Terminology of CCN1-6 should not be applicable for their fragments and be limited to only full length CCN1-6

It is well known that CCN proteins are cysteine-rich secreted proteins and composed of 4 distinct modules connected in tandem, i. e., IGF binding protein-like (IGFBP), von Willebrand type C (VWC), thrombospondin type 1 repeat (TSP1), and C-terminal modules, except for CCN5 which lacks CT-module (Takigawa et al 2003; Perbal and Takigawa 2005). Exploring active modules (sites) in this molecule has been great interest for CCN researchers for a long time and is still underway. Recently, Bruno G. et al. (2015) reported a paper entitled “CTGF/CCN2 exerts profibrotic action in myoblasts via the up-regulation of sphingosine kinase-1/S1P3 signaling axis: Implications in the action mechanism of TGFβ.” in Biochim Biophys Acta. In this article, the authors described “CTGF/CCN2” in the paper title but they used CTGF purchased from PeproTech in their experiments although they describe CTGF (lower molecular weight isoform of 11 kDa) in Materials and Methods section as described in the catalog of the company. Preceding this paper, Liu SC et al. (2012) had reported a paper entitled “CTGF increases IL-6 expression in human synovial fibroblasts through integrin-dependent signaling pathway” in PLosOne but in this paper the authors only described that human CTGF was purchased from PeproTech in Materials and Methods section. However, PeproTech (Rocky Hill, NJ, USA) sales only one type of CTGF named “recombinant human CTGF”, but it is an 11.0 kDa protein consisting of 97 amino acid residues, almost corresponding to C-terminal cystin-not module of CTGF/CCN2. The peptide is longer than CT-module and covers entire CT-module but not overlaps with TSP1 module. They reveal its a.a. sequence from which one can easily recognize that the peptide corresponds to the C-terminal module. Therefore, I am afraid that such a description in the above articles causes confusion in the field of CCN research. Indeed, C-terminal module of CCN2 has been shown to stimulate cell adhesion (Ball et al. 2003; Gao and Brigstock 2004; Hoshijima et al. 2006) and cell attachment (Ono et al. 2007), cell proliferation (Perbal and Takigawa 2005; Takigawa 2013) and is involved in dimerization (Hoshijima et al. 2012; Abd El Kader et al. 2014) and binding to integrin alpha(v)beta(3) (Gao and Brigstock 2004; Ono et al. 2007), heparan sulfate proteoglycan (Gao and Brigstock 2004) and fibronectin (Hoshijima et al. 2006). However other 3 modules also have biological activities and binding capacity to other molecules. For example, N-terminal half of CCN2/CTGF (IGFBP-VWC), but not C-terminal half (TSP1-CT) binds to aggrecan and overexpression of corresponding parts of CCN2 gene in chondrocytes revealed that IGFBP-VWC, but not TSP1-CT, stimulate aggrecan production (Aoyama et al. 2009). Not only CT module but also VWC and TSP modules as well as full length CCN2 are capable of efficiency activating the ERK signal transmission cascade in human umbilical vein vascular endothelial cells, whereas IGFBP is not (Kubota et al. 2006). In contrast, the IGFBP module prominently activates JNK in human chondrocytic HCS-2/8 cells (Kubota et al. 2006). Moreover, Abd El Kader et al. (2014) also reported that IGFBP module has the most potent activity in inducing gene expression of aggrecan in chondrocytes whereas TSP1 module has the most potent activity in inducing gene expression of collagen type II in the cells. Since TSP1, but not IGFBP, can bind gelatin hydrogel which is a slow releaser, the TSP1 modules mixed with gelatin hydrogel was applied to osteoarthritis animal model and found to have cartilage regenerative effect better than that of recombinant full length CCN2 (Abd El Kader et al. 2014) . Also in an early study, Abreu et al. (2002) reported that CCN2/CTGF bind TGF-β1 and BMP-4 through CR/VWC domain, inhibiting BMP signaling and activating TGF-β signaling. Moreover, Grotendorst and Duncan (2005) reported that individual domains of connective tissue growth factor regulate fibroblast proliferation and myofibroblast differentiation. More recently CCN2 is reported to stimulate collagen deposition by gingival fibroblast via module 3 (TSP1) and alpa6- and beta1 integrins (Heng et al. 2006). These findings clearly indicates that each module of CCN2 has specific activity and hence one should not say that CT-module of CCN2 is an active form of CCN2 protein although the CT-module is responsible for some of many activities of this multifunctional factor. Therefore, when CCN researchers publish their data obtained by using CT-module of CCN2, the authors should describe clearly what type of CCN2, full length or CT-module in their papers. This should be, of course, applied for other modules. Otherwise, confusion would spread over the field of CCN2 research. In addition, there is another example of possible confusion. B. Perbal’s group reported that amino-truncated variants of CCN3 are addressed to the nucleus and that the carboxyterminal (CT) module of CCN3 is responsible for the nuclear addressing (Planque et al. 2006). They also reported that the use of several clones expressing various portions of CCN3 established that the CT module of CCN3 is sufficient to induce cell growth inhibition (Bleau et al. 2007) but not sufficient for all the activities of the full length CCN3 (B. Perbal, personal communication). Since the full length of CCN3 inhibits proliferation of growth plate and epiphyseal chondrocytes in culture, this inhibition may be attributed to its CT module (Kawaki et al. 2008; Janune et al. 2011). However, CCN3 inhibited proteoglycan production in growth plate chondrocytes (Kawaki et al. 2008) in culture and stimulated it in epiphyseal chondrocytes in culture (Janune et al. 2011). Thinking about that the different modules of CCN2/CTGF are responsible for different activities of CCN2 described above, it is feasible that the other modules rather than the CT-module of CCN3 are involved in the effect of CCN3 on proteoglycan synthesis. Therefore, it is not concluded yet whether the CT-module of CCN3 is the active module of CCN3. So far, full-length recombinant CCN3 is commercially available but fortunately the CT-module is not commercially available. However, if someone in a company misunderstand that CT-module of CCN3 is the active form of CCN3 and if the company sales the CT-module of CCN3 as if it were the active form of CCN3, much more confusion would spread over the field of CCN research. This should be, of course, applied for all other CCN proteins and must be protected by careful effort of our scientific community.