Targeted TGF-β chemotherapies: friend or foe in treating human malignancies?

Transforming growth factor-β paradox

Transforming growth factor (TGF)-β is a multifunctional cytokine that plays essential roles in regulating virtually all aspects of mammalian development and differentiation, and in maintaining mammalian tissue homeostasis [1,2]. Indeed, the ubiquitous and multifunctional nature of TGF-β is highlighted by the fact that virtually every cell in the metazoan body is capable of both producing and responding to this cytokine. Although TGF-β was identified originally via its stimulation of morphological transformation and anchorage-independent growth in rat NRK-49 kidney fibroblasts, this cytokine is now recognized as being a potent tumor suppressor that prevents the dysregulated growth and survival of cells derived from epithelial, endothelial or hematopoietic lineages [1,2]. The process of tumorigenesis and its assortment of associated genetic and epigenetic events enable newly malignant cells to evade the cytostatic activities of TGF-β. As cancer cells continue down the evolutionary path towards advanced malignancy, they ultimately acquire the ability to transform the cytostatic signals produced by TGF-β into oncogenic activities, including enhanced proliferation, invasion and metastasis. The ability of malignant cells to convert the biological actions of TGF-β (i.e., tumor suppression) into pathological symptoms (i.e., tumor promotion) is commonly referred to as the ‘TGF-β paradox’, which remains the most important and unanswered question concerning physiological functions of this multifunctional cytokine. Moreover, the schizophrenic nature exhibited by TGF-β during tumorigenesis is not restricted solely to the cancer cells themselves, but also occurs in their accompanying stromal components, including fibroblasts, endothelial and infiltrating immune cells [1,2]. Indeed, when activated by TGF-β, these various stromal components conspire to create a tumor–host microenvironment that promotes the induction, selection and expansion of metastatic cells, thereby ensuring dissemination of the disease beyond its tissue of origin [1,2]. The consistent and repeated finding in nature of cancer cells that readily undergo invasion and metastasis in response to TGF-β underscores the importance of creating novel chemotherapeutics operant in targeting the oncogenic activities of TGF-β in developing and progressing human cancers.

Despite intense research efforts over the last decade by scientists in academic and industrial settings, complete success in targeting the TGF-β signaling system in human cancers has remained elusive. Indeed, the multifunctional nature of TGF-β probably represents the greatest barrier to effectively targeting TGF-β, its receptors and their downstream effectors in human tumors, which must acquire six distinct tumorigenic phenotypes in the course of their neoplastic journey. For instance, all cancer cells must develop the capacity to proliferate autonomously, ignore cytostatic signals, resist apoptotic signals, become angiogenic, initiate tissue invasion and metastasis, and become immortal [3]. We and others have shown that TGF-β plays a prominent role in regulating, either directly or indirectly, the acquisition of each of these individual traits by malignant cells [1,2,4,5]. Furthermore, the role of TGF-β in maintaining cell and tissue homeostasis necessitates that extreme caution be exercised during the delivery of targeted TGF-β therapies. For example, pan-antagonism of TGF-β function during the infancy of tumorigenesis may in fact promote cancer development by alleviating the cytostatic activities of TGF-β, not only in the malignant cells themselves [1,2], but also in their stromal counterparts where the loss of TGF-β function promotes neoplastic development [6,7]. This caveat is countered by the belief that administration of pan-TGF-β antagonists to advanced-stage cancers will inhibit the oncogenic activities of TGF-β, particularly its ability to induce tumor angiogenesis, invasion and metastasis, and its ability to inhibit host immunosurveillance. Thus, the timing, context and disease status initially encountered when deciding to launch targeted TGF-β treatment regimens will require critical information capable of identifying which cancer patients are potentially indicated or contraindicated for anti-TGF-β therapy.

Targeted transforming growth factor-β chemotherapies

At present, targeted TGF-β chemotherapies come in two flavors, namely large- and small-molecule TGF-β inhibitors [8]. Large-molecule inhibitors of TGF-β signaling include monoclonal anti-TGF-β antibodies (e.g., GC1008), soluble TβR-II–Fc fusion proteins, decorin and soluble TβR-III, as well as antisense TGF-β2 oligonucleotides (e.g., AP12009) [1,2,8]. In general, these molecules inhibit TGF-β signaling by neutralizing and/or sequestering TGF-β from its cell surface receptors or by limiting the production of TGF-β within tumor microenvironments. Other examples of large-molecule TGF-β antagonists are cystatin C, which binds TβR-II and prevents its binding to TGF-β [9,10] and fetuin/α-2HS glycoprotein, which binds TGF-β and prevents its binding to TβR-II [11,12]. As a group, these molecules effectively antagonize TGF-β signaling in a number of in vitro systems that model key tumorigenic processes (e.g., epithelial–mesenchymal transition [EMT], migration and invasion), as well as reduce the growth and metastasis of murine 4T1 breast cancer cells in mice [1,2]. Conversely, the growth of human MDA-MB-231 breast cancer cells in mice was enhanced by administration of neutralizing TGF-β antibody (e.g., human GC1008 or mouse ID11), as was the progression of indolent colon adenomas to highly aggressive adenocarcinomas [1,2]. Thus, abrogating TGF-β signaling within specific tumor microenvironments may in fact exacerbate disease development in a manner predicted by the TGF-β paradox. Despite these apparent concerns, GC1008 has entered human clinical trials to evaluate its effectiveness in treating patients with idiopathic pulmonary fibrosis, as well as those with metastatic skin and renal cancers [1,2].

In contrast to the TGF-β neutralizing activity of large-molecule TGF-β antagonists, those of the small-molecule flavor are comprised of compounds that bind competitively with varying selectivities to the ATP-binding site of the TβR-I protein kinase domain (e.g., LY580276, SB431542, A-83–01, or SD-208 and -093) [8]. As with the administration of large-molecule inhibitors, the use of small-molecule antagonists to manipulate TGF-β signaling in cancer cells has produced mixed results. For instance, TβR-I antagonists inhibit the ability of TGF-β to stimulate EMT and invasion in normal and malignant cells [1,2,8], suggesting that physiological EMT (i.e., wound healing and tissue remodeling) may be negatively impacted in cancer patients treated with these drugs. In addition, while administration of TβR-I antagonists to advanced-stage cancers inhibits oncogenic signaling by TGF-β, this same treatment protocol actually enhances the malignancy of early stage cancers that remain sensitive to the cytostatic actions of TGF-β. These findings underlie the necessity of designing and implementing rapid diagnostic tests capable of discriminating those cancer patients most likely to benefit from targeted TGF-β therapies from those individuals most likely to be harmed by TGF-β antagonism. This point is medically relevant as cancer cells that have undergone oncogenic EMT often develop resistance to standard cancer chemotherapies. Thus, administration of TβR-I antagonists in conjunction with erlotinib, an epidermal growth factor receptor antagonist, may afford new avenues to control the metastatic spread of developing and progressing cancers. Along these lines, TβR-I antagonists have recently been shown to inhibit the growth and metastasis of breast and pancreatic cancers implanted into mice. Interestingly, these pharmacological effects were not observed in immunocompromised mice [13], suggesting that TβR-I inhibitors play a prominent role in eliminating tumors by bolstering host immunosurveillance. Despite many of the uncertainties associated with the use of targeted TGF-β therapies, this novel class of cancer chemotherapeutics remains an attractive and potentially powerful approach to control the development and spread of human malignancies.

Future directions in targeting the transforming growth factor-β paradox

Generally speaking, the targeted TGF-β therapies discussed previously uniformly function as pan-TGF-β antagonists whose activities are subject to the phenomena underlying the TGF-β paradox. Although considerable progress in our understanding of the TGF-β paradox has been achieved in recent years, science and medicine still do not yet know precisely how carcinogenesis converts the cellular response to TGF-β. What is known is that canonical Smad2/3/4 signaling plays a prominent role in mediating the cytostatic function of TGF-β, while inappropriate or amplified activation of noncanonical signaling by TGF-β (e.g., mitogen-activated protein kinases [MAPKs], AKT and nuclear factor [NF]-κB), together with altered Smad2/3 signaling inputs, contribute to oncogenic signaling by TGF-β [1,2,14]. It therefore stands to reason that specific targeting of noncanonical pathways activated by TGF-β, as opposed to pan-antagonism of TGF-β signaling itself, may provide novel avenues to circumvent the TGF-β paradox and restore the tumor-suppressing function of TGF-β in human cancers. We recently discovered a novel αvβ3 integrin–Src–phospho-Y284–TβR-II–Grb2–p38 MAPK signaling axis whose activation mediates oncogenic signaling by TGF-β in normal and malignant mammary epithelial cells [15,16]. Importantly, the ability of TGF-β to stimulate the growth and pulmonary metastasis of breast cancers in mice absolutely requires activation of this oncogenic signaling complex Galliher AJ & Schiemann WP, unpublished DATA]. We also uncovered a novel TAB1–TAK1–IKKβ–xIAP–NF-κB signaling axis that forms aberrantly in breast cancer cells, but not in their normal counterparts, and enables oncogenic signaling by TGF-β [Neil JR & Schiemann WP, submitted]. Importantly, interdicting the ability of TGF-β to activate either oncogenic signaling axis abrogated the tumor-promoting activities of TGF-β without affecting its coupling to Smad2/3. The net effect of selective TGF-β antagonism was a partial restoration of the cytostatic function of TGF-β in breast cancer cells, a result most likely arising from the fact that oncogenic signaling by TGF-β appears to be evolutionarily and functionally redundant to that mediated by growth factor receptors. Future studies clearly need to address the validity of this hypothesis; they also need to identify the molecular targets activated by these oncogenic signaling axes, as well as determine their therapeutic potential as chemopreventive targets in alleviating cancer progression driven by TGF-β.

Parting thoughts

Solving and ultimately manipulating the TGF-β paradox to improve the human condition is in many respects the ‘Holy Grail’ for TGF-β biologists and pharmacologists. By continuing to unravel the mysteries that underlie the biology and pathology of TGF-β, it may one day be possible to selectively target the oncogenic activities of TGF-β and, consequently, to ‘normalize’ malignant tissues in such a way that cancer itself is converted from an acute, symptomatic and life-threatening disease to one that is chronic, asymptomatic and manageable through the normal lifespan of affected individuals.

Acknowledgements

Members of the Schiemann Laboratory are thanked for critical comments and reading of the manuscript. William P Schiemann is supported by grants from the National Institutes of Health (CA095519 and CA114039), the University of Colorado Cancer Center and the Cancer League of Colorado.