studies showed that GDF-5 could increase cell adhesiveness, and this may be a mechanism through which GDF-5 initiates condensation formation.

Conclusion:These studies show that GDF-5 acts at two stages of skeletal development and by two distinct mechanisms. First, GDF-5 promotes the initial stages of chondrogenesis by promoting cell adhesion, which is consistent with the expression of Gdf-5 in the cartilage condensation. Second, GDF-5 can increase the size of the skeletal elements by increasing proliferation within the epiphyseal cartilage adjacent to its expression within the joint interzone.

Figures in this Article

Introduction

Articular cartilage is particularly vulnerable to damage due to traumatic insult or as a result of diseases such as rheumatoid arthritis and osteoarthritis. Articular cartilage has a limited capacity for repair. Many studies have suggested that the repair of tissue recapitulates the developmental program. For example, following bone fractures the defect can be healed by chondrogenic precursors forming a cartilaginous callus that is replaced by bone, as occurs in normal endochondral bone development. The similarities between endochondral bone development and repair are seen at both a histological and a molecular level ^8,39 . It may be possible, ultimately, to recapitulate joint development with the aim of replacing damaged articular cartilage. Therefore, the identification of factors that control skeletogenesis and, importantly, joint development together with their mechanisms of action is likely to provide new therapeutic strategies for the repair of articular cartilage.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1:: Schematic diagram of the human growth/differentiation factor-5 (GDF-5) protein, indicating known human and mouse mutations. The protein possesses a 19 amino-acid leader sequence (shown in yellow), a 358 amino-acid pro-domain (blue), and a C-terminal mature domain of 124 amino acids (red). Positions refer to the cDNA, unless an amino acid is specified, in which case numbers refer to the full-length polypeptide. The five Bd-C mutations are shown above the protein. The mouse brachypod (bp) mutations and the human CGT and CHTT mutations are shown below the protein. ?G represents a single base deletion and InsG represents a single base insertion, both leading to frameshift mutations; FS denotes a complex frameshift by a deletion or insertion, or both.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2:: Expression of growth/differentiation factor-5 gene ( Gdf-5 ) during chick limb development. A , C , and E : Bright-field and B , D , and F : corresponding dark-field pictures of a stage 25 wing ( A and B ), 30 footplate ( C and D ), 44 foot ( E and F ), distal interphalangeal joint showing the expression of Gdf-5 (visualised by white grains in B , D , and F ). C, developing cartilage and j, developing joint. The perichondrium is marked by an arrowhead in C . Fig. 2 , A-D, are reprinted, with permission, from Francis-West PH, Abdelfattah A, Chen P, Allen C, Parish J, Ladher R, Allen S, MacPherson S, Luyten FP, Archer CW. Mechanisms of GDF-5 action during skeletal development. Development. 1999;126:1305.)

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3:: Expression of bone morphogenetic proteins (Bmps) and their antagonists. A : Bright-field view of an Alcian-blue stained tarsophalangeal joint at stage 44 and B-D : dark-field views of the same joint showing the expression of ( B ) noggin , ( C ) chordin , and ( D ) Bmp-7 . J, joint.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4:: The effect of over-expression of growth/differentiation factor-5 (GDF-5) in vivo . A : Virus was injected or infected cells were grafted into the presumptive humerus region at stage 19/20. B : The extent of retroviral spread (dark staining over the humerus, radius, and ulna) at day 10 of development. Skeletal development was analysed in the control ( C and E ) and GDF-5-infected wings ( D and F ) at 10 days ( C and D ) and stage 26/27 ( E and F ) of development. h, humerus; ha, handplate; r, radius; and u, ulna. The cartilage condensation in E and F is shown by arrows. Scale bar = 1 mm. ( Fig. 4 , B-F, are reprinted, with permission, from Francis-West PH, Abdelfattah A, Chen P, Allen C, Parish J, Ladher R, Allen S, MacPherson S, Luyten FP, Archer CW. Mechanisms of GDF-5 action during skeletal development. Development. 1999;126:1305.)

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5:: The effect of growth/differentiation factor-5 (GDF-5) on in vitro cultures of stage 22/23 limb bud cells. Top lane. Alcian blue staining of micromasses cultured either alone or together with 10 ng/ml of GDF-5. A and B : Phase contrast photographs of single cell suspensions cultured for 18 hours either ( A ) alone or ( B ) together with GDF-5 (10 ng/ml). Large cell aggregates (arrows) are characteristically seen in GDF-5 treated cultures. (The top lane of Fig. 5 is reprinted, with permission, from Francis-West PH, Abdelfattah A, Chen P, Allen C, Parish J, Ladher R, Allen S, MacPherson S, Luyten FP, Archer CW. Mechanisms of GDF-5 action during skeletal development. Development. 1999;126:1305.)

Development of Cartilage Elements and Synovial Joints

Development of the skeletal elements and joints begins with increased cell adhesion in a group of cells within the mesenchyme. These cells form prechondrogenic condensations that go on to produce matrix molecules such as aggrecan and type II collagen. In contrast, cells at the periphery of the condensation become elongated, are non-cartilaginous, and form a boundary, the perichondrium, which delineates the early skeletal elements ³¹ . The perichondrium expresses parathyroid hormone-related peptide (PTHrP), a peptide that prevents terminal differentiation of the epiphyseal chondrocytes in the developing skeleton ¹⁸ and also contributes to appositional growth of the skeleton. Therefore, the perichondrium is an important signalling layer coordinating skeletal growth and differentiation. Later, at the sites of presumptive joints, a flattened zone of cells known as the interzone forms within the cartilage anlagen. The centre of the joint interzone is loose and avascular and will form the joint cavity, while the outer layers, which are continuous with the perichondrium, will form the articular cartilage at the ends of the bones ^11,29 . The joint interzone expresses a number of growth factors, including growth differentiation factor-5 gene ( Gdf-5 ) (please note that the abbreviation for the gene is given in italics and the abbreviation for the protein expressed by the gene is given in capital letters) and Wnt-4^11,14 . As we will discuss later, GDF-5 expression in the joint is likely to control skeletal development by increasing epiphyseal chondrocyte proliferation. Recent evidence has also shown that Wnt-4 can accelerate terminal differentiation of chondrocytes ¹⁴ . This suggests that the developing joint is an important signalling centre controlling chondrocyte differentiation, as is the perichondrium. The chondrocytes within the developing skeleton differentiate into three zones: a proliferative rounded zone at the epiphyses, a flattened zone, and a hypertrophic zone in the diaphysis. The skeletal elements grow by a combination of proliferation, appositional growth from the perichondrium contributing to width, and hypertrophy, which mainly increases the size of the skeletal element along the long axis.

Analysis of the Mouse Gdf-5 Mutant

Insight into the role of GDF-5 during development was initially gained from analysis of the mouse Brachypod mutant, which is a result of mutations in the Gdf-5 gene ^13,33 ( Fig. 1 ). This mutant is characterised, as are the human syndromes (to be discussed later), by a shortened appendicular skeleton; the axial skeleton is relatively unaffected. In the brachypod mouse, the hindlimbs are more severely affected than the forelimbs. The skeletal elements in the manos and pes are the most severely affected, followed by the humerus and femur. In the manos, the digits are splayed, with digits two to five containing two phalanges. The proximal phalange is a relatively long thin rod (when compared with the equivalent phalange of an unaffected littermate) and appears to be equivalent to the basal and middle phalanx. Therefore, the proximal interphalangeal joint in these mice is absent. The pes abnormalities are similar but less severe than those seen in the manos. In addition to the shortening of the skeletal elements, there are also slight abnormalities in the shape. For example, the condyles of the femur and tibia are rudimentary. The fibula may not be joined to the tibia or it may be considerably shorter, resulting in abnormal bending of the tibia. As a result of these defects, together with the condylar abnormalities, the knee may be dislocated ¹³ .

Study of these mice has allowed analysis of the consequences of loss of GDF-5 function during development. Histological studies have shown that the defect is first apparent at E12, just after the formation of the cartilage precondensation and before the formation of the joint ^4,13,21 . The prechondrocytes are surrounded by less matrix and, secondarily to this, the perichondrium does not form ²¹ . Furthermore, development of the skeletal elements appears to be delayed ^13,21 . With the exception of the "fused phalanges," which remain unossified, differentiation of all the skeletal elements eventually occurs and the elements ossify normally. Other analyses have supported these histological observations, which have shown that there is a developmental delay in hyaluronan synthesis and changes in the expression of cell surface molecules such as those with galactosyl transferase activity and ConA reactivity ^7,15,16,32 .

Defects in Humans

Genomic mutations in the human Gdf-5 gene have been shown to cause three related chondrogenic dysplasias: chondrodysplasia Grebe type (CGT) and acromesomelic chondrodysplasia Hunter-Thompson type (CHTT), which are autosomal recessive disorders, and brachydactyly type C, which is autosomal dominant ^30,35,36 . The most striking feature of these disorders is the restriction of defects to the limbs and, within the limbs, the increasing severity toward the distal tip. Grebe type chondrodysplasia (CGT) and acromesomelic chondrodysplasia Hunter-Thompson type (CHTT) are similar, characterised by extremely short limbs (forearms and hands relatively shorter than the upper arms and likewise for the thighs and shanks), multiple dislocations of the joints, and loss of articulation of the digits. CGT is more severe and brachydactyly has been described in heterozygotes ³⁵ . Brachydactyly type-C again affects the distal limbs, with the second, third, and fifth middle phalanges being shorter in heterozygotes. The precise mutations in GDF-5 that cause these defects have been identified in a number of kindreds and will be described in detail.

CHTT has been shown to be caused by a 22-bp tandem duplication that results in addition of 43 aberrant amino acids followed by a stop codon ³⁶ ( Fig. 1 ). This mutation arises in the mature region of the polypeptide in the middle of the cysteine knot, which is predicted to create a null phenotype. In this family, haploinsufficiency does not appear to give a phenotype but, as described previously, homozygote individuals are severely affected.

A number of geographically related families with CGT have been found to have another mutation, Cys400Tyr, in GDF-5 ³⁵ ( Fig. 1 ). This mutation affects the first of the seven conserved cysteines of the cysteine knot. In vitro assays have shown that this mutation can abolish the secretion of GDF-5, together with other bone morphogenetic proteins (BMPs), suggesting that GDF-5 may form heterodimers in vivo . This idea is supported by the more severe phenotype of CGT, which may be due to the loss of function of other BMPs in addition to GDF-5. As this mutation acts in a dominant-negative manner, heterozygotes are also affected, but with the much milder defects of brachydactyly.

Five other distinct mutations of GDF-5 have been identified in five separate families with the autosomal dominant brachydactyly type-C. Of these mutations, three result in frameshifts, due to deletion or insertion of a nucleotide, and one mutation introduces a termination codon N-terminal to the mature domain. Another mutation results in an Arg438Cys substitution in the cysteine knot region of the mature moiety, and this is associated with a more severe phenotype ³⁰ ( Fig. 1 ).

Thus, these mutations show that GDF-5 is essential for normal appendicular skeletal development and that some skeletal elements are more sensitive to the loss of GDF-5 function, i.e., the digits can be affected in heterozygotes. This may be because in other regions of the limb, different BMPs/GDFs are able to compensate for the loss of GDF-5.

Aims

In light of the data showing that GDF-5 has an important physiological role during appendicular skeletal and joint development, we aimed to further determine the role and mechanism of action of GDF-5 using the developing chick embryo as a model system to study the effect of GDF-5 on skeletogenesis. This experimental model has been particularly powerful for elucidating the role of growth factors during skeletogenesis. For example, the role of Indian Hedgehog in controlling chondrocyte differentiation by the expression of PTHrP in the perichondrium was identified with use of this system ³⁸ .

Materials and Methods

In Situ Analysis

Gene expression was determined by in situ hybridisation to tissue sections as previously described ¹⁰ . The Bmp-7 , Gdf-5 , noggin , and chordin probes are as described ^10-12 .

Over-expression Studies

Early chick limb buds were infected at different stages in ovo with a replication competent retrovirus, RCAS(BP), which expresses mouse GDF-5 (RCAS-GDF-5). The virus was introduced by grafting cells infected with the RCAS-GDF-5 retrovirus or by injecting concentrated viral stocks as described ¹⁰ ( Fig. 2 -A). Control embryos were infected with a RCAS(BP) retrovirus encoding alkaline phosphatase. The embryos were allowed to develop until various time points and were fixed for histological or proliferation analysis.

Micromass Studies

Stage 22/23 limb mesenchymal cells were plated at high density in the presence or absence of recombinant GDF-5 (10 ng/ml) and were cultured for 1 to 3 days as described ³ . The number of chondrogenic nodules was counted and matrix production was assessed by Alcian blue binding to glycosaminoglycans ¹⁰ .

Cell Adhesion Studies

Stage 22/23 forelimbs and hindlimbs were disaggregated to single cell suspensions by trypsinisation and passage through Nitex filters. Single cell suspensions were then roller cultured in the presence or absence of GDF-5 (10 ng/ml) for 18 hours ¹⁰ . The number and size of the cell aggregates were then determined.

Proliferation Analysis

Stage 19/20 limb buds were infected with the GDF-5 or control retrovirus. Cells in S-phase of the cell cycle were labelled at stage 27/28 for 8 hours with tritiated thymidine. The embryos were subsequently fixed and sectioned for autoradiographic analysis. The percentage of proliferating cells was compared between injected and contralateral, control limbs.

Results

The Expression of Gdf-5

In situ studies revealed that Gdf-5 is expressed in the early cartilage condensation, developing joint interzone, and, briefly, in the perichondrial layer ( Fig. 2 ; see also ^10,23,34 ). As the joint cavitates, Gdf-5 expression is down-regulated ¹⁰ (data not shown). Similar expression patterns have been reported in humans and mice ^2,33 . In addition, because other studies have suggested that GDF-5 may heterodimerise with other members of the BMP family ³⁵ , we also analysed the expression of Bmps , together with their antagonists, noggin and chordin . These studies revealed that Bmp-2 , 4 , and 7 are expressed in the joint interzone ( Fig. 3-D ; data not shown; see also ^11,22 ). Noggin is initially expressed in the developing cartilage anlagen, including the developing joint (data not shown and see ^11,28 ). Later, it is expressed in a narrow domain in the epiphyseal layer of chondrocytes a few cell layers from the articular surface ( Fig. 3-B ; also see ^11,28 ). Chordin is also expressed in the developing joint ( Fig. 3-C ; also see ^11,28 ).

Over-expression of GDF-5

To investigate how GDF-5 controls skeletal development, we over-expressed GDF-5 in the developing chick limb using a replication competent retrovirus, RCAS(BP). This virus will spread through the developing limb bud, resulting in misexpression of GDF-5 around and in the developing skeletal elements ( Fig. 4-A and Fig. 4-B ). The embryos infected with GDF-5 were allowed to develop for 10 days and were analysed for changes in skeletal development. This, together with subsequent histological analysis, showed that GDF-5 over-expression increased the size of the skeletal elements, due to an increase in the number of chondrocytes ( Fig. 4-C and Fig. 4-D ; data not shown). The arrangement of the chondrocytes within the skeletal element itself appeared to be normal, i.e., there was a hypertrophic, rounded, and flattened layer of cells (data not shown; see also ¹⁰ ). The increase in size as a result of GDF-5 expression was first apparent at stage 26, i.e., during the early precartilage condensation ( Fig. 4-E and Fig. 4-F ). Injecting the RCAS-GDF-5 virus in the developing limbs, once the skeletal elements and joints had formed, also increased the size of the skeletal elements (data not shown; see also ¹⁰ ). Therefore, these studies show that GDF-5 can increase the size of the developing skeletal elements at two distinct stages: during condensation and once the appendicular skeleton has formed.

GDF-5 can Initiate Chondrogenesis

The over-expression studies in ovo showed that GDF-5 can increase the size of the early cartilage condensation, suggesting that GDF-5 may control the early stages of chondrogenesis. To further investigate, we used chick micromass cultures, which are a model system to study chondrogenesis. Micromasses cultured in the presence of GDF-5 (10 ng/ml) formed 30% more nodules than control cultures (n = 4; 370 24, control; 481 25, GDF-5 treated; Student's t test, p = 0.01). This shows that GDF-5 can initiate chondrogenesis. The nodules were also larger after 3 days of culture, and this increase in chondrogenesis occurs in a dose-dependent fashion ( Fig. 5 ; data not shown; see also ¹⁰ ).

GDF-5 can Increase Cell Adhesion

GDF-5 may initiate chondrogenesis by increasing cell proliferation or cell adhesion, or both. Because studies of brachypod mice have indicated that changes in the expression of cell surface molecules may be responsible, at least in part, for the defect, we tested whether GDF-5 could alter cell adhesion. Thus, single cell limb mesenchymal cells were cultured in the presence or absence of GDF-5 (10 ng/ml). After 18 hours of culture, we detected a 2.3-fold increase in the number of aggregates (>100 cells) in GDF-5 treated cultures compared with the controls ( Fig. 5 ).

GDF-5 can also Increase Proliferation of Chondrocytes In Vivo

Developing limbs were infected with the GDF-5 retrovirus at stages 17-20 and were then pulsed for 8 hours with tritiated thymidine at stages 27/28 to label proliferating cells. Fixed and sectioned embryos were analysed by autoradiography, which revealed that GDF-5 infection resulted in a 53.6% increase in cells in S-phase. There was no significant difference in proliferation between control retroviral-infected and uninfected humeri.

Discussion

These studies, together with other analyses, have shown that GDF-5 is an important modulator of chondrogenesis and that levels of active GDF-5 may control the size of the skeletal elements. These over-expression studies are similar to those performed in the developing mouse ³⁷ . In the latter studies, GDF-5 was expressed under the control of a collagen type II promoter such that it was over-expressed in the early developing condensation at E12.5 and later in resting and proliferating chondrocytes. In these mouse studies, over-expression of GDF-5 increased the size of the skeletal elements in only two animals. Typically, the skeletal elements were smaller and were characterised by fusion and bone differentiation across the joints. In the latter group of animals, histological and in situ analysis revealed that the hypertrophic zone was larger and the proliferative layer was smaller, suggesting that GDF-5, when misexpressed in the epiphyseal region, accelerates chondrocyte differentiation. The ability of GDF-5 to promote chondrocyte differentiation may not reflect its true physiological function, as GDF-5 is not expressed near or in the hypertrophic zone. However, the closely related protein GDF-6 is expressed in the hypertrophs in humans, and it is possible that GDF-6 promotes terminal differentiation in vivo² .

The discrepancy between the chick and mouse studies may be due to the methods used to over-express GDF-5. In the mouse, GDF-5 was misexpressed in the developing skeletal elements themselves. In contrast, whereas the retrovirus spread throughout the chick limb, in situ studies showed that actual infection of the chondrocytes was variable and limited. In general, GDF-5 was misexpressed in the mesenchyme around the developing skeletal elements and not in the actual prechondrocytes/chondrocytes themselves (unpublished observations). However, consistent with the chick data presented here, over-expression of GDF-5 in the mouse increased the size of the precartilage condensation due to an increase in cell number.

The chick studies presented here have shown that GDF-5 can initiate chondrogenesis. GDF-5 has also been shown to initiate chondrogenesis in the ATDC5 cell line, which undergoes progressive chondrogenesis in the presence of insulin, and in rat limb bud cells ^17,24 . Furthermore, in vitro studies of brachypod mesenchymal cells have shown that these cells can form the initial aggregates of the chondrogenic pathway but are unable to form nodules ^5,27 . This indicates that the defective step in chondrogenesis in the brachypod mutant is not the first stage of initiation but occurs very shortly after, indicating a physiological role for GDF-5 in this process.

Mechanism of Action

In vitro and in vivo studies have indicated that GDF-5 may act by two mechanisms-increasing cell proliferation and intercellular adhesion-both of which are important modulators of chondrogenesis. In vitro studies of brachypod limb cells have shown that they have different cell surface properties from wild-type cells ^4-7,15 . For example, mutant cells aggregate more rapidly than wild-type cells when cultured in vitro and display more affinity for each other than for wild-type cells ^4,5 . These studies suggest that brachypod cells are more adhesive than wild-type cells at equivalent stages. However, because brachypod cells are developmentally delayed compared with wild-type cells, this difference in cell adhesivity may just reflect the developmental delay. Furthermore, our in vitro studies using chick mesenchymal cells suggest that GDF-5 actually increases cell adhesion. Single cell suspensions form large aggregates when cultured in the presence, but not in the absence, of GDF-5. This shows that a subpopulation of cells (presumably those expressing the GDF-5 receptor ALK6) can respond to GDF-5 signalling by increasing cell interactions. The targets of GDF-5 expression, i.e., what cell surface proteins it induces/modulates, are currently unknown, but they are likely to be important modulators of chondrogenesis (e.g., NCAM ⁴⁰ ).

In addition to controlling cell adhesion, GDF-5 can increase proliferation. In the transgenic mouse, GDF-5 increased cell proliferation around the condensation and later in the perichondrial cells themselves ³⁷ , whereas the chick over-expression studies revealed that GDF-5 can increase proliferation of chondrocytes ¹⁰ . In contrast to the mouse studies, which detected proliferation histologically by silver staining for nucleolar organizer regions, we did not detect an increase in proliferation in the chick micromass studies by pulse-labelling experiments ¹⁰ . However, this does not rule out a possible proliferative effect of GDF-5 on the prechondrocytes/chondrocytes in the chick micromasses. GDF-5 may induce a transient increase in proliferation that was not revealed by the time points analysed. For example, the studies that will be described show that some of the effects of GDF-5 signalling are very transient. Alternatively, the effect of GDF-5 on the proliferation of chondrocyte precursors may have been masked by its effect on proliferation of other cells in the culture.

The chondrogenic ATDC5 cell line has been used to dissect the intracellular signalling pathways of GDF-5. In this cell line, GDF-5 promotes initiation of chondrogenesis, Alcian blue matrix staining, and alkaline phosphatase activity ²⁴ . The latter two are markers of terminal differentiation. This cell line has been used to show that GDF-5 activates two MAP kinases, p38 and ERK1/2. Activation of the former is slow and sustained, whereas activation of the latter is transient, lasting for 5 minutes. Inhibition studies revealed that the p38 activation is required for the ability of GDF-5 to enhance nodule formation and subsequent collagen type II expression. However, inhibition of p38 did not block the ability of GDF-5 to stimulate alkaline phosphatase activity. Analysis and identification of these pathways is important to understand how growth factors such as GDF-5 mediate their pleiotropic effects. Thus, it may ultimately be possible to block or stimulate specific pathways, promoting "desirable" effects of GDF-5 while blocking "undesirable" effects for therapeutic purposes.

Fitting This all Together

The Role of GDF-5 during Skeletal Development

The combined functional and in situ analyses suggest that GDF-5 acts at the early steps of chondrogenesis to promote condensation. This is consistent with the early developmental defect seen in Brachypod mice and the ability of GDF-5 to initiate chondrogenesis. Combining the mouse and chick data, GDF-5 would appear to induce chondrogenesis by increasing cell adhesion and proliferation and possibly by promoting recruitment of mesenchymal cells into the developing cartilage element. Later, GDF-5 in the perichondrium may control proliferation of the adjacent chondrocytes and the perichondrial cells themselves. In addition, GDF-5 may induce recruitment of the perichondrial cells into the chondrogenic lineage. Finally, GDF-5 in the joint interzone may signal to the chondrocytes in the adjacent epiphyses to increase their proliferation.

These mechanisms are also consistent with the expression of ALK6 in the early cartilage condensation and later in the epiphyseal chondrocytes ^19,42 . Furthermore, over-expression of the constitutively activated ALK6 receptor can initiate chondrogenesis ⁴² , suggesting that ALK6 may be the receptor that mediates GDF-5 signalling during early chondrogenesis. This is supported by recent gene-inactivation studies in mice. Loss of ALK6 function results in abnormal chondrogenesis of the digits. These defects include loss of phalanges one and two in all digits or their replacement by a single rudimentary element, together with elbow dislocations ^1,41 . Histological and molecular analyses have shown that the initial condensation forms but fails to mature. This phenotype is reminiscent of the GDF-5 mutant. However, in contrast to the brachypod mutant, loss of ALK6 does not affect the length of the skeletal elements, suggesting that GDF-5 controls skeletal growth by way of one of its other receptors, such as Bmpr1A or ActRI ²⁶ .

The Role of GDF-5 in the Developing Joint

The loss of joints in humans and mice as a result of mutations in GDF-5, combined with its expression in the joint interzone, has led to the suggestion that GDF-5 may specify joint development. However, at present the role of GDF-5 during joint development is unclear. The loss of joints may be a secondary consequence of the initial chondrogenic defect. Indeed, the BMP antagonists chordin and noggin are expressed in developing joints, and they may be able to bind GDF-5 and block its activity. Alternatively, GDF-5 may control cell death in the developing joint or hyaluronan synthesis and CD44 expression, as does BMP-7 in articular chondrocytes ²⁵ . Evidence in support of the former possibility has been obtained from bmpr1b knockout mice, which show increased/sustained Gdf-5 expression correlating with apoptosis of the prechondrogenic cells ¹ . GDF-5, either alone or with other BMPs, may induce programmed cell death, although clearly by an alternative BMP receptor.

Future Perspectives

GDF-5 is an important physiological mediator of appendicular skeletal development. It has an activity distinct from BMPs, which are currently being used in therapeutic trials. Dissection of their distinct effects, mechanisms of action, and intracellular pathways utilised will afford insight into how these growth factors mediate their pleiotropic effects. Thus, the long-term goal is to elucidate these signalling cascades and mechanisms, with the aim of maximising repair of articular cartilage.

Note: We thank Adel Abdelfattah for technical help; the Genetics Institute for the gift of recombinant GDF-5; David Kingsley, Frank P. Luyten, Randy Johnson, and Tom Jessel for the gift of cDNA probes; and Andrew Yeudall for critical reading of the manuscript. This work was funded by the Arthritis and Rheumatism Campaign.

References

BaurST, Mai JJ,Dymecki SM. Combinatorial signaling through BMP receptor IB and GDF5: shaping of the distal mouse limb and the genetics of distal limb diversity. Development,2000;127: 605-S19. 127605 2000 [PubMed]

ChangSC, Hoang B, Thomas JT, Vukicevic S, Luyten FP, Ryba NJ, Kozak CA, Reddi AH,Moos M Jr. Cartilage-derived morphogenetic proteins: new members of the transforming growth factor-beta superfamily predominantly expressed in long bones during human embryonic development. J Biol Chem,1994;269: 28227-S34. 26928227 1994 [PubMed]

CottrillCP, Archer CW,Wolpert L. Cell sorting and chondrogenic aggregate formation in micromass culture. Dev Biol,1987;122: 503-S15. 122503 1987 [PubMed]

DukeJ,Elmer WA. Effect of the brachypod mutation on early stages of chondrogenesis in mouse embryonic hind limbs: an ultrastructural analysis. Teratology,1979;19: 367-S75. 19367 1979 [PubMed]

DukeJ,Elmer WA. Effect of the brachypod mutation on cell adhesion and chondrogenesis in aggregrates of mouse limb mesenchyme. J Embryol Exp Morph,1977;42: 209-S17. 42209 1977

ElmerWA, Pennybacker MF, Knudsen TB,Kwasigroch TE. Alterations in cell surface galactosyltransferase activity during limb chondrogenesis in brachypod mutant mouse embryo. Teratology,1988;38: 475-S84. 38475 1988 [PubMed]

ElmerWA,Wright JT. Changes in plasma membrane proteins and glycoproteins during normal and brachypod mouse limb development. Prog Clin Biol Res,1983;110 Pt A: 355-S64. 110 Pt A355 1983 [PubMed]

FergusonC, Alpern E, Miclau T,Helms JA. Does adult fracture repair recapitulate embryonic skeletal formation? . Mech Dev,1999;87: 57-S66. 8757 1999 [PubMed]

FrancisPH, Richardson MK, Brickell PM,Tickle C. Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development,1994;120: 209-S18. 120209 1994 [PubMed]

Francis-West PH, Abdelfattah A, Chen P, Allen C, Parish J, Ladher R, Allen S, MacPherson S, Luyten FP,Archer CW. Mechanisms of GDF-5 action during skeletal development. Development,1999;126: 1305-S15. 1261305 1999 [PubMed]

Francis-West PH, Parish J, Lee K,Archer CW. BMP/GDF-signalling interactions during synovial joint development. Cell Tissue Res,1999;296: 111-S19. 296111 1999 [PubMed]

Francis-West PH, Robertson KE, Ede DA, Rodriguez C, Izpisua-Belmonte JC, Houston B, Burt DW, Gribbin C, Brickell PM,Tickle C. Expression of genes encoding bone morphogenetic proteins and sonic hedgehog in talpid (ta3) limb buds: their relationships in the signalling cascade involved in limb patterning. Dev Dyn,1995;203: 187-S97. 203187 1995 [PubMed]

Grüneberg H,Lee AJ. The anatomy and development of brachypodism in the mouse. J Embryol Exp Morphol,1973;30: 119-S41. 30119 1973 [PubMed]

HartmannC,Tabin CJ. Dual roles of Wnt signaling during chondrogenesis in the chicken limb. Development,2000;127: 3141-S59. 1273141 2000 [PubMed]

HewittAT,Elmer WA. Reactivity of normal and brachypod mouse limb mesenchymal cells with con A. Nature,1976;264: 177-S8. 264177 1976 [PubMed]

HewittAT,Elmer WA. Developmental modulation of lectin-binding sites on the surface membranes of normal and brachypod mouse limb mesenchymal cells. Differentiation,1978;10: 31-S8. 1031 1978 [PubMed]

Hötten GC, Matsumoto T, Kimura M, Bechtold RF, Kron R, Ohara T, Tanaka H, Satoh Y, Okazaki M, Shirai T, Pan H, Kawai S, Pohl JS,Kudo A. Recombinant human growth differentiation factor 5 stimulates mesenchyme aggregation and chondrogenesis responsible for the skeletal development of limbs. Growth Factors,1996;13: 65-S74. 1365 1996 [PubMed]

IwamotoM, Enomoto-Iwamoto M,Kurisu K. Actions of hedgehog proteins on skeletal cells. Crit Rev Oral Biol Med,1999;10: 477-S86. 10477 1999 [PubMed]

KawakamiY, Ishikawa T, Shimabara M, Tanda N, Enomoto-Iwamoto M, Iwamoto M, Kuwana T, Ueki A, Noji S,Nohno T. BMP signaling during bone pattern determination in the developing limb. Development,1996;122: 3557-S66. 1223557 1996 [PubMed]

KimuraS,Shiota K. Sequential changes of programmed cell death in developing fetal mouse limbs and its possible roles in limb morphogenesis. J Morphol,1996;229: 337-S46. 229337 1996 [PubMed]

KwasigrochTE, Curtis SK, Knudsen TB, Barrach H-J,Elmer WA. Morphological analysis of abnormal digital chondrogenesis in the brachypod (bp ^H ) mouse limb in organ culture. Anat Embryol,1992;185: 307-S15. 185307 1992 [PubMed]

MaciasD, Ganan Y, Sampath TK, Piedra ME, Ros MA,Hurle JM. Role of BMP-2 and OP-1 (BMP-7) in programmed cell death and skeletogenesis during chick limb development. Development,1997;124: 1109-S17. 1241109 1997 [PubMed]

MerinoR, Macias D, Ganan Y, Economides AN, Wang X, Wu Q, Stahl N, Sampath KT, Varona P,Hurle JM. Expression and function of Gdf-5 during digit skeletogenesis in the embryonic chick leg bud. Dev Biol,1999;206: 33-S45. 20633 1999 [PubMed]

NakamuraK, Shirai T, Morishita S, Uchida S, Saeki-Miura K,Makishima F. p38 mitogen-activated protein kinase functionally contributes to chondrogenesis induced by growth/differentiation factor-5 in ATDC5 cells. Exp Cell Res,1999;250: 351-S63. 250351 1999 [PubMed]

NishidaY, Knudson CB, Eger W, Kuettner KE,Knudson W. Osteogenic protein 1 stimulates cell-associated matrix assembly by normal human articular chondrocytes: up-regulation of hyaluronan synthase, CD44, and aggrecan. Arthritis Rheum,2000;43: 206-S14. 43206 2000 [PubMed]

NishitohH, Ichijo H, Kimura M, Matsumoto T, Makishima F, Yamaguchi A, Yamashita H, Enomoto S,Miyazono K. Identification of type I and type II serine/threonine kinase receptors for growth/differentiation factor-5. J Biol Chem,1996;271: 21345-S52. 27121345 1996 [PubMed]

OwensEM,Solursh M. Cell-cell interaction by mouse limb cells during in vitro chondrogenesis: analysis of the brachypod mutation. Dev Biol,1982;91: 376-S88. 91376 1982 [PubMed]

PathiS, Rutenberg JB, Johnson RL,Vortkamp A. Interaction of Ihh and BMP/Noggin signaling during cartilage differentiation. Dev Biol,1999;209: 239-S53. 209239 1999 [PubMed]

PitsillidesAA, Skerry TM,Edwards JC. Joint immobilization reduces synovial fluid hyaluronan concentration and is accompanied by changes in the synovial intimal cell populations. Rheumatology (Oxford),1999;38: 1108-S12. 381108 1999 [PubMed]

PolinkovskyA, Robin NH, Thomas JT, Irons M, Lynn A, Goodman FR, Reardon W, Kant SG, Brunner HG, van der Burgt I, Chitayat D, McGaughran J, Donnai D, Luyten FP,Warman ML. Mutations in CDMP1 cause autosomal dominant brachydactyly type C. Nat Genet,1997;17: 18-S19. 1718 1997 [PubMed]

Rooney P, Archer CW, Wolpert L Morphogenesis of cartilaginous long bone rudiments. In The Role of Extracellular Matrix in Development , edited by R. Treslstad, pp 305-322. New York, Wiley Liss, 1984.

ShambaughJ,Elmer WA. Analysis of glycosaminoglycans during chondrogenesis of normal and brachypod mouse limb mesenchyme. J Embryol Exp Morphol,1980;56: 225-S38. 56225 1980 [PubMed]

StormEE, Huynh TV, Copeland NG, Jenkins NA, Kingsley DM,Lee SJ. Limb alterations in brachypodism mice due to mutations in a new member of the TGF beta-superfamily. Nature,1994;368: 639-S43. 368639 1994 [PubMed]

StormEE,Kingsley DM. GDF5 coordinates bone and joint formation during digit development. Dev Biol,1999;209: 11-S27. 20911 1999 [PubMed]

ThomasJT, Kilpatrick MW, Lin K, Erlacher L, Lembessis P, Costa T, Tsipouras P,Luyten FP. Disruption of human limb morphogenesis by a dominant negative mutation in CDMP1. Nat Genet,1997;17: 58-S64. 1758 1997 [PubMed]

ThomasJT, Lin K, Nandedkar M, Camargo M, Cervenka J,Luyte FP. A human chondrodysplasia due to a mutation in a TGF-beta superfamily member. Nat Genet,1996;12: 315-S17. 12315 1996 [PubMed]

TsumakiN, Tanaka K, Arikawa-Hirasawa E, Nakase T, Kimura T, Thomas JT, Ochi T, Luyten FP,Yamada Y. Role of CDMP-1 in skeletal morphogenesis: promotion of mesenchymal cell recruitment and chondrocyte differentiation. J Cell Biol,1999;144: 161-S73. 144161 1999 [PubMed]

VortkampA, Lee K, Lanske B, Segre GV, Kronenberg HM,Tabin CJ. Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science,1996;273: 613-S22. 273613 1996 [PubMed]

VortkampA, Pathi S, Peretti GM, Caruso EM, Zaleske DJ,Tabin CJ. Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev,1998;71: 65-S76. 7165 1998 [PubMed]

WidelitzRB, Jiang TX, Murray BA,Chuong CM. Adhesion molecules in skeletogenesis: II. Neural cell adhesion molecules mediate precartilaginous mesenchymal condensations and enhance chondrogenesis. J Cell Physiol,1993;156: 399-S411. 156399 1993 [PubMed]

YiSE, Daluiski A, Pederson R, Rosen V,Lyons KM. The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development,2000;127: 621-S30. 127621 2000 [PubMed]

ZouH, Wieser R, Massague J,Niswander L. Distinct roles of type I bone morphogenetic protein receptors in the formation and differentiation of cartilage. Genes Dev,1997;11: 2191-S203. 112191 1997 [PubMed]

Topics