RAPID REVIEW
Rapid review
CCN proteins: multifunctional signalling regulators
Bernard Perbal
Context Although little is known as yet about the processes that coordinate cell-signalling pathways, matrix proteins are probably major players in this type of global control. The CCN (cyr61, ctgf, nov) proteins are an important family of matricellular regulatory factors involved in internal and external cell signalling. This family participates in angiogenesis, chondrogenesis, and osteogenesis, and they are probably involved in the control of cell proliferation and differentiation. Starting point Runping Gao and David Brigstock (Hepatol Res 2003; 27: 214–20) recently showed that CCN2 (CTGF, connective tissue growth factor) is a cell-adhesion factor for hepatic stellate cells. On exposure to transforming growth factor , hepatic stellate cells produce distinct CCN2 isoforms. Gao and Brigstock assign to CCN2 module 3 the capacity to mediate binding to low-density-lipoprotein receptorrelated protein (LRP), which was previously reported to interact with CCN2 and to be involved in various types of signalling. They also establish that CCN2 binding to LRP is heparindependent and that module 4 of CCN2 promotes LRPindependent adhesion of hepatic stellate cells. The differential binding of CCN2 isoforms to LRP highlights the importance of functional interactions between individual modules, and reinforces the concept that different module combinations might confer agonistic or antagonistic activities. Where next? It is essential to understand how the distinct configuration of the various CCN isoform affects their biological activities and bioavailability, and to explore the mechanisms and the regulatory processes involved in the production of truncated CCN isoforms. A better understanding of the structural basis for their multifunctionality is a prerequisite to wider use of CCN proteins in molecular medicine.
The CCN (cyr61, ctgf, nov) family of proteins contains six members in humans (figure 1). The CCN proteins contain up to four modules that resemble functional domains previously identified in major regulatory proteins. Each CCN protein is now numbered by their order of discovery.1 CCN1 and CCN2 were the first to be discovered, with increased expression on stimulation of cell growth.2,3 The gene encoding the chicken CCN1 (CEF10) was induced in v-src-transformed chicken-embryo fibroblasts, and the gene encoding the murine CCN2 (FISP12) had been identified as a secreted protein expressed in serum-induced NIH3T3 fibroblasts.2,3 Human CCN2 was isolated with an antiserum Lancet 2004; 363: 62–64 Laboratoire d’Oncologie Virale et Moléculaire, UFR de Biochimie, Université Paris 7-D Diderot, 75005 Paris, France (Prof B Perbal PhD) (e-mail:
[email protected])
62
directed against platelet-derived growth factor, and as a chondrocyte-specific gene product.2,3 The discovery of ccn3 provided the first evidence for aberrant expression in tumours of a CCN protein with antiproliferative activity.3 CCN4 (ELM1) and CCN5 (RCOP1) also showed antiproliferative activities. Their human orthologues were identified later as Wnt-induced secreted proteins (WISP) with an additional member (CCN6/WISP3). Except for CCN5, which lacks the C-terminal module, the other CCN proteins contain the four structural modules originally described in CCN1–3. Variant proteins lacking one or more module are also produced in normal and pathological conditions (see below and figure 2). Their high degree of homology suggested that CCN proteins might have similar or redundant functions, a view which has been challenged by recent results.4 The presence of four potentially functional domains raised fundamental questions about their contribution to the biological properties of the CCN proteins. The current view is that each of the four modules acts both independently and interdependently.4 Module 3 of CCN1 binds integrin61.5 Runping Gao and David Brigstock6 recently showed that module 3 of CCN2 binds low-density-lipoprotein receptor-related protein (LRP). In CCN2, module 4 binds heparin7 whereas it binds integrin alphaMbeta2 in CCN1.8 In CCN3 module 4 binds fibulin 1C9and Notch1.10
Sites of expression Early studies showed ccn gene expression in a wide variety of tissues during normal development and in various species.2,3 The amounts of RNA species appeared to be controlled by tight spatiotemporal regulation. High concentrations of ccn1 and ccn2 RNA were detected in most tissues that stained positive, whereas low to high levels of ccn3 RNA were generally present in most positive tissues.3 ccn5 and ccn6 have a more restricted expression pattern.4 Because the CCN proteins contain a signal peptide driving their secretion, a central question is to determine whether they act at their site of production or can be transported away to execute their functions elsewhere. The CCN proteins and corresponding RNA were detected in tissues originating from the three germ layers, with major sites of expression suggesting these proteins have a role in the differentiation and functioning of nervous system, vasculature, muscle, and bone.2,3 In some cases, cells positive for ccn3 RNA did not stain positive for CCN3 protein. In others, cells strongly positive for CCN3 protein expressed barely detectable ccn3 RNA,3 which suggests that the accumulation of CCN3 protein at specific sites might be related to biological function. That CCN3 modulates calcium and sodium ion-chanelling supports this hypothesis.11 CCN proteins have also been detected in normal biological fluids.2,3,12 Although the role of these circulating proteins remains obscure, their presence suggests functions away from their production site. Findings in mesangial cells suggest a paracrine role for CCN2.13
THE LANCET • Vol 363 • January 3, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.
RAPID REVIEW
e1
e2
SP
IGFBP
e4
e3
VWC
e5
TSP1
H
CT
Figure 1: CCN family of proteins Prototypic multimodular structure of CCN proteins. Five exons encode signal peptide (SP) and IGBFP, VWC, TSP1, and CT modules. At junction between domain II and III in CCN1, long peptidic stretch was proposed to act as hinge (H). The great variability of primary sequences between domains II and III in CCN proteins suggest this region may not be acting in similar manner in all CCN proteins.
How many CCN proteins? Originally regarded as a simple group of related proteins, the CCN family is now known to be complex (figure 2), with biologically active CCN isoforms generated either by posttranslational processing or alternative splicing. Detection in pathological conditions suggested that these variant proteins have critical biological functions.14 Characterisation of CCN5 provided the first indication that CCN proteins expressed in normal conditions might show important structural differences. A carboxy-truncated CCN2 protein, lacking the CT module, was also isolated in high concentration in primary human osteoblasts.3 Previous studies had established that the cystin knot motif, which is also contained in the CT domain of several growth factors and extracellular matrix proteins, was responsible for dimerisation. The CT domain is essential for CCN3 interaction with several other regulatory proteins, including fibulin 1C,
Notch1, CCN2, and CCN3 itself.3,10 Whether the absence of CT modulates or alters the capacity of these proteins to interact with targets that are recognised by other CCN proteins remains an important point to clarify. In CCN2, the absence of module 4 does not abrogate its ability to promote adhesion of osteoblasts. However, this module is responsible for the ability of CCN2 to induce dose-dependent adhesion of different cell types.7,15 The existence of two biologically active truncated isoforms of CCN2 in normal uterine secretory fluids2 provided the first clue for post-translational processing having an important role. CCN3 is similarly proteolytically cleaved.3 The biological importance of truncated CCN isoforms in tumours was established by the identification of an aminotruncated CCN3 protein expressed in MAV-induced nephroblastoma.3 Although the full-length CCN3 protein was antiproliferative, the aminotruncated protein induced the morphological transformation of normal fibroblasts. Thus modification of subcellular addressing could uncover potential oncogenic activity. An aminotruncated CCN3 protein in tumour cell nuclei3 suggests that accumulation there of CCN3 is associated with tumorigenesis. The interaction of CCN3 with a subunit of RNA polymerase II,3 and the transcriptional transactivation mediated by truncated CCN3 recombinants (our work), suggest direct involvement of CCN3 isoforms in regulating gene expression. Other CCN proteins might also be directed to the nucleus.13
Biological functions of CCN proteins CCN1, 2, 3, 4, 6 SP IGFBP
VWC
TSP1
SP IGFBP
VWC
TSP1
CT
CCN5
CCN2-V1* TSP1
CT
CCN2-V2* CT CCN2-V3 SP IGFBP
VWC
TSP1
VWC
TSP1
CT
TSP1
CT
CCN3-V1 CCN3-V2* CCN1-V1 SP IGFBP
VWC
CT
SP IGFBP
TSP1
CT
SP IGFBP
VWC
CCN4-V1
CCN6-V1
Figure 2: Structure of full-length and truncated CCN proteins Full-length CCN proteins and CCN variants have been aligned to highlight structural differences that might be of biological significance. Isoforms are designated after recommendations of International CCN Society. CCN2-V1, CCN2-V2, and CCN3-V2 are believed to originate from post-translational processing. They are present in biological fluids and cell-culture medium. CCN2-V3, also designated CTGF-L, was identified as novel regulator of osteoblast function. CCN3-V1 was detected in nucleus of cancer cells. CCN1-V1 was identified in normal human fibroblasts, CCN4-V1 was expressed in scirrhous gastric carcinoma, and CCN6-V1 was identified in colorectal carcinomas.14
THE LANCET • Vol 363 • January 3, 2004 • www.thelancet.com
Expression patterns of CCN proteins indicate a role in angiogenesis, chondrogenesis, and wound healing.2,3,15,16 They also participate in cell proliferation, migration, and differentiation. Despite a similar structure, CCN proteins have a great variety of biological functions, highly dependent on the cellular context. Profibrotic activity of CCN2 is interesting because high levels of CCN2 were detected in many fibrotic lesions, (skin, kidney, lung, liver).17 CCN2 might act as a co-factor for transforming growth factor , responsible for induction of fibrogenesis. CCN2 activates transforming growth factor ,18 to mediate extracellular matrix activity downstream. However, fibrotic activity does not seem to be a general feature of the CCN family. CCN1 and CCN2 promote endothelial cell growth, migration, adhesion, and survival in vitro, and their action in angiogenesis is mediated at least partly through interactions with integrins.16 They regulate the activity and production of other angiogenic proteins. CCN3 is also proangiogenic in endothelial cells,19 whereas CCN6 inhibits angiogenesis in the chick aortic-ring assay.4 The proangiogenic activity of CCN1 and CCN2 supports their role in the establishment and functioning of the vasculature and in vascular diseases. Knockout mutations in ccn1 and ccn2 are lethal.20,21 Abrogation of CCN2 induced major skeletal defects because of impaired chondrocyte proliferation and matrix remodelling in the hypertrophic zone. The involvement of CCN proteins in skeletal development agreed with observations that linked CCN3 expression with a late stage in cartilage differentiation3 and CCN6 mutation to pseudorheumatoid dysplasia.22 In-vivo induction of osteogenesis by delivery of rCCN2 into the femoral marrow cavity leading to increased angiogenesis suggests a role for CCN2 in osteoblast proliferation and differentiation.23 By contrast, CCN6-null mice were fully viable4 except for mature endplates, a phenotypic trait suggesting premature ossification. The opposite effects of CCN2 on normal biological processes were shown in transgenic mice that overproduced CCN2 under the control of type XI collagen promoter.15 In
63
For personal use. Only reproduce with permission from The Lancet publishing Group.
RAPID REVIEW
these mice, excess CCN2 led to dwarfism soon after birth, which contrasts with the lack of skeletal development linked with absent CCN2 in null mice. In tumour samples, abnormal expression of CCN proteins is associated with cancer development.14 In many cases, CCN proteins showed converse effects in vivo and ex vivo, with increased expression associated with increased proliferation, inhibition of growth, or both.3 The multimodular architecture of the CNN proteins could allow them to interact with several other proteins whose bioavailability may differ among tissues and with developmental stages.3 Different combinations would thus be the key to biological variety and would confer on the CCN proteins a flexible multifunctionality. In agreement with this model, CCN1, CCN2, and CCN3 interact with signalling and regulatory proteins that have critical roles in the regulation of cell growth. These signalling and regulatory proteins include extracellular-matrix protein (fibulin 1C),9 different receptor types (LRP, Notch 1, integrins),6,10,5,19 a calcium-binding protein (S100A4),11 ion channels (calcium voltage-independent and Cx43 gap junction), a subunit of RNA polymerase II,3 and probably a few more, as suggested by work in our laboratory. Although the interaction of CCN proteins with cell receptors suggested that they might participate in signaling, the modulation of intracellular calcium concentration by CCN3 and CCN2 established these proteins as genuine signalling factors.11,24 The regulation of ion channels by CCN311 was the first function attributed to this protein. The functional interaction of CCN proteins with various proteins and ligands involved in different signaling pathways means that, as matricellular proteins, CCN3 and other CCN family members might act as anchors to bring together regulatory circuits. In such a model, the interaction of CCN3 with fibulin 1C could be a critical step in building multifunctional complexes that would physically connect membrane receptors to extracellular-matrix proteins. CCN3 increases cell adhesion and migration,25 activates focal adhesion kinase,10 and binds to integrins19 that cluster at focal adhesions. The interaction of the cytoskeleton with the extracellular matrix might also involve such multimolecular complexes. Thus CCN proteins could coordinate signalling pathways governing intercellular communication needed for efficient control of cell growth, differentiation, and death.26
Potential clinical applications The involvement of CCN proteins in fundamental biological processes suggests that they might be useful targets for molecular medicine,27 with potential applications in vasculogenesis, chondrogenesis and osteogenesis, nerve conduction, and muscular contraction. CCN proteins are also targets for human carcinogenesis,14 and are promising tools for early diagnosis, typing, and therapy of cancers.27 Monoclonal anti-CCN2 antibodies could become useful fibrosis inhibitors.4 The measurement of CCN protein concentrations, in tumour samples or biological fluids, might help to target treatment.12 However, the variety of biological properties of the different full-length CCN proteins and the identification of CCN isoforms with unknown functions must be taken into account.
References 1
2
3
4
5
6
7
8
9
10
11 12
13
14
15
16 17 18
19
20
21
22
23
24
25 The work in my laboratory was supported by grants from Ligue Nationale Contre le Cancer (Comités du Cher et de l'Indre), the Association pour la Recherche Contre le Cancer (ARC), and the French Ministère de l’Education et de la Recherche. I thank Annick Perbal for editorial help, and M Mochino and Assorca for financial support.
64
26 27
Brigstock DR, Goldschmeding R, Katsube KI, et al. Proposal for a unified CCN nomenclature. J Clin Pathol Mol Pathol 2003; 56: 127–28. Brigstock DR. The connective tissue growth factor/cysteine-rich 61/nephroblastoma overexpressed (CCN) family. Endocr Rev 1999; 20: 189–206. Perbal B. NOV (nephroblastoma overexpressed) and the ccn family of genes: structural and functional issues. Mol Pathol 2001; 54: 57–79. Perbal B, Brigstock DR, Lau L. Report of the second international workshop on the ccn family of genes. J Clin Pathol Mol Pathol 2003; 56: 80–85. Leu SJ, Liu Y, Chen N, et al. Identification of a novel integrin alpha6 beta1 binding site in the angiogenic inducer CCN1(CYR61). J Biol Chem 2003; 278: 33801–08. Gao R, Brigstock D. Low density lipoprotein receptor-related protein (LRP) is a heparin-dependent adhesion receptor for connective tissue growth factor (CTGF) in rat activated hepatic stellate cells. Hepatol Res 2003; 27: 214–20. Ball DK, Rachfal AW, Kemper SA, Brigstock D. The heparinbinging 10 kDa fragment of connective tissue growth factor (CTGF) containing module 4 alone stimulates cell adhesion. J Endocrinol 2003; 176: R1–R7. Schober JM, Lau LF, Ugarova TP, Lam SC. Identification of a novel integrin alphaMbeta2 binding site in CCN1(CYR61), a matricellular protein expressed in healing wounds and atherosclerotic lesions. J Biol Chem 2003; 278: 25808–15. Perbal B, Martinerie C, Sainson R, et al. The C-terminal domain of the regulatory protein NOVH is sufficient to promote interaction with fibulin 1C. Proc Natl Acad Sci USA 1999; 96: 869–74. Sakamoto K, Yamaguchi S, Ando R, et al. The nephroblastoma overexpressed gene (nov/ccn3) protein associates with Notch1 extracellular domain and inhibits myoblast differentiation via Notch signaling pathway. J Biol Chem 2002; 277: 29399–405. Li CL, Martinez V, He B, Lombet A, Perbal B. A role for CCN3 (NOV) in calcium signalling. Mol Pathol 2002; 55: 250–61. Riser BL, Cortes P, Denichilo M, et al. Urinary CCN2 (CTGF) as a possible predictor of diabetic nephropathy: preliminary report. Kidney Int 2003; 64: 451–58. Abdel-Wahab N, Weston BS, Roberts T, Mason RM. Connective tissue growth factor and regulation of the mesangial cell cycle: role in cellular hypertrophy. J Am Soc Nephrol 2002; 13: 2437–45. Planque N, Perbal B. A structural approach to the role of CCN proteins in tumorigenesis. Cancer Cell Int 2003; 3: 15. http://www. cancerci.com/content/3/1/15 (accessed Dec 1, 2003). Takigawa M. CTGF/Hcs24 as a multifunctional growth factor for fibroblasts, chondrocytes and vascular endothelial cells. Drug News Perspect 2003; 16: 11–21. Lau LF, Lam SC. The CCN family of angiogenic regulators: the integrin connection. Exp Cell Res 1999; 248: 44–57. Moussad EE, Brigstock DR. Connective tissue growth factor: what’s in a name? Mol Genet Metab 2000; 71: 276–92. Abreu JG, Ketpura NI, Reversade B, De Robertis EM. Connective-tissue growth factor (CTGF) modulates cell signalling by BMP and TGF-beta. Nat Cell Biol 2002; 4: 599–604. Lin C, Leu SJ, Chen N, et al. CCN3 (NOV) is a novel angiogenic regulator of the CCN protein family. J Biol Chem 2003; 278: 24200–08. Ivkovic S, Yoon BS, Popoff SN, et al. Connective tissue growth factor coordinates chondrogenesis and angiogenesis during skeletal development. Development 2003; 130: 2279–91. Mo FE, Muntean AG, Chen CC, et al. CYR61 (CCN1) is essential for placental development and vascular integrity. Mol Cell Biol 2002; 22: 8709–20. Hurvitz JR, Suwairi WM, Van Hul W, et al. Mutations in the CCN gene family member WISP3 cause progressive pseudorheumatoid dysplasia. Nat Genet 1999; 23: 94–98. Safadi FF, Xu J, Smock SL, et al. Expression of connective tissue growth factor in bone: its role in osteoblast proliferation and differentiation in vitro and bone formation in vivo. J Cell Physiol 2003; 196: 51–62. Lombet A, Planque N, Bleau AM, Li C.L, Perbal B. CCN3 and calcium signaling. Cell Commun Signal 2003; 1: 1. http://www. biosignaling.com/content/1/1/1 (accessed Dec 1, 2003). Scotlandi K, Benini S, Manara MC, et al Biological role of NOVH in Ewing’s sarcoma cells. Mol Pathol 2003; 56: 73 (abstr). Perbal B. Communication is the key. Cell Commun Signal 2003; 1: 3. http://www.biosignaling.com/content/1/1/3 (accessed Dec 1, 2003). Perbal B. The CCN3(NOV) cell growth regulator: a new tool for molecular medicine. Expert Rev Mol Diag 2003; 3: 597–604.
THE LANCET • Vol 363 • January 3, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.