TGF-β is widely expressed during development to regulate the interactions between epithelial and mesenchymal cells, particularly those in the lung, kidney, and mammary gland. Inappropriate reactivation of EMT during tumorigenesis is now recognized as an important process necessary for the acquisition of invasive and metastatic phenotypes by tumors [1
]. By cooperating with oncogenes and growth factors, TGF-β potently induces EMT and serves to stabilize this transition by means of autocrine signaling. Moreover, these events seem to underlie the oncogenic activities of TGF-β and its ability to promote cancer progression [20
]. A comprehensive understanding of how TGF-β both suppresses and promotes tumorigenesis remains an unknown and fundamental question that directly affects our ability to effectively target the TGF-β signaling system during the treatment of human malignancies, particularly those of the breast. Indeed, solving this paradox remains the most important aspect of the biological and pathological actions of this multifunctional cytokine.
Despite recent advances in understanding the molecular mechanisms underlying EMT, the question of how to prevent this process effectively in response to TGF-β remains unanswered. We recently discovered that CystC antagonizes TGF-β signaling in normal and cancer cells by interacting physically with TβR-II, thereby preventing TGF-β binding [12
]. Importantly, we demonstrated the effectiveness of CystC in inhibiting the invasion of cancer cells and the TGF-β-stimulated invasion of fibroblasts [12
]. Because EMT is necessary for the acquisition of invasive and metastatic phenotypes by cancer cells, and because CystC inhibited TGF-β-stimulated invasion, we proposed CystC as a potential antagonist of EMT stimulated by TGF-β.
Accordingly, in this study we show that CystC and Δ14CystC both prevent EMT and its associated increase in MEC motility (Figs , , ), as well as antagonizing TGF-β signaling in human breast cancer cells (Figs and ). We further show for the first time that these CystC molecules inhibit TGF-β signaling in NRK fibroblasts, thus preventing their morphological transformation and invasion through synthetic basement membranes (Fig. ). Although our understanding of CystC function in regulating TGF-β signaling is in its infancy, our findings suggest that this protease inhibitor might provide an innovative model for the development of novel TβR-II antagonists designed to combat the stimulation of tumor progression and EMT by TGF-β. We further propose that the chemopreventive effectiveness of CystC will be potentiated by its inhibition of cathepsin B-mediated invasion and metastasis [27
] and its inhibition of the cathepsin B-mediated activation of latent TGF-β [31
], which co-localizes with cathepsin B to the invading face of malignant tumors [34
]. Moreover, CystC-mediated cathepsin B inactivation will reduce the activity of the urokinase plasminogen system, which enhances tumor cell extracellular matrix degradation, as well as growth factor and latent TGF-β activation [38
]. Cumulatively, the chemopreventive activities of CystC will antagonize cancer cell responses to TGF-β by inhibiting TGF-β binding [12
] and by reducing TGF-β bioavailability within tumor microenvironments, thereby alleviating the stimulation of EMT and tumor metastasis in late-stage tumors by TGF-β.
Molecular dissection of TGF-β signaling systems necessary for its induction of EMT has clearly established a role for Smad2/3 in mediating EMT, particularly when coupled with signals emanating from oncogenic Ras [13
]. However, Smad2/3-independent signaling has also been implicated in TGF-β stimulation of EMT. For instance, TGF-β stimulates EMT in cancers of the breast and other tissues by activating phosphoinositide 3-kinase, Akt, RhoA, p160(ROCK), and p38 mitogen-activated protein (MAP) kinase [40
]. In addition, EMT in TGF-β-treated MECs is abrogated by measures that inhibit β1 integrin activity [42
], thus establishing the necessity of β1 integrin expression for EMT stimulated by TGF-β. Finally, by repressing Id2 and Id3 expression [45
], inducing Snail and SIP1 expression [46
], and stimulating nuclear factor-κB activity [47
], TGF-β regulates transcription factor activity operant in mediating the transition from epithelial to mesenchymal cell markers. We show that CystC inhibits the stimulation of Smad2 phosphorylation by TGF-β and the subsequent induction of reporter gene expression in normal and cancer MECs. Thus, reduced Smad2/3 signaling mediated by CystC probably underlies part of its ability to inhibit EMT stimulated by TGF-β. Future studies need to address the role of CystC in regulating Smad2/3-independent signaling stimulated by TGF-β, as well as determining their contribution in preventing the oncogenic activities of TGF-β in human breast cancer cells.
Finally, we were quite surprised to find that CystC, despite its ability to inhibit Smad2/3 signaling, failed to alter the growth-suppressing activities of TGF-β. Although the molecular mechanism(s) underlying this unexpected CystC activity remains to be elucidated, our findings suggest that CystC does not function to abrogate all TGF-β signaling, but may instead specifically alter and/or modulate certain aspects of TGF-β signaling when complexed to TβR-II. In support of this supposition, we find that activation of MAP kinases and Akt by TGF-β requires high cytokine concentrations, whereas that of Smad2/3 requires markedly lower cytokine concentrations (more than 10-fold lower; data not shown). Mechanistically, we propose that maximal stimulation of Smad2/3 by TGF-β requires minimal receptor occupancy (that is, large receptor reserves), whereas maximal stimulation of MAP kinases and AKT requires maximal receptor occupancy (that is, no receptor reserves). Thus, manipulations designed to antagonize the binding of TGF-β to its receptors might elicit disproportionate inhibition of TGF-β signaling systems, resulting in greater inhibition of Smad2/3-independent versus Smad2/3-dependent pathways. Future studies need to address this important question and determine which TGF-β signaling systems are preferentially inhibited by the formation of CystC–TβR-II complexes in normal and cancerous MECs.