Smad7 Is an Activin-inducible Inhibitor of Activin-induced Growth Arrest and Apoptosis in Mouse B Cells*

Members of the transforming growth factor-β (TGF-β) family, which includes the activins, relay signals from serine/threonine kinase receptors in membrane to nucleus via intracellular Sma- and Mad-related (Smad) proteins. Inhibitory Smad proteins were found to prevent the interaction between the serine/threonine kinase receptors and pathway-restricted Smad proteins. Smad7 was identified as a TGF-β-inducible antagonist of TGF-β signaling, and it may participate in a negative feedback loop to control TGF-β signaling. Here we demonstrate that the mRNA expression of Smad7 is induced by activin A in mouse B cell hybridoma HS-72 cells, which undergo growth arrest and apoptosis upon exposure to activin A. The ectopic expression of mouse Smad7 in HS-72 cells suppressed the activin A-induced cell cycle arrest in the G1 phase by abolishing the activin A-induced expression of p21CIP1/WAF1 and hypophosphorylation of retinoblastoma protein. Furthermore, Smad7 expression suppressed activin A-induced apoptosis in HS-72 cells. Thus, our data indicate that Smad7 is an activin A-inducible antagonist of activin A-induced growth arrest and apoptosis of B lineage cells.

Activins belong to the transforming growth factor-␤ (TGF-␤) 1 family (1) and were originally isolated as factors stimulating secretion of a follicle-stimulating hormone from anterior pituitary cells (2,3). In addition to regulation of the reproductive endocrine system, activins are also implicated in regulation of erythroid differentiation (4), mesoderm induction of an embryo (5)(6)(7)(8), and negative cell growth of various cell types, including gonadal cells and adrenal cells (9 -12). Recently, activin A has been isolated from the cultured media of activated mouse macrophages as a factor that inhibits the growth of plasmacytic cells, including mouse B cell hybridoma cells and mouse and human myeloma cells (13)(14)(15). We have reported that activin A induces the cell cycle arrest in the G 1 phase and apoptosis in mouse B cell hybridoma HS-72 cells (16) and that signals for both growth arrest and apoptosis induced by activin A are mediated through a type IB activin receptor (ActR-IB) in HS-72 cells (17). However, regulation of intracellular signal transduction for activin remains to be clarified.
TGF-␤ family members elicit their multifunctional effects through heteromeric complexes of type I and type II serine/ threonine kinase receptors (18 -20). Two type I receptors, i.e. ActR-I and in particular ActR-IB (21)(22)(23)(24) and two type II receptors, i.e. ActR-II and ActR-IIB (25)(26)(27), have been implicated in transducing activin signals. Upon activin binding to type II receptor with constitutively active kinase, the type I receptor is recruited and phosphorylated and activated by type II receptor kinase (28,29).
Members of the Sma-and Mad-related (Smad) protein family are known to play pivotal roles in intracellular TGF-␤ family signaling (30). Smad1, Smad2, Smad3, and Smad5 become phosphorylated by specific activated type I serine/threonine kinase receptors and thus act in a pathway-restricted fashion. Smad4 forms hetero-oligomeric complexes with pathway-restricted Smad proteins, which translocate into the nucleus and activate transcriptional responses. Smad6 and Smad7 function as inhibitors of TGF-␤ family signaling (31)(32)(33). Smad7 is most closely related to Smad6, with 36 and 56% sequence identities in the amino-terminal domain and the carboxyl-terminal Mad homology 2 domain, respectively. They bind to type I receptors and interfere with the phosphorylation of the pathway-restricted Smad proteins (30 -33) or interfere with complex formation between pathway-restricted and common-mediator Smad proteins (34).
In this study, the levels of Smad7 mRNA were examined in HS-72 cells stimulated with activin A. In addition, HS-72 cells were transfected with mouse Smad7 expression plasmid to examine the effect of Smad7 on activin A-induced responses.
Here we report the functional role of Smad7 in the regulation of growth arrest and apoptosis induced by activin A.
Plasmid and Transfection-Construction of the Smad7 expression plasmid pcDNA3-FLAG-Smad7 was reported previously (31). HS-72 cells were transfected with plasmid by electroporation using an Electroporator II (Invitrogen, San Diego, CA) at 200 V, 1000 microfarads, and then selected by cultivation with G418 (1 mg/ml). Single-cell clones were obtained by limiting dilution.
MTT Assay-The cells (2 ϫ 10 4 cells/well in 96-well plates) were incubated with Iscove's modified Dulbecco's medium containing 5% fetal calf serum and antibiotics in the presence of various concentrations of activin A or human bone morphogenic protein (hBMP)-2 for 48 h and then examined for cell viability by a calorimetric assay with MTT (Sigma) (13). Absorbance was determined at a wavelength of 570 nm with background subtraction at 620 nm.

RESULTS AND DISCUSSION
Northern blot analysis of Smad7 revealed that Smad7 mRNA was induced in HS-72 cells stimulated with activin A (Fig. 1). The expression of Smad7 mRNA was most vigorously induced after 3 h of activin A stimulation, and decreased timedependently, indicating that Smad7 is an activin A-responsive gene. TGF-␤1 and activin A were found to induce Smad7 mRNA expression in the mink epithelial cell line Mv1lu (31). 2 Since the kinase domains of ActR-IB and type I TGF-␤ receptor (T␤R-I) are nearly identical (24), activin A and TGF-␤ may share the same or similar signals for induction of Smad7 mRNA expression.
To investigate whether Smad7 controls the growth arrest and apoptosis induced by activin A, HS-72 cells were stably transfected with mouse Smad7 expression plasmid (pcDNA3-FLAG-Smad7) or control plasmid (pcDNA3).   HS-72 cells were exposed to activin A (50 ng/ml) for various times. The total RNA (10 g) were separated and analyzed for levels of Smad7 mRNA by Northern blotting. The same blot was serially hybridized with GAPDH cDNA probe.

Smad7 Inhibits Activin-induced G 1 Arrest and Apoptosis
found to express high levels of mouse FLAG-Smad7 (48 kDa) ( Fig. 2A). Control clones (HS-72C1, HS-72C2, HS-72C3, HS-72C4) showed growth arrest in response to activin A, however, Smad7 overexpressing clones showed no growth arrest in response to activin A in a MTT assay (Fig. 2B). Furthermore, cultivation with activin A for 24 h increased the population of the cells in the G 1 phase from 33 to 81% with a reduction of those in the S phase from 61 to 11% in HS-72C4 cells. In contrast, activin A showed no effect on the cell cycle distributions of Smad7 overexpressing clones (Fig. 2C). These results indicate that an overexpression of Smad7 in HS-72 cells abolishes the cell cycle arrest in the G 1 phase caused by the activin A treatment. p21 CIP1/WAF1 is known to suppress the activity of cyclin-dependent kinase (CDK)-4 (39). We have demonstrated that activin A induces the expression of p21 CIP1/WAF1 and blocks the Rb kinase activity of CDK-4, resulting in the accumulation of hypophosphorylated Rb (pRb) (16). pRb binds to, and negatively regulates, the activities of transcriptional factors of the E2F family, whose function is important for the G 1 to S transition (40,41).
To obtain more insight into mechanism by which Smad7 suppresses the activin A-induced cell cycle arrest in the G 1 phase, we examined the effect of Smad7 on the activin A-induced expression of p21 CIP1/WAF1 by immunoblot and Northern blot analyses. Immunoblot analysis showed that the untreated control clone (HS-72C4) contained undetectable levels of p21 CIP1/WAF1 (Fig. 3A). Upon exposure to activin A (50 ng/ml), expression of p21 CIP1/WAF1 (21 kDa) was detected as early as 6 h, and its level increased time-dependently in the control clone. However, Smad7 overexpressing clones (HS-72SM1 and HS-72SM2) contained undetectable levels of p21 CIP1/WAF1 , even when cultured with activin A (50 ng/ml). Northern blot analysis showed that p21 CIP1/WAF1 mRNA was rapidly induced in the control clone (HS-72C4) after 3 h of activin A stimulation (Fig. 3B). However, p21 CIP1/WAF1 mRNA was not detected in the Smad7 overexpressing clone (HS-72SM1 and HS-72SM2), even when cultured with activin A (50 ng/ml). These results suggest that Smad7 abolishes p21 CIP1/WAF1 expression by attenuating the activin signal, which induces the p21 CIP1/WAF1 mRNA expression in HS-72 cells. We examined the effect of Smad7 on the activin A-induced hypophosphorylation of Rb by immunoblot analysis. As shown in Fig. 3C, activin A (50 ng/ml) decreased the levels of hyperphosphorylated Rb (ppRb) in the control clone (HS-72C4) at 12 and 24 h, resulting in an increase in the levels of pRb time-dependently. On the contrary, activin A showed no effect on the phosphorylation status of Rb in Smad7 expressing clones (HS-72SM1 and HS-72SM2) during the 24-h culture. Taken together, these findings suggest that the decreased level of p21 CIP1/WAF1 by overexpression of Smad7 in HS-72 cells results in continuous phosphorylation of Rb by activated CDK-4.
In addition, we examined the effect of Smad7 on activin A-induced apoptosis, which is usually seen after the cell cycle arrest in the G 1 phase. Gel electrophoresis of cellular DNA showed that fragmented DNA was detected much less in ac- The cells were cultured with activin A (50 ng/ml) for the times indicated. A, p21 CIP1/WAF1 expression was analyzed by immunoblotting. B, the total RNA (5 g) were separated and analyzed for levels of p21 CIP1/WAF1 mRNA by Northern blotting. The same blot was serially hybridized with GAPDH cDNA probe. C, Rb expression was analyzed by immunoblotting.
Smad7 Inhibits Activin-induced G 1 Arrest and Apoptosis tivin A-treated Smad7 overexpressing clones (HS-72SM1, HS-72SM2, HS-72SM3, HS-72SM4) than in HS-72 and control clones (HS-72C2, HS-72C4) (Fig. 4A). As demonstrated in Fig.  4B, cultivation of the control clone (HS-72C4) with activin A (50 ng/ml) increased the population of nuclei with reduced DNA content representing apoptotic nuclei to 47%. In contrast, Smad7 overexpressing clones (HS-72SM1 and HS-72SM2) showed no increases in the population of apoptotic nuclei, even when cultured with activin A for 36 h. These results indicate that overexpression of Smad7 suppresses not only activin A-induced growth arrest in the G 1 phase but also activin A-induced apoptosis in HS-72 cells.
We examined the effect of TGF-␤ family, such as TGF-␤, BMP-2, and inhibin, on the responses of HS-72 cells. Among these factors, hBMP-2 induced the growth arrest in HS-72 cells, and the overexpression of Smad7 in HS-72 cells inhibited the hBMP-induced growth arrest (data not shown). The precise mechanism by which Smad7 suppresses the hBMP-2-induced growth arrest is under the investigation in our laboratory. We have reported that concanamycin A induces apoptotic cell death of HS-72 cells (42). Concanamycin A is known as a strong inhibitor of the vacuolar type H ϩ -ATPases in vitro and in vivo (43). In this study, we found that concanamycin A sensitivities of Smad7 overexpressing clones (HS-72SM1, HS-72SM2) were essentially the same as that of the control clone (HS-72C4) concerning the induction of apoptosis (data not shown). These results suggest that the ability of activin A and concanamycin A to promote apoptosis could be differently regulated in HS-72 cells.
In summary, we demonstrated that Smad7 mRNA in HS-72 cells was induced in response to activin A stimulation. Overexpression of mouse Smad7 was found to suppress activin A-induced cell cycle arrest in the G 1 phase by abolishing the activin A-induced expression of p21 CIP1/WAF1 and hypophosphorylation of Rb. Furthermore, Smad7 suppressed activin A-induced apoptosis in HS-72 cells. These results indicate that Smad7 may play an important role in regulating the growth arrest and apoptosis induced by activin A stimulation.