New tricks for an old fox: Impact of TGFβ on the DNA damage response and genomic stability

See allHide authors and affiliations

Science Signaling  02 Sep 2014:
Vol. 7, Issue 341, pp. re5
DOI: 10.1126/scisignal.2005474


Transforming growth factor–β (TGFβ) is a well-known master regulator of cellular proliferation and is a critical factor in the maintenance of tissue homeostasis. TGFβ is classically defined as a tumor suppressor that functions in the early stages of carcinogenesis, yet paradoxically it functions as a tumor promoter in established cancers. Less well studied is its role in maintaining genomic stability through its participation in the DNA damage response (DDR). Deletion of Tgfb1 in murine epithelium increases genomic instability (GIN) as measured by gene amplification, aneuploidy, and centrosome aberrations; likewise, GIN is increased by depleting the TGFβ ligand or inhibiting TGFβ pathway signaling in human epithelial cells. Subsequent studies demonstrated that TGFβ depletion compromises cell survival in response to radiation and impairs activation of the DDR because of severely reduced activity of ataxia telangiectasia mutated (ATM), a serine/threonine protein kinase that is rapidly activated by DNA double-strand breaks. The SMAD transcription factors are intermediaries in the crosstalk between the TGFβ and ATM pathways in the DDR. Recent studies have shown that SMAD2 and SMAD7 participate in the DDR in a manner dependent on ATM or TGFβ receptor type I, respectively, in human fibroblasts and epithelial cells. Understanding the role of TGFβ in the DDR and suppressing GIN is important to understanding its seemingly paradoxical roles in tumorigenesis and thus has therapeutic implications for improving the response to DNA damage–inducing therapy.

Control of Genomic Stability by TGFβ

Transforming growth factor–β (TGFβ) is a pluripotent cytokine. The activity of TGFβ is restrained by its secretion as a latent complex, which consists of the 24-kD cytokine and an 80-kD dimer of its pre-pro region called latency-associated peptide (LAP). Cells secrete the latent complex, which is abundant in the extracellular space, bound to extracellular matrix and in circulation, particularly in platelets. The latent complex requires activation to release TGFβ from LAP, which is the essential control of its bioactivity. We use TGFβ to refer to its effects or the mature cytokine. Active TGFβ1 and TGFβ3 ligands are able to directly bind the type II TGFβ receptor (TβRII), whereas TGFβ2 is able to efficiently bind TβRII only in the presence of the type III TGFβ receptor (TβRIII). The ligand-bound TβRII is able to subsequently transactivate the type I TGFβ receptor (TβRI) to promote the majority of downstream signaling (1, 2). The downstream signaling is primarily due to serine/threonine kinase activity localized to a glycine- and serine-rich region TβRI termed the GS domain (2). Atypical signaling, such as phosphorylation of tyrosine residues, can also be directly attributed to activation of the TGFβ receptor complex (3).

TGFβ promotes signaling through a family of transcription factor proteins called SMADs (2, 4, 5). TβRI phosphorylates receptor-associated SMAD family members 2 and 3 (R-SMADs), thereby enabling a conformational change that permits assembly of hetero- and homo-oligomeric complexes with SMAD4 (also referred to as co-SMAD or common mediator SMAD). R-SMAD activity is cell type– and context-dependent. Moreover, TGFβ signaling can also occur through a noncanonical network currently known to include the adaptor protein ShcA; the guanosine triphosphatases RHO, RAC, CDC42, and RAS; the E3 ubiquitin ligases tumor necrosis factor receptor–associated factor 6 (TRAF6) and mitogen-activated protein kinase kinase kinase 1 (MAP3K1); TGFβ-activated kinase 1 (TAK1); phosphoinositide 3 kinase (PI3K); the cell polarity protein PAR6; the death domain–associated protein DAXX; and the protein phosphatase PP2A (27).

Activated SMAD complexes formed in the cytoplasm shuttle into the nucleus in response to intrinsic nuclear localization signals, but exit the nucleus as monomers through alternate mechanisms (8). At present, it is thought that the SMAD complex is actively involved in the regulation of transcriptional activation and repression. The specific outcome of SMAD activation is a function of the presence or absence of other transcription factors and parallel signaling pathways (2, 4, 5). SMAD1 and SMAD5 activation has also been described in response to TGFβ stimulation in epithelial cells, carcinoma cells, and fibroblasts (9, 10). These phosphorylation events now link SMAD1 and SMAD5, traditionally considered to be regulated by bone morphogenetic protein (BMP) signaling, to motility and invasion of cancer cells in response to TGFβ stimulation (9, 10). The net activation of SMAD-dependent and SMAD-independent pathways, together with the interactions derived from the presence or absence of other parallel signaling cascades [such as estrogen, epidermal growth factor, hepatocyte growth factor, and wingless (WNT) ligands], determines the functional response to TGFβ.

An important determinant of SMAD2 and SMAD3 activation is the status of SMAD7 expression, which is a target of R-SMAD transcriptional regulation. SMAD7 binds TβRI and blocks the activation of SMAD2 and SMAD3 through competitive inhibition of the common active site on TβRI (11). SMAD7 also promotes dephosphorylation of the activated receptor complex, thus attenuating TGFβ signaling through association with the SMURF (SMAD ubiquitin regulatory factor) E3 ubiquitin ligase proteins, SMURF1 and SMURF2 (12, 13).

TGFβ is often described as a canonical tumor suppressor because escape from TGFβ growth regulation is pervasive in a wide range of cancers. Loss of response to TGFβ as a growth inhibitor and increased levels of TGFβ are associated with malignant conversion and progression in breast, gastric, endometrial, ovarian, and cervical cancers, as well as glioblastoma and melanoma (14). The primary intracellular mediators of TGFβ signaling are the SMAD family of proteins, which are observed to be activated by receptor-mediated phosphorylation in various human cancers. Inactivation of SMAD4 through homozygous deletion or intragenic mutation occurs frequently in association with malignant progression in pancreatic and colorectal cancer (15). Expression of SMAD2 and SMAD4 is frequently decreased in human squamous cell carcinomas, often because of loss of heterozygosity (16), and disruption of Smad2 can lead to malignant transformation in mice (17). However, mutation of the genes that encode TGFβ ligand or receptors occurs only occasionally in most human cancers. For example, in a study of more than 500 breast cancers, Reiss and colleagues showed that 92% were positive for nuclear, phosphorylated SMAD2, indicating activation of the TGFβ pathway (18). Indeed, many TGFβ-mediated transcriptional responses are intact in cancer cells that have escaped the control of proliferation. More importantly, it is clear that increased TGFβ in cancer cells can act in various ways to promote neoplastic progression. Production of TGFβ in malignant cells acts on the host to suppress antitumor immune responses, to enhance extracellular matrix production, and to augment angiogenesis [reviewed in (19)]. These activities resemble those induced by TGFβ during wound healing and may create a “permissive” microenvironment that promotes malignant growth by acting on the host.

Genomic instability (GIN) is a less well-recognized consequence of TGFβ loss, yet deletion of Tgfb1 greatly increases GIN in murine epithelial cells (20). Using cultured keratinocytes isolated from newborn Tgfb1-null, heterozygote, and wild-type mice, Yuspa and colleagues showed that Tgfb1-null cells spontaneously immortalized more readily than TGFβ-competent cells. Compared with wild-type cells, Tgfb1-null cells gave rise to 1000-fold more mutant clones that were resistant to N-phosphonacetyl-l-aspartate (PALA), an event requiring amplification of the gene encoding dihydrofolate reductase. This unexpected phenotype was difficult to place within the pathways known to be controlled by TGFβ. Following up on this finding, our laboratory found increased amounts of centrosome aberrations, chromosomal instability, and spontaneous DNA damage in nonmalignant human epithelial cells in which TGFβ signaling was inhibited by a small-molecule inhibitor of the TGFβ type I receptor kinase (TβRI) (21). We also showed that heterozygous Tgfb1 mammary epithelium, which has only 10 to 30% the abundance of TGFβ compared with wild-type epithelium, exhibits more GIN, comparable to that in epithelium that is heterozygous for Trp53 (encoding tumor suppressor protein p53).

Studies have shown that radiation induces TGFβ activity in vitro and in vivo in both normal and cancer cells (2228). This observation gained more importance with the finding that TGFβ regulates the expression of genes that encode key DNA damage response (DDR) proteins (29) and with our study showing that epithelial tissues from Tgfb1-null embryos fail to undergo apoptosis or cell cycle arrest in response to high-dose (5 Gy) radiation (30). Our subsequent study found that the absence of cell death could be due to a failure of the DDR, attributed to a marked reduction of radiation-induced activity of ataxia telangiectasia mutated (ATM) (31).

ATM is a PI3K-related serine/threonine kinase that mediates an extensive DDR signaling network, which includes DNA repair, specific cell cycle checkpoints, and programmed cell death. This network is most vigorously activated in response to DNA double-strand breaks (DSBs) (32). Mutations in human ATM lead to the GIN syndrome ataxia telangiectasia (A-T), which is characterized, among others, by extreme radiosensitivity (33). A-T cell lines have increased sensitivity to the cytotoxic effect of ionizing radiation (IR) and exhibit high amounts of simple and complex chromosomal aberrations and unrejoined DNA fragments (32). DSBs elicit the induction of sensing and processing proteins that can be observed by immunofluorescence in IR-induced foci (IRIF) (34). IRIF may contain many proteins involved in ongoing repair or checkpoint control, such as the tumor suppressor p53-binding protein 1 (53BP1), the DSB repair protein RAD51, checkpoint protein CHK2, and activating transcription factor 2 (ATF2) (3540). ATM also phosphorylates H2A histone family member X (H2AX, then called γH2AX) at and near DSBs (41). ATM is activated in response to DSBs and in turn phosphorylates numerous substrates, thereby modulating the DDR and cell survival decisions. ATM precisely controls its downstream pathways, often by mediating the same effect through several different paths; for example, each of the cell cycle checkpoints is regulated by several ATM-mediated pathways (32). Notably, in addition to ATM’s versatility as a protein kinase with numerous substrates, the ATM signaling nexus contains protein kinases (such as CHK1 and CHK2) that are themselves capable of targeting several downstream effectors simultaneously and thereby concomitantly control subsets of pathways. A prototypic example is the direct and indirect ATM-mediated phosphorylation of p53 that results in its activation and stabilization, which is critical to its action as a major player both in the G1-S cell cycle checkpoint and in damage-induced apoptosis (32).

We and others have shown that TGFβ regulates the kinase activity of ATM. TGFβ depletion by genetic knockout in mouse cells, or TGFβ inhibition of signaling in human cells, compromises ATM activity and autophosphorylation, leading to reduced phosphorylation of critical ATM targets, abrogation of DNA damage–induced cell cycle checkpoints, and increased cellular radiosensitivity (30, 31, 42). The ability of exogenous TGFβ to restore these responses indicates that the effect is both cell-intrinsic and distal to TGFβ signaling. Inhibition of TGFβ receptor signaling with a small-molecule inhibitor in irradiated human cells phenocopies the molecular and cellular consequences of TGFB1 deletion. This requirement for TGFβ in the genotoxic stress program provides a previously unsuspected avenue to modulate radiotherapy.

SMADs and recognition of DNA damage

Exploration of the mechanisms involved in TGFβ-mediated modulation of the DDR has extended to the canonical TGFβ-SMAD signaling pathway. In nonstimulated cells, the receptor-associated SMADs (designated R-SMADs 1, 2, 3, 5, and 8) are predominantly localized to the cytoplasm (2). Once the TGFβ receptor complex is activated by ligand, it aids in phosphorylation of the R-SMADs that then complex with co-SMAD4 and promote the nuclear translocation and retention of R-SMADs to activate target gene expression (43). SMAD2 is then degraded or dephosphorylated and exported out of the nucleus (44). Inhibitory SMAD7 is not phosphorylated after TGFβ activation, because it lacks the type I receptor phosphorylation site (45). It is a general antagonist of TGFβ signaling, and it regulates the formation of SMAD2/SMAD4 complexes, blocking the nuclear accumulation of SMAD2 and SMAD3. It also binds SMURF2 to form an E3 ubiquitin ligase that targets TβRI for degradation, thereby inhibiting the activation of SMAD2 (46). Two histone deacetylases, HDAC1 and SIRT1, are reported to deacetylate SMAD7 (47, 48). This in turn forms a feedback loop and may coordinate with SMAD7 to remodel chromatin structure because SMAD7 interacts with DNA through an MH2 domain (49).

Monitoring early γH2AX nuclear focus formation and other IRIF is a fairly accurate means to estimate the protein dynamics during DSB recognition and repair (5052). In experiments using different types of IR-induced DSBs, phosphorylated SMAD2 (at Ser465 and Ser467) and SMAD7, but not SMAD3, co-localize at IRIF with γH2AX and DDR proteins that are recruited to DSB sites (including 53BP1, pATF2, and RAD51) in human epithelial and fibroblast cells (53). The decay of SMAD-containing foci was similar to that of γH2AX foci, as was the spatial localization and the delayed disappearance compared with γ-rays of both phosphorylated SMAD2 and SMAD7 foci along high linear energy transfer (LET) particle tracks, similar to other DDR proteins. These observations support the specificity of SMAD-containing foci at sites of DNA damage. However, SMAD7 foci formed as early as 1 hour after irradiation and appeared in cells in all phases of the cell cycle, whereas phosphorylated SMAD2 foci were not detectable until 4 hours after exposure, were observed primarily in G1 cells, and colocalized with the homologous recombination repair protein RAD51 in G2 cells (53).

The relative appearance of SMAD7 foci perhaps indicates direct binding to DNA at or near DSB breaks because other studies show that it can bind to the DNA elements containing the minimal SMAD-binding element CAGA box (54). In contrast, phosphorylated SMAD2, which is unable to directly bind DNA, is likely to be indirectly localized to DSB sites through interactions with other repair molecules at a later stage, perhaps as a result of chromosome remodeling during repair or an additional role in the activation of transcription. The observation that phosphorylated SMAD2 foci are mainly observed in G1 cells might be explained by the primary role of phosphorylated SMAD2 in the nonhomologous end joining (NHEJ) pathway or its role in the G1-S checkpoint in addition to its role as a transcription factor. SMAD2/3 and p53 physically interact, which suggests that p53 activation might serve as a bridge connecting TGFβ signaling and the IR response (55, 56).

To characterize the potential role for phosphorylated SMAD2 in the DDR, phosphorylated SMAD2 relocalization kinetics after DNA damage induction was studied as a function of ATM kinase activity. ATF2 is a phosphorylation target of ATM at Ser490/498 (39, 40). Phospho-ATF2 foci are not observed in cells treated with an ATM kinase inhibitor or in A-T cells, supporting the notion that ATF2 activation is dependent on ATM. Similar to pATF2, phosphorylated SMAD2 focus formation was completely blocked by addition of the ATM kinase inhibitor as well as in A-T cells, indicating that phosphorylated SMAD2 foci require ATM kinase activity in response to radiation-induced DNA damage. A small-molecule inhibitor of TβRI did not diminish phosphorylated SMAD2 focus formation, which suggests that the fraction of phosphorylated SMAD2 at IRIF is ATM-dependent but TβRI-independent. Although the colocalization of SMAD3 with DSB proteins had been observed previously in response to very high doses (more than 10 Gy) of radiation or other DNA-damaging agents (57), neither total nor phosphorylated SMAD3 was observed to form foci and colocalize with other DDR proteins in IRIF induced by lower doses.

In contrast, SMAD7 foci formed promptly after radiation but were neither inhibited upon ATM inhibition nor diminished in A-T cells (53). However, TβRI inhibition ablated SMAD7 localization to IRIF, indicating that it is dependent on TGFβ signaling (53). An additional role for SMAD7 is suggested by a SMAD7-dependent increase in the phosphorylation of ATM at Ser1981 in prostate cancer cells stimulated with TGFβ (49). γH2AX and SMAD7 colocalize after TGFβ treatment, suggesting that SMAD7 may act as a scaffold for ATM and its substrate H2AX (49). A recent study has shown in mice that Smad7 enhances cell survival against DNA damage by accelerating ATM-dependent DNA repair signaling (58). Here, deletion of Smad7 in mouse embryonic fibroblast cells decreases the activation of ATM and inhibits the recruitment of ATM to sites of DSBs by blocking the interaction between ATM and Nbs1, a member of the MRN complex. A model of the roles of SMAD2 and SMAD7 in the DDR suggests that complex crosstalk occurs between the TGFβ-SMAD and ATM response pathways (Fig. 1A).

Fig. 1 Schematic of signaling that is mediated by TGFβ in response to DNA damage.

(A) IR and other agents that elicit DSB, activate the DDR that is mediated by the kinase activity of ATM. ATM phosphorylates itself and myriad other substrates, including p53, RAD17, CHK2, and BRCA1, that dictate cellular decisions to repair, arrest, or die. ATM also participates in the recognition of DSBs, marking them by phosphorylating H2AX (γH2AX). Concomitantly, TGFβ is activated from the latent form (LTGFβ) to enable binding to its signaling receptors, type I and type II, which leads to the phosphorylation of receptor-mediated SMADs (SMAD2 and SMAD3) and complex formation with SMAD4. The translocation of the SMAD complex to the nucleus initiates gene transcription, including that of SMAD7, which acts to inhibit SMAD2/3 signaling. (B) Recent studies have shown that TGFβ is necessary for DSB recognition and ATM activation in irradiated epithelial cells, and that SMAD7 localizes to early IRIF at DSBs. TGFβ inhibition compromises the autophosphorylation and kinase activity of ATM and inhibits γH2AX and SMAD7 IRIF formation.

In addition to colocalizing with γH2AX in IRIF, SMAD7 foci are also present in spontaneous or radiation-induced micronuclei (MN) (59). Human prostate cancer cells carry a higher proportion of SMAD7-positive MN compared to human normal epithelial cells and fibroblasts (59). A higher frequency of MN was detected in heavily irradiated cells, indicating that MN may contain chromatin with unrepaired and complex DNA damage. SMAD7 is suggested to be a potential oncogene by studies showing endogenous overexpression of SMAD7 in skin, pancreatic, or colon cancer cells and that ectopic overexpression of SMAD7 can induce malignancy (60, 61). A role for SMAD7 in the activation of ATM is also indicated by a SMAD7-dependent increase in Ser1981-phosphorylated ATM and TGFβ-induced cell cycle arrest and genetic stability in prostate cancer cells stimulated with TGFβ (49). The frequency of MN with γH2AX or SMAD7 foci 24 hours after IR is significantly increased compared with MN without γH2AX or SMAD7 foci in human prostate cancer PC3 cells (59). Compared to normal epithelial cells and fibroblasts, spontaneous MN occur more frequently in PC3 cells, possibly because of loss of the G2-M checkpoint as a result of p53 mutations, which enables cells to enter mitosis with unrepaired DSB. Phosphorylated ATM was also found in spontaneous and irradiation-induced MN in PC3 cells (59); however, only a small portion of newly formed MN contain phosphorylated ATM foci 24 hours after high LET particle radiation. Unlike SMAD7, phosphorylated SMAD2 does not appear in MN induced by radiation.

TGFβ inhibition in cancer therapy

Given its control of the DDR, the translational potential of TGFβ inhibition in the context of radiotherapy is promising (62). One of the most striking phenotypes of A-T cells is extreme cellular radiosensitivity demonstrated in clonogenic survival assays. Consistent with reduced ATM kinase activity, Tgfb1 deletion in murine cells or inhibition of TGFβ signaling in human cells increases radiosensitivity. Most human and murine cancer cell lines treated with a small molecular TβRI inhibitor are more sensitive to irradiation, with the dose of IR needed to reduce clonogenic survival to 10% is up to 70% less when TGFβ is inhibited before irradiation (63, 64). Consistent with impaired ATM kinase activity, ATM autophosphorylation and γH2AX focus formation are also decreased in cells treated with a TβR1 inhibitor before IR compared with those treated only with IR. Tumors irradiated in mice treated with preclinical pan-neutralizing TGFβ antibodies show reduced γH2AX foci and increased tumor growth delay. Thus, even though most epithelial cancers are resistant to TGFβ-mediated control of growth, TGFβ appears to be required in these cells to effectively mount a DDR and survive.

An interesting twist on the role of TGFβ in cancer comes from the work of Bhowmick and colleagues, who showed that stromal abundance of TβRII is decreased in 70% of prostate cancers because of increased IL-6. Loss of TβRII in mouse tumors then leads to Dnmt1-dependent methylation of DDR proteins (65). The authors propose that an increased amount of DNA damage caused by DDR gene silencing in the prostatic stromal cells promotes tumor progression, leading them to conclude that TGFβ safeguards DNA fidelity by epigenetic regulation in fibroblastic cells.

There is a growing body of evidence that TGFβ is a therapeutic target in glioblastoma multiforme (GBM), a cancer characterized by a high degree of radioresistance and inevitable local or disseminated recurrence. Several clinical trials are under way testing the efficacy of combining TGFβ inhibition with radiotherapy and chemotherapy, including a phase 2 trial in glioblastoma patients. Mengxian et al. reported GBM radiosensitization, tumor growth delay, and improved patient survival with the addition of the small-molecule inhibitor of TβRI kinase, LY2109761 (66). The magnitude of radiosensitization by LY2109761 is similar to that induced by cotreatment with temozolomide (67). Zhang et al. reported that the addition of LY2109761 to the current standard therapy (radiation and the oral alkylating agent temozolomide) provided significant benefit and inhibited angiogenesis and glioma cell migration (68). Also, Mengxian et al. reported that TGFβ inhibition and irradiation synergize to decrease the self-renewal of glioma stem-like cells in a neurosphere assay (66). Hardee et al. demonstrated that glioma-initiating cells produce high amounts of autocrine TGFβ that potentiates their effective molecular DDR, whereas irradiation-induced TGFβ mediates self-renewal signals (64). These initial findings suggest that TGFβ inhibition can provide specific benefit in the context of radiotherapy, particularly by acting on ATM (Fig. 1B).


System biology modeling of the intersection of TGFβ signaling and the DDR suggests some degree of retroactivity (69, 70), meaning that downstream components affect the dynamic state of the upstream component (70). It has been postulated that modular structures found in biological systems enable basic functions of a module to be robust to change; however, modifications in molecular connections between modules lead to alterations in cell and tissue phenotypes (69, 70). As discussed above, mutation of TGFβ itself is far less frequent than loss of response and increased abundance of TGFβ in cancer. This can be explained by evoking retroactivity in TGFβ signaling to its large and diverse downstream components. For example, the various inactivation mutations of SMADs observed in human cancers lead to a much greater potential for signaling rewiring in cancer cells through their retroactivity on TGFβ function than single-gene deletions alone would suggest. Furthermore, the role of SMAD7 and other SMAD proteins in the canonical pathway and the recent observations of their interplay with the DDR (29, 49, 57, 59) along with retroactivity suggest a mechanism that likely extends the reach of TGFβ into areas not considered in the past.

The conundrum of why tumors maintain TGFβ abundance and signaling when it is an extremely potent growth inhibitor gains clarity when control of genome stability is incorporated. Cells maintain genomic stability in the face of relentless challenges by environmental stresses that induce DNA breaks and activate DDR pathways mediated by ATM and its downstream effectors that lead to damage-induced cell cycle checkpoints and DNA repair (32). Early malignant lesions in the breast accumulate genomic damage, evidenced by abnormal centrosomes, chromosome instability, and activation of the DDR, often without mutations in the gene encoding p53 (71, 72). Women with “low signaling” TGFB1 polymorphisms have a twofold increased risk for breast cancer (73, 74). It is thought that GIN markedly increases the likelihood of malignant transformation (72). Studies by Glick and colleagues were the first to show that Tgfb1-null murine keratinocytes exhibit markedly increased GIN measured by gene amplification (20, 75). Centrosome aberrations and tetraploidy are also significantly increased with age in Tgfb1-heterozygote mammary epithelium, and TGFβ inhibition in human epithelial cells increases centrosome aberrations, aneuploidy, tetraploidy, and spontaneous DNA damage, all measures of GIN (21). These data support the idea that attenuated TGFβ signaling may predispose to cancer by increasing genomically unstable epithelial cells. Tgfb1-heterozygote mice stressed with increased expression of oncogene or exposure to chemical carcinogens exhibit increased tumor incidence and size (76) as well as decreased tumor latency (77, 78). Thus, attenuated TGFβ signaling can amplify the possibility of neoplastic transformation by decreasing genome stability.

However, although some degree of instability appears to promote tumorigenesis, cancer cells that have high GIN fail to progress (79). The necessity for TGFβ signaling to maintain genomic stability, and the recognition that TGFβ regulates ATM kinase and other DDR components, suggests that, although TGFβ acts to initially suppress tumorigenesis by inhibiting proliferation and endorsing genomic integrity, maintenance of TGFβ signaling protects malignant cells both by limiting GIN and by enabling recovery from radiation and other therapeutically induced DNA damage. Thus, the growing evidence of the importance of TGFβ signaling in mediating the response to DNA damage and genome stability opens a new perspective on TGFβ biology in carcinogenesis.


Funding: The authors are supported by funding from the Low Dose Radiation Program of the Office of Biological and Environmental Research (M.H.B.-H. and F.A.C.) and Varian Medical Systems Inc. (M.H.B.-H.). Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Signaling

Navigate This Article