Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. 2009 May; 29(9): 2443–2455.
Published online 2009 Feb 17. doi:  10.1128/MCB.01443-08
PMCID: PMC2668365

Transforming Growth Factor β Depletion Is the Primary Determinant of Smad Signaling Kinetics


A cell's decision to growth arrest, apoptose, or differentiate in response to transforming growth factor β (TGF-β) superfamily ligands depends on the ligand concentration. How cells sense the concentration of extracellular bioavailable TGF-β remains poorly understood. We therefore undertook a systematic quantitative analysis of how TGF-β ligand concentration is transduced into downstream phospho-Smad2 kinetics, and we found that the rate of TGF-β ligand depletion is the principal determinant of Smad signal duration. TGF-β depletion is caused by two mechanisms: (i) cellular uptake of TGF-β by a TGF-β type II receptor-dependent mechanism and (ii) reversible binding of TGF-β to the cell surface. Our results indicate that cells sense TGF-β dose by depleting TGF-β via constitutive TGF-β type II receptor trafficking processes. Our results also have implications for the role of the TGF-β type II receptor in disease, as tumor cells harboring TGF-β type II receptor mutations exhibit impaired TGF-β depletion, which may contribute to the overproduction of TGF-β and a consequently poor prognosis in cancer.

Transforming growth factor β (TGF-β) is the prototypical cytokine of a namesake superfamily of cytokines that regulate diverse aspects of cellular homeostasis. TGF-β signaling begins with the binding of a ligand dimer to two type II TGF-β receptors (TβRII), followed by binding of this complex to two type I receptors (TβRI) (52). The TβRII is a constitutively active kinase that phosphorylates residues within the GS domain of the TβRI (52). Upon activation, the TβRI exhibits increased kinase activity toward the intracellular Smad transcription factors (52). Eight Smad isoforms exist, which are functionally classified as receptor-regulated Smads (R-Smads; Smad isoforms 1, 2, 3, 5, and 8), the common mediator Smad (co-Smad; Smad isoform 4), and inhibitory Smads (I-Smads; Smad isoforms 6 and 7) (34). In the absence of TGF-β signaling, the Smads constitutively shuttle between the cytoplasm and nucleus, with predominant localization in the cytoplasm (40, 45). During TGF-β signaling, Smads 2 and 3 are phosphorylated by the TβRI, which facilitates their binding to Smad4. Smad complexes accumulate in the nucleus, where they carry out transcriptional regulation of TGF-β target genes. Within the nucleus, the Smad complexes reversibly dissociate and the monomeric phospho-R-Smads are dephosphorylated by a nuclear phosphatase (29), upon which they join the pool of R-Smads available for nuclear export. The cycle of Smad activation and deactivation persists for as long as receptors are active (19).

The responses of cells to TGF-β depend on the ligand concentration to which they are exposed (10). Several members of the TGF-β superfamily are morphogens, which are secreted molecules that determine the developmental fate of cells based on concentration (3). Cells are exquisitely sensitive to morphogen concentration, such that subtle differences in the concentration can induce different cell fates. For example, in Xenopus laevis (frog) development, five distinct cell fates are determined by scarcely overlapping ranges of Activin concentration within a total span of 0 to 20 units/ml (17). Therefore, cells are somehow able to sense, or to “read,” the concentration of TGF-β ligands at the exterior of the cell and orchestrate a specific response. How cells read, interpret, and respond to TGF-β concentration is thus a question of important relevance to understanding TGF-β biology.

Once a cell reads the TGF-β concentration, it must interpret the signal and transduce it inside the cell. Quantitative studies of Activin signal transduction in dissociated Xenopus cells indicate that the signal is transmitted directly from the plasma membrane to the nucleus. Specifically, the Activin cue is transduced into an absolute number of active receptor complexes (13), whose kinase activities phosphorylate the Smads at a rate proportional to the number of active receptors (6). In turn, the phosphorylated Smads accumulate in the nucleus and regulate gene expression, with the degree of Smad nuclear accumulation directly proportional to Activin concentration (6). The concentration of Smad complexes in the nucleus determines which Activin target genes are regulated (53), whereby the genes are thought to be activated in an all-or-none manner once a signal strength threshold is met (3, 49). Activin target genes then mediate the cell response. Therefore, Smad dynamics appears to be a direct readout of the Activin cue, whereby the Activin dose specifies the strength and duration of the Smad signal, which subsequently determines the genes whose expression is regulated and the ultimate cell response. Analogous studies conducted with bone morphogenetic protein (BMP) ligands have conferred similar results (54), suggesting that TGF-β superfamily signals are interpreted in a similar manner.

The cell has various means to regulate Smad signal intensity and duration, with loss of functional receptors through negative feedback and Smad dephosphorylation in the nucleus being the predominant hypothesized mechanisms for terminating the Smad signal (21). Receptors are downregulated at the cell surface in the presence of TGF-β (58, 61). Receptors are constitutively degraded via the lysosomal and ubiquitin-proteosome pathways, depending on whether the receptors are internalized in clathrin-coated pits or caveolae (11, 22, 36). Smad7 is thought to be a crucial mediator of receptor deactivation and degradation, because Smad7 can recruit phosphatases and the Smurf2 ubiquitin ligase to the receptor complex and can also competitively bind to the TβRI to prevent R-Smad binding (21, 39, 51). While potential molecular mechanisms for receptor deactivation have been well-characterized, the physiological relevance of receptor deactivation for regulating the strength and duration of Smad signaling has yet to be verified. Specifically, downregulation of receptors at the cell surface does not necessarily imply loss of activity, because receptors can continue to signal from the endosomes (20, 42) and many of the conclusions about Smad7-related inhibition have been based on overexpression studies (49). As for Smad dephosphorylation, knockdown experiments have shown that the duration of the Smad signal inversely correlates with the abundance of the putative Smad phosphatase, PPM1A/PP2C (29). However, the activity of PPM1A/PP2C does not appear to change during signaling (29), suggesting that under endogenous conditions the observed decline of phospho-Smad levels (19) is due to some other factor.

An overlooked aspect of TGF-β signal transduction is the strength and duration of the input itself. In most TGF-β studies involving cultured cells, a bolus of TGF-β of known concentration is added to the cells at the start of the experiment. Thereafter, how the concentration of TGF-β varies in the medium has not been systematically studied. While it is known that TGF-β is internalized and degraded by cells (33), it is unknown whether this activity significantly impacts the TGF-β concentration in the medium. Published mathematical models of TGF-β signaling assume that TGF-β concentration remains constant in the medium (35, 57). However, this assumption was untested in both studies, such that if TGF-β levels do not remain constant over time, then the validity of the model predictions might be compromised. Progress in developing predictive systems models of TGF-β signaling depends on correct assumptions about the input signal. In addition, studies attempting to attribute specific phenotypes or gene expression programs to properties of the intracellular signal would benefit from knowledge of the quantitative relationship between the input and intracellular signals. Therefore, studies of TGF-β input dynamics are urgently needed.

In this paper, we studied the properties of the TGF-β input and its interpretation at the level of the Smad signal in order to obtain insight into the mechanism by which cells read TGF-β concentration. Specifically, we found that the potency of a given TGF-β dose depends on the number of cells to which it is applied. This phenomenon results from the cells depleting TGF-β from the culture medium, and the duration of the subsequent Smad signal correlates with the duration of TGF-β presence in the medium. TGF-β depletion is mediated by the TβRII and reversible binding to the cell surface. Furthermore, we found that neither receptor loss nor alterations to the rate of Smad dephosphorylation account for dose-dependent Smad kinetics. We therefore conclude that TGF-β depletion principally determines Smad signal kinetics.


Cell lines and reagents.

PE25 cells are mink lung epithelial cells (Mv1Lu or CCL-64) that harbor a stably transfected luciferase reporter gene (18). HCT116 and HCT116+TβRII cells were a generous gift of LuZhe Sun. All cell lines were maintained using standard procedures and incubated at 37°C with 5% CO2. PE25, HaCaT, HeLa S3, DR27, and R1B cells were cultured in Dulbecco's modified Eagle medium, and HCT116 cell lines were cultured in McCoy's 5A medium. All media were supplemented with 10% fetal bovine serum, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine.

Recombinant human TGF-β1 (isoforms 1, 2, and 3) was obtained from R&D Systems (240-B, 302-B2, and 243-B3). The TβRI inhibitor SB-431542 (Tocris) was dissolved in dimethyl sulfoxide and used at a concentration of 10 μM. Three antibodies against phospho-Smad2 were used at dilutions ranging from 1:150 to 1:250 and were obtained from the following sources: (i) a generous gift from Aris Moustakas (Uppsala University), (ii) Cell Signaling Technology (3101), or (iii) Cell Signaling Technology (3108). Anti-Smad2 antibody was a generous gift from Ed Leof (Mayo Clinic) and was used at a 1:1,000 dilution. Anti-α-tubulin antibody (catalog no. 691251; MP Biomedicals) was also used at a 1:1,000 dilution. Secondary antibodies against rabbit and mouse immunoglobulin G were purchased from Amersham (NA934V [used at dilutions ranging from 1:3,000 to 1:5,000] and NA931V [used at a 1:6,000 dilution]).

Luciferase reporter gene activity assay.

For the luciferase reporter gene activity assay, cells were seeded in 12-well plates (200,000 cells/well) and were lysed with 200 μl passive lysis buffer (catalog no. E1501; Promega) and frozen at −80°C for >2 h. Cells were then scraped into microcentrifuge tubes and centrifuged at 13,200 rpm for 10 min at 4°C. One hundred microliters of lysate was combined with 50 μl of substrate (catalog no. E1501; Promega) and the resulting luminescence was read using a Dynex Technologies MLX microtiter plate luminometer. The linear dynamic range of the assay was confirmed in control experiments using serial dilutions of concentrated lysate.

TGF-β reporter assay.

A TGF-β reporter assay was used to measure TGF-β concentration in the culture medium, similar to the approach of Abe et al. (1). The assay uses reporter cells (PE25 cells, which express luciferase in a TGF-β dose-dependent manner [see Fig. Fig.1B,1B, below]) to measure TGF-β concentration remaining in the culture medium from a separate group of test cells. Specifically, at the end of an experimental treatment, 0.75 ml of the test cell medium was transferred onto the reporter cells seeded in 12-well plates (200,000 cells/well). In parallel, media containing known concentrations of TGF-β were added to a separate group of reporter cells in order to generate a standard curve. The PE25 cells were incubated for 24 h and processed for luciferase measurements. The standard curve was generated by fitting the luciferase activities as a function of TGF-β concentration using a three-parameter Hill equation. Numerical solution was then used to interpolate the TGF-β concentrations in the media of the experimental samples (Microsoft Excel Solver tool).

FIG. 1.
Ligand concentration specifies the cellular responses to TGF-β. (A) Schematic of the luciferase reporter assay. Cells possess a stably integrated luciferase reporter gene driven by a Smad-sensitive promoter. Addition of TGF-β induces production ...

Immunoblotting analysis.

Cells were frozen in N2 upon completion of each experiment. Cell lysis procedures were carried out at 4°C. Cells were lysed in lysis buffer (50 mM Tris-HCl, 400 mM NaCl, 1 mM EDTA, 1% NP-40, 15% glycerol, 2 mM sodium fluoride, 1 mM sodium orthovanadate, and 1× protease cocktail [catalog no. 1697498; Roche]), scraped into microcentrifuge tubes, rotated for 45 min, and then centrifuged at 13,200 rpm for 15 min. Lysate protein concentration was measured by the bicinchoninic acid method (catalog no. 23225; Pierce). Equal protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotted onto a nitrocellulose membrane (Whatman Protran BA 83 10-401-396). Membranes were blocked for 60 min with 4% nonfat milk in Tris-buffered saline-0.05% (vol/vol) Tween (TBS-T), incubated in primary antibody for 2 h, washed three times for 5 min in TBS-T, incubated in secondary antibody for 1 h, and washed three times for 10 min in TBS-T. Signal was detected by chemiluminescence (either ThermoScientific SuperSignal West Dura [34076] or Amersham ECL Plus [RPN 2132]). To obtain appropriately exposed blots, several exposures of various durations were captured for each blot. Optical band densities were quantified using ImageJ (NIH).

Quantitative immunoblotting.

We used quantitative immunoblotting to estimate the absolute numbers of phospho-Smad2 molecules per cell. We used recombinant phospho-MH2 (P-MH2) polypeptides as a standard. The P-MH2 was generated using expressed protein ligation (38), in which the Smad2 MH2 domain (residues 241 to 462) lacking the C-terminal phosphorylation motif is expressed in bacteria and a chemically synthesized phosphorylated C-terminal peptide is chemically ligated to the MH2 domain (60). The reaction proceeds to completion such that virtually 100% of the recombinant protein is phosphorylated, although the P-MH2 protein appears as a doublet, which may reflect singly phosphorylated P-MH2 species (60). We quantified the P-MH2 stock concentration using Coomassie staining with bovine serum albumin standards whose concentration was quantified using the extinction coefficient method (catalog no. 23209; Thermo Scientific Pierce). We performed optimization experiments to find a range of dilutions of the P-MH2 that conferred immunoblot band densities that spanned the observed range of phospho-Smad2 from the cell lysates. To control for any effects of cell lysis on the detectability of phosphorylated Smad2, we diluted the P-MH2 domain into PE25 cell lysate that had been treated with 100 pM TGF-β for 50 min. Serial dilutions of the P-MH2 in the lysate were used as standards from which the phospho-Smad2 content in the cell lysates could be interpolated using a power law function (46; D. Clarke and X. Liu, submitted for publication).

Cell growth assay.

Growth inhibition assays were performed using PE25 cells plated at 15,000 cells per well in six-well plates. After incubation for 8 to 18 h to permit cell adherence, the medium was replaced with fresh medium containing 0, 1, 5, 10, 25, or 50 pM TGF-β. Medium and TGF-β were replaced every 3 to 4 days. After 2 weeks, the cells were washed with phosphate-buffered saline, fixed with 70% ethanol for 1 hour, crystal violet stained for 1 hour, and stored in water with 10% glycerol.


An enzyme-linked immunosorbent assay (ELISA) kit for human TGF-β1 detection was purchased from R&D Systems (DB100B) and used according to the manufacturer's instructions.


TGF-β concentration is the signal to which cells respond.

To demonstrate the effect of ligand concentration on the cellular response to TGF-β, we performed two experiments. First, we established the dose-response relationship between TGF-β concentration and luciferase reporter activity in mink lung epithelial cells stably infected with a luciferase reporter gene driven by a Smad-sensitive promoter (hereafter referred to as PE25 cells [30]) (Fig. (Fig.1A).1A). We found that luciferase reporter gene expression varied as a sigmoidal function of the log10 of TGF-β concentration, with a dynamic range of ∼1 pM to ∼50 pM (Fig. (Fig.1B).1B). Second, TGF-β-mediated growth inhibitory responses in PE25 cells are also concentration dependent, with crystal violet-stained cells visible only after exposure to 0 and 1 pM TGF-β, but not higher concentrations (Fig. (Fig.1C).1C). Therefore, cell responses to TGF-β are concentration dependent, implying that the signal to which cells respond is TGF-β concentration.

The Smad signal is a function of the number of TGF-β molecules per cell.

Interestingly, the growth response and luciferase reporter assays revealed differing sensitivities to the concentrations of TGF-β. This result was particularly puzzling since both the luciferase reporter activity and inhibitory growth response depend on transcriptional regulation driven by Smad-sensitive promoters, such that a similar strength of input should lead to similar magnitudes of response. A possible source for the discrepancy is that the assays employ different experimental conditions, implying that TGF-β concentration, expressed in moles per unit volume, is insufficient to specify the Smad signal. Alternatively, the number of TGF-β molecules per cell might be the variable that determines the Smad signal. To address this possibility, we performed a two-level factorial experiment in which four experimental parameters (TGF-β concentration, cell number seeded, plate type, and medium volume) were varied in all possible combinations (Table (Table1).1). Note that each of the variables affects the number of TGF-β molecules per cell. Phospho-Smad2 levels were measured by immunoblotting at 30 min and 8 h after addition of TGF-β (Fig. (Fig.2A).2A). Quantitation of the immunoblot assay data showed reduced scatter in the data points when the phospho-Smad2 levels were plotted versus the number of TGF-β molecules per cell instead of TGF-β concentration alone for both time points (compare Fig. 2B and C). Therefore, we conclude that TGF-β dose expressed as molecules per cell is a better predictor of the phospho-Smad2 signal than TGF-β concentration per unit volume. This result has two implications: (i) that cells can interpret absolute numbers of TGF-β molecules per cell and (ii) TGF-β potency (as reflected by the intensity of the phospho-Smad2 signal) depends inversely on the number of cells present. We also used this result to standardize the conditions for subsequent experiments, choosing to use six-well plates seeded with 1.5 × 106 cells per well and 1.5 ml medium.

FIG. 2.
The number of TGF-β molecules per cell predicts the Smad signal. (A) Immunoblots of the indicated proteins in whole-cell lysates of PE25 cells exposed under the indicated conditions. (B and C) Plots of the ratio between the immunoblot band densities ...
Design of the factorial experimenta

TGF-β is depleted from the culture medium during signaling and the presence of TGF-β correlates with the duration of Smad signaling.

The data above indicate that the reduced potency of TGF-β with higher cell density is more pronounced at 8 h than at 30 min, suggesting that the cells might actively reduce the potency of TGF-β over time. Previous studies have shown that cells internalize and degrade TGF-β (33); however, the effect of this degradation on the amount of TGF-β in the culture medium was not addressed. We hypothesized that TGF-β depletion from the cells' environment could be a way to reduce the potency of TGF-β over time in order to control the duration of Smad signaling.

To test this hypothesis, we measured the time courses of TGF-β depletion and phospho-Smad2 levels in response to three TGF-β doses: 10, 25, and 200 pM, which correspond to 6,020, 15,050, and 120,400 TGF-β molecules per cell under the experimental conditions listed above. TGF-β depletion was measured using our TGF-β reporter assay (see Materials and Methods), for which results of control and validation experiments are shown in Fig. 3A and B. In accordance with our hypothesis, TGF-β was depleted from the culture medium for each initial dose (Fig. (Fig.3C).3C). To confirm that TGF-β depletion occurs with cell types other than just PE25 cells, we performed the depletion experiment with HaCaT and HeLa S3 cells, which are both TGF-β-sensitive cell lines. Both cell types deplete TGF-β with kinetics similar but not identical to those of PE25 cells (Fig. (Fig.3D).3D). We also sought to confirm that depletion was not exclusive to the TGF-β1 isoform by testing whether TGF-β2 and TGF-β3 isoforms were also depleted. We found that PE25 cells deplete both isoforms similar to, but slower than, TGF-β1 (Fig. (Fig.3E).3E). Therefore, we conclude that TGF-β depletion is a property of TGF-β-responsive cell lines and of all TGF-β ligands.

FIG. 3.
TGF-β is depleted from the culture medium during TGF-β signaling. (A) The TGF-β reporter assay specifically measures TGF-β in conditioned medium. PE25 cells were exposed to either fresh medium or 25 pM TGF-β for ...

To determine how the phospho-Smad2 signal varies as a function of TGF-β dose, we used quantitative immunoblotting to measure phospho-Smad2 time courses in lysates of the cells that were used in the depletion experiment above. We observed that the signal amplitudes (defined as the maximal levels of phospho-Smad2 during the time course) were similar between the 10 and 25 pM TGF-β groups (Fig. 4A and B), whereas the signal duration (defined as the phospho-Smad2 levels at the later time points) was prolonged in the 25 pM samples in the 4- to 8-h time span compared to the 10 pM samples (Fig. 4A and B). Phospho-Smad2 amplitude and duration were both elevated in response to 200 pM TGF-β compared with the lower doses. Strikingly, the kinetics of the decay in the phospho-Smad2 signal correlate with those of TGF-β depletion. For example, comparing Fig. Fig.3C3C and and44 for the 10 pM TGF-β group reveals that the depletion of TGF-β by 6 to 8 h coincides with the time of disappearance of the phospho-Smad2 signal. Conversely, some TGF-β still remained after 24 h in the 200 pM group, and this coincided with the continued presence of elevated phospho-Smad2 (Fig. (Fig.3C3C and and4).4). Furthermore, the duration of the Smad signal correlates with the duration of Smad nuclear activity, with luciferase reporter activity increasing steadily until the Smad signal ends (note the different time scale in Fig. Fig.4C).4C). The correlation between the abundance of TGF-β in the culture medium and the presence of elevated phospho-Smad2 levels suggests that the rate of TGF-β depletion could be a key regulator of Smad signal duration and subsequent Smad nuclear activity.

FIG. 4.
Phospho-Smad2 time course as a function of TGF-β dose. (A) Raw immunoblot assay data, showing phospho-Smad2 (P-Smad2) levels at the indicated times in response to 10, 25, and 200 pM TGF-β. In each gel, known amounts of the phosphorylated ...

TGF-β depletion is caused by a TβRII-dependent mechanism and reversible binding to the cell surface.

TGF-β receptor trafficking is an important feature of TGF-β signaling. TGF-β-receptor complexes are internalized into the early endosomal compartment of the cells, followed either by recycling of the receptors back to the plasma membrane or lysosomal degradation (36). Receptors may also be degraded via the caveolar pathway (11). We hypothesized that the mechanism of TGF-β depletion depends on TGF-β binding to its cognate receptors, followed by internalization and subsequent degradation of the ligand. To investigate the mechanism of TGF-β depletion, we first assayed the ability of cells lacking functional TGF-β receptors to deplete TGF-β from the medium. Specifically, we employed the chemically mutagenized clones of Mv1Lu cells that lack functional TβRI (R1B cells) and TβRII (DR27 cells) (24). In both cases, the mutations introduced to either receptor prevent binding of TGF-β to that receptor such that performing the depletion assay with these cells should reveal the dependence of depletion on the TGF-β receptors. In response to an initial dose of 25 pM TGF-β, we observed that depletion was impaired in DR27 cells, but not R1B cells, compared to PE25 cells (Fig. 5A and B). Therefore, the TβRII appears necessary for TGF-β depletion, whereas depletion is independent of the TβRI. (This result is further supported by the TβRII rescue experiment presented below in Fig. Fig.88).

FIG. 5.
TGF-β depletion is mediated by a TβRII-dependent mechanism and reversible binding to the cell surface. (A) The TβRII, but not the TβRI, is necessary for TGF-β depletion. TGF-β depletion time courses (25 ...
FIG. 8.
TβRII mutations in cancer impair TGF-β depletion. (A) TGF-β depletion is impaired in HCT116 cells and restored upon expression of the TβRII. Cells were exposed to a 25 pM concentration of TGF-β for 8 h, with TGF-β ...

Although depletion in DR27 cells is impaired, it is not completely eliminated. Partial depletion occurs in DR27 cells, with depletion kinetics that mirror those of PE25 cells up to about 60 min, after which depletion ceases and a steady state of TGF-β concentration ensues (Fig. (Fig.5B).5B). Such behavior is consistent with a reversible binding mechanism whereby equilibrium establishes after about 60 min. To test this hypothesis, we performed a washout experiment in which we applied an initial dose of 25 pM TGF-β to DR27 and PE25 cells for 60 min, followed by exchanging the medium with fresh medium containing no TGF-β. If reversible binding occurs, then removal of free TGF-β should drive the equilibrium toward dissociation and TGF-β should reappear in the fresh medium. In accordance with our hypothesis, TGF-β quickly reappeared in the medium (by 5 to 10 min) and a steady-state concentration of about 5 pM TGF-β remained in the medium for at least 4 h (Fig. (Fig.5C).5C). When the same treatment was applied to PE25 cells, TGF-β reappeared in the medium but then decreased over time, reflecting continued TGF-β depletion (Fig. (Fig.5C).5C). These results are consistent with reversible binding of TGF-β to the cell surface. To verify that TGF-β reversibly binds to the cell surface in PE25 cells, we performed a depletion time course with PE25 cells at 37°C, the temperature at which our experiments are normally performed, and at 4°C. The cold temperature blocks endocytic processes and hence receptor internalization, which should allow us to isolate whether partial depletion due to reversible binding occurs in PE25 cells. As expected, partial depletion of TGF-β occurred at 4°C in a manner similar to the DR27 cells (Fig. (Fig.5D).5D). This result confirmed that TGF-β depletion in PE25 cells results from at least two processes: (i) an active TβRII-dependent mechanism involving receptor internalization, and (ii) reversible binding to the cell surface.

TGF-β receptor function is preserved throughout signaling.

Receptor degradation is a commonly cited mechanism for Smad signaling termination (52, 55). The current view is that Smad7, an inhibitory Smad, deactivates TGF-β signaling in a negative feedback manner by targeting the receptors for degradation (34). Indeed, receptor downregulation from the cell surface has been shown using radiolabeled TGF-β binding assays (58, 61); however, whether such downregulation corresponds to reduced functional capacity of the receptors is not known. Our data show that prolonged Smad signaling accompanies increased TGF-β dose, which suggests that significant receptor loss does not occur during signaling since continuous receptor activity is required to maintain elevated phospho-Smad levels in the presence of TGF-β (19). Therefore, the role of receptor loss in determining the duration of Smad signaling bears examination.

Since the rate of Smad phosphorylation depends directly on the levels of functional receptor complexes, we used a functional approach to ascertain whether loss of receptors is actually occurring during signaling. Specifically, we performed a double TGF-β stimulation experiment (Fig. (Fig.6A).6A). The rationale of the experiment is as follows: if TGF-β signaling is terminated by receptor loss, then the cells should be unresponsive to a second dose of TGF-β. Since TGF-β depletion depends on the TβRII, loss of the TβRII should manifest itself as a reduced rate of TGF-β depletion in response to a second dose of TGF-β. If significant loss of the TβRI occurs, then depletion should be unaffected but Smad2 phosphorylation should be lessened in response to a second dose of TGF-β. Furthermore, if the magnitude of the receptor loss is proportional to ligand dose, then a higher dose of TGF-β should more profoundly reduce cellular responsiveness to a second dose of TGF-β.

FIG. 6.
TGF-β receptor levels are preserved during signaling. (A) Design of the double-stimulation experiment. Two groups of cells were seeded for comparison: a single-stimulation group and a double-stimulation group. First, the indicated dose of TGF-β ...

The data indicate that little or no net loss of receptors accompanied 8 h of TGF-β signaling. TGF-β depletion occurred at the same rate in response to a dose of 25 pM TGF-β, regardless of whether cells were preexposed to either 25 or 200 pM TGF-β (Fig. (Fig.6B).6B). Similarly, a second dose of 25 pM TGF-β restored phospho-Smad2 levels in cells preexposed to 25 pM TGF-β (Fig. (Fig.6C).6C). (We did not perform the same analysis for cells preexposed to 200 pM TGF-β because the phospho-Smad2 levels from the first dose would remain elevated throughout the second 8 h [Fig. [Fig.4].)4].) The restoration of phospho-Smad2 levels by a second dose of TGF-β was eliminated by applying the SB-431542 inhibitor (Fig. (Fig.6C),6C), implying that the TβRI is responsible for the additional Smad phosphorylation. These results imply that neither the TβRII nor the TβRI are lost to a significant extent in 8 h of TGF-β signaling. Therefore, receptor loss or deactivation cannot account for the observed decrease of Smad phosphorylation levels during signaling (Fig. (Fig.44).

The Smad dephosphorylation rate is preserved during signaling.

Another possible mechanism for the observed phospho-Smad2 kinetics during TGF-β signaling is the regulation of the nuclear phosphatase(s) that dephosphorylates the Smads. If this were the case, the phospho-Smad2 time course data (Fig. (Fig.4)4) indicate that the phosphatase activity would have to be repressed in a manner proportional to TGF-β dose. To ascertain if the observed rate of Smad2 dephosphorylation varies as a function of either TGF-β dose or signaling duration, we measured the kinetics of phospho-Smad2 loss after blocking TβRI activity during signaling. Specifically, PE25 cells were exposed to 25 or 200 pM TGF-β for either 30 min or 6 h, followed by applying TβRI inhibitor (Fig. (Fig.7A).7A). Phospho-Smad2 levels were measured by immunoblotting, and we observed that phospho-Smad2 was virtually completely dephosphorylated under all conditions by 60 min after the TβRI inhibitor was applied (Fig. (Fig.7B).7B). We conclude that neither the TGF-β dose nor signaling duration affects the phospho-Smad2 dephosphorylation rate.

FIG. 7.
Phospho-Smad2 is dephosphorylated at a similar rate throughout signaling. (A) PE25 cells were exposed to either 25 or 200 pM TGF-β for either 30 min or 6 h and then exposed to 10 μM TβRI inhibitor SB-431542. Phospho-Smad2 (P-Smad2) ...

TβRII mutations in cancer impair TGF-β depletion.

Many cancer cell lines possess inactivating mutations in the TβRII, which leads to TGF-β resistance. Our results predict that loss of the TβRII would also impair TGF-β depletion in these cell lines, which could contribute to the well-characterized increase in TGF-β levels both locally in tumors and systemically in cancer patients. We thus decided to investigate whether rescuing TβRII expression could restore the cells' ability to deplete TGF-β.

HCT116 cells are colon cancer cells that harbor a deletion mutation for the TβRII (7, 32). We measured the TGF-β depletion kinetics and phospho-Smad2 levels in HCT116 cells and HCT116 cells stably expressing the TβRII (HCT116+TβRII cells) (59). In response to an initial dose of 25 pM TGF-β, we found that the levels of bioactive TGF-β in the culture medium of HCT116 cells increased over time after a brief initial decrease (Fig. (Fig.8A).8A). This observed increase is consistent with the notion that cancer cells often upregulate TGF-β expression and secretion (43). However, reintroduction of the TβRII into the HCT116 cells reverted the depletion phenotype to one displayed by healthy cells (Fig. (Fig.8A).8A). These results confirm that the TβRII is necessary for TGF-β depletion and that cancer cell lines deficient in TβRII expression exhibit an impaired ability to deplete TGF-β from their environment.


In this study, we discovered mechanisms by which cells read TGF-β concentration and transduce this signal into an intracellular Smad signal. Specifically, we found that the potency of a given concentration of TGF-β depends on the number of cells that are exposed to the TGF-β, such that TGF-β dose is best expressed in units of molecules per cell. The dependence of TGF-β potency on the number of cells in part reflects the continuous depletion of TGF-β by the cells from the medium, such that the duration of the Smad signal is proportional to the dose of TGF-β and inversely proportional to the number of cells present. From a mechanistic standpoint, we found that TGF-β depletion is mediated by a TβRII-dependent mechanism and by reversible binding to the cell surface (Fig. (Fig.8B).8B). Finally, we establish TGF-β depletion as the primary determinant of Smad signal duration, because receptor loss, Smad2 loss, or changes in the phospho-Smad2 dephosphorylation rate do not account for the decrease in phospho-Smad2 levels over time in response to TGF-β. Therefore, under standard cell culture conditions, Smad signaling is terminated predominantly due to the disappearance of ligand. Our results indicate that TGF-β concentration is sensed by constitutive TβRII trafficking processes in which cycling of the receptors to and from the cell surface bind and internalize TGF-β molecules at a constant rate, such that higher concentrations would take longer to deplete. This mechanism is akin to Smad cycling between the cytoplasm and nucleus to sense the number of active receptors (19). Therefore, it appears that TGF-β signaling is regulated by two dynamic cyclic processes to transduce TGF-β dose into a corresponding amount of Smad nuclear accumulation.

Our finding that TGF-β depletion is a principal determinant of Smad signal duration adds significantly to our understanding of the negative regulation of TGF-β signaling. Specifically, we contend that ligand depletion is the primary means by which the Smad signal is terminated. (Here we define signal termination in the macroscopic sense, i.e., the return of the cell's entire TGF-β/Smad signaling system to the basal state, rather than the microscopic sense, i.e., terminating the signals delivered by individual Smad proteins.) Currently, Smad7-mediated negative feedback, which putatively acts at multiple levels within the signaling pathway, and Smad dephosphorylation are the prominent means by which Smad signaling is thought to be inhibited and terminated. We argue that these mechanisms are responsible for inhibiting but not terminating signaling, because our results show that as long as ligand is present, the cells remain competent to TGF-β signals and the Smad dephosphorylation rate does not appear to change. Our contention is further supported by published data, which show that cells remain competent for TGF-β signaling hours to weeks after the initial TGF-β dose (16, 41), long after the putative time course of Smad7 negative feedback has taken place (2, 39, 56). In addition, our results indicate that eliminating the TβRII through Smad7-mediated negative feedback could actually prolong signaling by lessening the rate of TGF-β depletion. Our data argue against a role for Smad7 in TGF-β depletion because the kinetics of TGF-β depletion are similar between wild-type and R1B cells, which lack functional TGF-β type I receptors and hence the capacity to signal and induce Smad7 (48). Therefore, it appears that Smad7 and TGF-β depletion have distinct roles in inhibiting and terminating TGF-β signaling, the details of which await future study. With respect to dephosphorylation, the activity of the putative R-Smad phosphatase, PPM1A/PP2C, does not appear to change in response to TGF-β signaling (29). Therefore, we propose that Smad7-mediated negative feedback and Smad dephosphorylation serve to restrain Smad signaling, while the duration of the presence of ligand determines the duration of Smad signaling.

We infer from our quantitative Smad phosphorylation data that negative feedback and dephosphorylation mechanisms tightly restrain Smad signaling, such that the rate of phosphorylation only slightly exceeds the rate of dephosphorylation during signaling. Specifically, we estimate that only about 20% of the total cellular Smad2 is phosphorylated during signaling, because in this study we observed a maximum amplitude of ∼20,000 phospho-Smad2 molecules per cell in response to 120,400 molecules per cell of TGF-β, out of a total of about 100,000 Smad2 molecules per cell (9). Our results agree reasonably well with imaging data showing that 36% of an enhanced green fluorescent protein-Smad2 fusion is phosphorylated and accumulates in the nucleus in response to TGF-β in HaCaT cells (48) and model fitting of similar data estimates about 30,000 phospho-Smad2 molecules per cell (50). Functionally, it has been observed that only a small amount of phospho-Smad1 is required to regulate differentiation of dissociated Xenopus ectodermal cells (23). Therefore, the negative regulators of TGF-β appear to keep the number of phospho-Smad molecules to a modest percentage of the available molecules in order to tightly control the cellular responses to TGF-β.

Ligand depletion is emerging as an important mechanism for specifying cell responses. For example, TGF-α and epidermal growth factor (EGF) both signal through the EGF receptor, yet EGF more potently induces mitogenesis because TGF-α is more rapidly depleted (47). In Drosophila melanogaster embryonic development, experimental and theoretical studies have demonstrated that proper morphogen gradient formation for the BMP homolog Decapentaplegic requires ligand depletion mediated by receptor internalization followed by degradation through the endolysosomal pathway (15, 25, 37) and through binding to cell surface proteoglycans (5, 26). In addition, an inhibitory role for the Drosophila BMP homolog Glass bottom boat (Gbb) type I receptor Saxophone (Sax) was recently identified, presumably because it helps to deplete Gbb and regulate its spatial gradient (4). Similarly, proper formation of Wnt homolog Wingless gradients also require endocytosis and lysosomal degradation (12). Finally, ligand depletion enhanced by feedback can contribute to the robustness of morphogen gradients (14). In general, in the context of cultured cells, ligand depletion regulates the temporal properties of the input signal, whereas in the context of developing embryos, ligand depletion regulates both the spatial and temporal aspects of the input signal.

Some mathematical models that incorporate binding of BMP to nonsignaling cell surface proteoglycans assume that turnover of the proteoglycan molecules contributes to additional ligand internalization and degradation (26). We did not observe such behavior in our experiments, although this may reflect a cell-type-dependent feature because heparan sulfate proteoglycans are important for BMP signaling in Drosophila wing discs but not in dorsal patterning (28). Determining the molecules responsible for reversibly binding TGF-β to the cell surface would be helpful because such knowledge might allow one to predict whether a specific cell type could deplete TGF-β by this mechanism. Likely candidates include decorin, biglycan, and beta-glycan (the TβRIII) (31) in addition to as-yet-unidentified TGF-β-binding cell surface proteins (8, 44).

Our results also have implications for the study of TGF-β in physiology and disease. First, the relevance of the notion of “physiological concentration,” which is usually expressed on a per-volume basis, may need to be reconsidered if the result that TGF-β molecules per cell predicts phospho-Smad2 levels can be generalized to the in vivo setting. Similarly, assuming the result can be extended to BMP signaling in Drosophila, it implies that the potency of a given concentration (by volume) of BMP in the perivitelline space of the embryo would depend on the number of cells that are exposed to BMP. This could be relevant in accounting for any differences that might exist in BMP signaling in dorsal versus imaginal disc patterning. In addition, our finding that the TβRII is responsible for actively depleting TGF-β and that tumor cell lines with mutant TβRII exhibit impaired TGF-β depletion may have implications for the role of TGF-β in cancer (Fig. (Fig.8C).8C). Tumor cells are known to overproduce TGF-β (43). Several tumor cell lines have deletion mutations for the TβRII (27), which we showed would be unable to deplete TGF-β. The inability to deplete TGF-β would therefore contribute to the accumulation of TGF-β in the tumor microenvironment and systemically throughout the organism, a scenario that correlates with poor prognosis (43). Therefore, maximizing TGF-β overproduction, in addition to the loss of signaling responses, may underlie the selection for TβRII mutations in cancer.


This work was supported by a National Institutes of Health research grant (GM083172) to X.L. M.L.B. was supported by the Boettcher Foundation and the Undergraduate Research Opportunities Program through the University of Colorado at Boulder. D.C.C. was supported by a Doctoral Research Award from the Canadian Institutes for Health Research.


Published ahead of print on 17 February 2009.


1. Abe, M., J. G. Harpel, C. N. Metz, I. Nunes, D. J. Loskutoff, and D. B. Rifkin. 1994. An assay for transforming growth factor-beta using cells transfected with a plasminogen activator inhibitor-1 promoter-luciferase construct. Anal. Biochem. 216276-284. [PubMed]
2. Afrakhte, M., A. Moren, S. Jossan, S. Itoh, K. Sampath, B. Westermark, C. H. Heldin, N. E. Heldin, and P. ten Dijke. 1998. Induction of inhibitory Smad6 and Smad7 mRNA by TGF-β family members. Biochem. Biophys. Res. Commun. 249505-511. [PubMed]
3. Ashe, H. L., and J. Briscoe. 2006. The interpretation of morphogen gradients. Development 133385-394. [PubMed]
4. Bangi, E., and K. Wharton. 2006. Dual function of the Drosophila Alk1/Alk2 ortholog Saxophone shapes the Bmp activity gradient in the wing imaginal disc. Development 1333295-3303. [PubMed]
5. Belenkaya, T. Y., C. Han, D. Yan, R. J. Opoka, M. Khodoun, H. Liu, and X. Lin. 2004. Drosophila Dpp morphogen movement is independent of dynamin-mediated endocytosis but regulated by the glypican members of heparan sulfate proteoglycans. Cell 119231-244. [PubMed]
6. Bourillot, P. Y., N. Garrett, and J. B. Gurdon. 2002. A changing morphogen gradient is interpreted by continuous transduction flow. Development 1292167-2180. [PubMed]
7. Brattain, M. G., W. D. Fine, F. M. Khaled, J. Thompson, and D. E. Brattain. 1981. Heterogeneity of malignant cells from a human colonic carcinoma. Cancer Res. 411751-1756. [PubMed]
8. Butzow, R., D. Fukushima, D. R. Twardzik, and E. Ruoslahti. 1993. A 60-kD protein mediates the binding of transforming growth factor-beta to cell surface and extracellular matrix proteoglycans. J. Cell Biol. 122721-727. [PMC free article] [PubMed]
9. Clarke, D. C., M. B. Betterton, and X. Liu. 2006. Systems theory of Smad signalling. Systems Biol. (Stevenage) 153412-424. [PubMed]
10. Clarke, D. C., and X. Liu. 2008. Decoding the quantitative nature of TGF-β signaling. Trends Cell Biol. 18430-442. [PMC free article] [PubMed]
11. Di Guglielmo, G. M., C. Le Roy, A. F. Goodfellow, and J. L. Wrana. 2003. Distinct endocytic pathways regulate TGF-β receptor signalling and turnover. Nat. Cell Biol. 5410-421. [PubMed]
12. Dubois, L., M. Lecourtois, C. Alexandre, E. Hirst, and J. P. Vincent. 2001. Regulated endocytic routing modulates wingless signaling in Drosophila embryos. Cell 105613-624. [PubMed]
13. Dyson, S., and J. B. Gurdon. 1998. The interpretation of position in a morphogen gradient as revealed by occupancy of activin receptors. Cell 93557-568. [PubMed]
14. Eldar, A., D. Rosin, B. Z. Shilo, and N. Barkai. 2003. Self-enhanced ligand degradation underlies robustness of morphogen gradients. Dev. Cell 5635-646. [PubMed]
15. Entchev, E. V., A. Schwabedissen, and M. Gonzalez-Gaitan. 2000. Gradient formation of the TGF-β homolog Dpp. Cell 103981-991. [PubMed]
16. Gal, A., T. Sjoblom, L. Fedorova, S. Imreh, H. Beug, and A. Moustakas. 2008. Sustained TGF beta exposure suppresses Smad and non-Smad signalling in mammary epithelial cells, leading to EMT and inhibition of growth arrest and apoptosis. Oncogene 271218-1230. [PubMed]
17. Green, J. B., H. V. New, and J. C. Smith. 1992. Responses of embryonic Xenopus cells to activin and FGF are separated by multiple dose thresholds and correspond to distinct axes of the mesoderm. Cell 71731-739. [PubMed]
18. Hua, X., X. Liu, D. O. Ansari, and H. F. Lodish. 1998. Synergistic cooperation of TFE3 and smad proteins in TGF-β-induced transcription of the plasminogen activator inhibitor-1 gene. Genes Dev. 123084-3095. [PMC free article] [PubMed]
19. Inman, G. J., F. J. Nicolas, and C. S. Hill. 2002. Nucleocytoplasmic shuttling of Smads 2, 3, and 4 permits sensing of TGF-β receptor activity. Mol. Cell 10283-294. [PubMed]
20. Itoh, F., N. Divecha, L. Brocks, L. Oomen, H. Janssen, J. Calafat, S. Itoh, and P. Dijke. 2002. The FYVE domain in Smad anchor for receptor activation (SARA) is sufficient for localization of SARA in early endosomes and regulates TGF-β/Smad signalling. Genes Cells 7321-331. [PubMed]
21. Itoh, S., and P. Ten Dijke. 2007. Negative regulation of TGF-β receptor/Smad signal transduction. Curr. Opin. Cell. Biol. 19176-184. [PubMed]
22. Kavsak, P., R. K. Rasmussen, C. G. Causing, S. Bonni, H. Zhu, G. H. Thomsen, and J. L. Wrana. 2000. Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGF beta receptor for degradation. Mol. Cell 61365-1375. [PubMed]
23. Kuroda, H., L. Fuentealba, A. Ikeda, B. Reversade, and E. M. De Robertis. 2005. Default neural induction: neuralization of dissociated Xenopus cells is mediated by Ras/MAPK activation. Genes Dev. 191022-1027. [PMC free article] [PubMed]
24. Laiho, M., M. B. Weis, and J. Massague. 1990. Concomitant loss of transforming growth factor (TGF)-beta receptor types I and II in TGF-β-resistant cell mutants implicates both receptor types in signal transduction. J. Biol. Chem. 26518518-18524. [PubMed]
25. Lander, A. D., Q. Nie, and F. Y. Wan. 2002. Do morphogen gradients arise by diffusion? Dev. Cell 2785-796. [PubMed]
26. Lander, A. D., Q. Nie, and F. Y. Wan. 2007. Membrane-associated non-receptors and morphogen gradients. Bull. Math. Biol. 6933-54. [PMC free article] [PubMed]
27. Levy, L., and C. S. Hill. 2006. Alterations in components of the TGF-β superfamily signaling pathways in human cancer. Cytokine Growth Factor Rev. 1741-58. [PubMed]
28. Lin, X. 2004. Functions of heparan sulfate proteoglycans in cell signaling during development. Development 1316009-6021. [PubMed]
29. Lin, X., X. Duan, Y. Y. Liang, Y. Su, K. H. Wrighton, J. Long, M. Hu, C. M. Davis, J. Wang, F. C. Brunicardi, Y. Shi, Y. G. Chen, A. Meng, and X. H. Feng. 2006. PPM1A functions as a Smad phosphatase to terminate TGFβ signaling. Cell 125915-928. [PubMed]
30. Macdonald, M., Y. Wan, W. Wang, E. Roberts, T. H. Cheung, R. Erickson, M. T. Knuesel, and X. Liu. 2004. Control of cell cycle-dependent degradation of c-Ski proto-oncoprotein by Cdc34. Oncogene 235643-5653. [PubMed]
31. Macri, L., D. Silverstein, and R. A. Clark. 2007. Growth factor binding to the pericellular matrix and its importance in tissue engineering. Adv. Drug Deliv. Rev. 591366-1381. [PubMed]
32. Markowitz, S., J. Wang, L. Myeroff, R. Parsons, L. Sun, J. Lutterbaugh, R. S. Fan, E. Zborowska, K. W. Kinzler, B. Vogelstein, et al. 1995. Inactivation of the type II TGF-β receptor in colon cancer cells with microsatellite instability. Science 2681336-1338. [PubMed]
33. Massague, J., and B. Kelly. 1986. Internalization of transforming growth factor-beta and its receptor in BALB/c 3T3 fibroblasts. J. Cell. Physiol. 128216-222. [PubMed]
34. Massague, J., J. Seoane, and D. Wotton. 2005. Smad transcription factors. Genes Dev. 192783-2810. [PubMed]
35. Melke, P., H. Jonsson, E. Pardali, P. ten Dijke, and C. Peterson. 2006. A rate equation approach to elucidate the kinetics and robustness of the TGF-β pathway. Biophys. J. 914368-4380. [PMC free article] [PubMed]
36. Mitchell, H., A. Choudhury, R. E. Pagano, and E. B. Leof. 2004. Ligand-dependent and -independent transforming growth factor-beta receptor recycling regulated by clathrin-mediated endocytosis and Rab11. Mol. Biol. Cell 154166-4178. [PMC free article] [PubMed]
37. Mizutani, C. M., Q. Nie, F. Y. Wan, Y. T. Zhang, P. Vilmos, R. Sousa-Neves, E. Bier, J. L. Marsh, and A. D. Lander. 2005. Formation of the BMP activity gradient in the Drosophila embryo. Dev. Cell 8915-924. [PMC free article] [PubMed]
38. Muir, T. W. 2003. Semisynthesis of proteins by expressed protein ligation. Annu. Rev. Biochem. 72249-289. [PubMed]
39. Nakao, A., M. Afrakhte, A. Moren, T. Nakayama, J. L. Christian, R. Heuchel, S. Itoh, M. Kawabata, N. E. Heldin, C. H. Heldin, and P. ten Dijke. 1997. Identification of Smad7, a TGF-β-inducible antagonist of TGF-β signalling. Nature 389631-635. [PubMed]
40. Nicolas, F. J., K. De Bosscher, B. Schmierer, and C. S. Hill. 2004. Analysis of Smad nucleocytoplasmic shuttling in living cells. J. Cell Sci. 1174113-4125. [PubMed]
41. Nicolas, F. J., and C. S. Hill. 2003. Attenuation of the TGF-β-Smad signaling pathway in pancreatic tumor cells confers resistance to TGF-β-induced growth arrest. Oncogene 223698-3711. [PubMed]
42. Panopoulou, E., D. J. Gillooly, J. L. Wrana, M. Zerial, H. Stenmark, C. Murphy, and T. Fotsis. 2002. Early endosomal regulation of Smad-dependent signaling in endothelial cells. J. Biol. Chem. 27718046-18052. [PubMed]
43. Pardali, K., and A. Moustakas. 2007. Actions of TGF-β as tumor suppressor and pro-metastatic factor in human cancer. Biochim. Biophys. Acta 177521-62. [PubMed]
44. Piek, E., P. Franzen, C. H. Heldin, and P. ten Dijke. 1997. Characterization of a 60-kDa cell surface-associated transforming growth factor-beta binding protein that can interfere with transforming growth factor-beta receptor binding. J. Cell. Physiol. 173447-459. [PubMed]
45. Pierreux, C. E., F. J. Nicolas, and C. S. Hill. 2000. Transforming growth factor beta-independent shuttling of Smad4 between the cytoplasm and nucleus. Mol. Cell. Biol. 209041-9054. [PMC free article] [PubMed]
46. Pitre, A., Y. Pan, S. Pruett, and O. Skalli. 2007. On the use of ratio standard curves to accurately quantitate relative changes in protein levels by Western blot. Anal. Biochem. 361305-307. [PMC free article] [PubMed]
47. Reddy, C. C., A. Wells, and D. A. Lauffenburger. 1996. Receptor-mediated effects on ligand availability influence relative mitogenic potencies of epidermal growth factor and transforming growth factor alpha. J. Cell Physiol. 166512-522. [PubMed]
48. Schmierer, B., and C. S. Hill. 2005. Kinetic analysis of Smad nucleocytoplasmic shuttling reveals a mechanism for transforming growth factor beta-dependent nuclear accumulation of Smads. Mol. Cell. Biol. 259845-9858. [PMC free article] [PubMed]
49. Schmierer, B., and C. S. Hill. 2007. TGFβ-SMAD signal transduction: molecular specificity and functional flexibility. Nat. Rev. Mol. Cell Biol. 8970-982. [PubMed]
50. Schmierer, B., A. L. Tournier, P. A. Bates, and C. S. Hill. 2008. Mathematical modeling identifies Smad nucleocytoplasmic shuttling as a dynamic signal-interpreting system. Proc. Natl. Acad. Sci. USA 1056608-6613. [PMC free article] [PubMed]
51. Shi, W., C. Sun, B. He, W. Xiong, X. Shi, D. Yao, and X. Cao. 2004. GADD34-PP1c recruited by Smad7 dephosphorylates TGF-β type I receptor. J. Cell Biol. 164291-300. [PMC free article] [PubMed]
52. Shi, Y., and J. Massague. 2003. Mechanisms of TGF-β signaling from cell membrane to the nucleus. Cell 113685-700. [PubMed]
53. Shimizu, K., and J. B. Gurdon. 1999. A quantitative analysis of signal transduction from activin receptor to nucleus and its relevance to morphogen gradient interpretation. Proc. Natl. Acad. Sci. USA 966791-6796. [PMC free article] [PubMed]
54. Simeoni, I., and J. B. Gurdon. 2007. Interpretation of BMP signaling in early Xenopus development. Dev. Biol. 30882-92. [PubMed]
55. ten Dijke, P., and C. S. Hill. 2004. New insights into TGF-β-Smad signalling. Trends Biochem. Sci. 29265-273. [PubMed]
56. Ulloa, L., J. Doody, and J. Massague. 1999. Inhibition of transforming growth factor-beta/SMAD signalling by the interferon-gamma/STAT pathway. Nature 397710-713. [PubMed]
57. Vilar, J. M., R. Jansen, and C. Sander. 2006. Signal processing in the TGF-β superfamily ligand-receptor network. PLoS Comput. Biol. 2e3. [PMC free article] [PubMed]
58. Wakefield, L. M., D. M. Smith, T. Masui, C. C. Harris, and M. B. Sporn. 1987. Distribution and modulation of the cellular receptor for transforming growth factor-beta. J. Cell Biol. 105965-975. [PMC free article] [PubMed]
59. Wang, J., L. Sun, L. Myeroff, X. Wang, L. E. Gentry, J. Yang, J. Liang, E. Zborowska, S. Markowitz, J. K. Willson, et al. 1995. Demonstration that mutation of the type II transforming growth factor beta receptor inactivates its tumor suppressor activity in replication error-positive colon carcinoma cells. J. Biol. Chem. 27022044-22049. [PubMed]
60. Wu, J. W., M. Hu, J. Chai, J. Seoane, M. Huse, C. Li, D. J. Rigotti, S. Kyin, T. W. Muir, R. Fairman, J. Massague, and Y. Shi. 2001. Crystal structure of a phosphorylated Smad2. Recognition of phosphoserine by the MH2 domain and insights on Smad function in TGF-β signaling. Mol. Cell 81277-1289. [PubMed]
61. Zwaagstra, J. C., Z. Kassam, and M. D. O'Connor-Mccourt. 1999. Down-regulation of transforming growth factor-beta receptors: cooperativity between the types I, II, and III receptors and modulation at the cell surface. Exp. Cell. Res. 252352-362. [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Compound
    PubChem chemical compound records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records. Multiple substance records may contribute to the PubChem compound record.
  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...