SARA localizes to EEA1-containing endosomes. (Top) HeLa cells were stained with antibodies to endogenous SARA (left) and EEA1 (middle). (Bottom) HeLa cells were transfected with myc-SARA and stained with a mouse antibody against myc (left) and a monoclonal antibody against EEA1 (middle). The overlap between the signals is displayed in yellow in the far right panel.

TGFβ receptor internalization into to EEA1-positive endosomes is blocked by dominant-negative dynamin. (A) Cos-7 cells cotransfected with HA-tagged type I receptor (RI) and myc-tagged type II TGFβ receptor, and either wild-type (WtDyn) or K44E dynamin (left, WT; right, K44E) were incubated for 1 h at 4°C with 100 pM TGFβ and antibodies against the HA epitope, and then for 60 min at 37°C. Cells were fixed and stained with a human antiserum against EEA1 (indicated as EEA1, red) and a mouse antiserum against dynamin I (blue), and secondary antibodies to the anti-HA antibody (RI, green). The overlap between the three signals is displayed in the panel labeled Dyn + EEA1 + RI. The overlap between EEA1 + RI is shown as indicated, and the large bottom panel represents an enlarged area of this panel (EEA1+RI (wtDyn or DynK44E). (B) Cos-7 cells cotransfected with HA-tagged type I receptor and myc-tagged type II TGFβ receptor (RII), and either wild-type (WtDyn) or K44E dynamin (left, WT; right, K44E) were incubated for 1 h at 4°C with 100 pM TGFβ and antibodies against the myc epitope, and then for 60 min at 37°C. Cells were fixed and stained and panels labeled as described in Fig. 2 A for RI.

Dominant-negative dynamin inhibits Smad2 translocation. HeLa cells transfected with myc-tagged dynamin K44E were incubated with 100 pM TGFβ for 30 min. Cells were fixed and stained with rabbit antibodies against myc (left) and mouse antibodies against endogenous Smad2/3 (right). Overlap is shown in far right column.

Quantification of the effects of dominant-negative dynamin on Smad2 translocation and transferrin uptake. (A) HeLa cells transfected with myc-tagged dynamin K44E were incubated with Alexa 594–labeled transferrin and 100 pM TGFβ for 30 min. Cells were fixed and stained to detect transfected Dynamin (left) and endogenous Smad2/3 (right). Internalized transferrin is visualized in the middle panels. Top and bottom panels represent two representative fields. (B) The three images shown in A were merged. The overlap for the top row is shown on the left; the overlap for the bottom row is shown on the right. Regions within the cytoplasm and nucleus selected as exemplified by the circles in cells 1 and 6. All the cells are numbered. (C) The green fluorescence intensity was recorded for the cytoplasmic and nuclear regions and the nuclear/cytoplasmic ratio was calculated for each cell indicated by number in B. The red bars indicate the transfected cells; the black bars indicate the nontransfected cells.

Dominant-negative dynamin inhibits Smad2 signaling. (A) The nuclear/cytoplasmic ratio of Smad2/3 was measured in 24 independent fields derived from two separate experiments. Plotted are the means ± SEM, and asterisk indicates a statistically significant (P < .01) difference from the nontransfected cohort. (B) TGFβ-induced transcription of 3Tp-Lux was measured in HeLa cells stably expressing wild-type (WT) or K44A dynamin under a tetracycline repressible promoter. Cells were transfected with reporter plasmids and grown in the continued presence or absence of tetracycline (Tet) as indicated. Bars represent the fold-stimulation over basal luciferase activity induced by a 14–16 h incubation with 100 pM TGFβ. Bars are means and lines represent standard errors of the mean of three experiments performed in triplicate.

Potassium depletion inhibits TGFβ receptor endocytosis. (A) Cells expressing myc-tagged type II receptors (RII) were subjected to the potassium- depletion protocol in the absence (−KCl) or presence (+KCl) of 10 mM KCl, and incubated with 100 pM TGFβ and anti-myc antibodies for 60 min at 4°C. After incubation at 37°C for 20 min cells were fixed, permeabilized, and stained to detect RII (top row) and endogenous EEA1 (second row). The overlap is shown in the third row and a region from this panel magnified five times in the fourth row. Arrows indicate endosomes stained for both EEA1 and RII in the control cells; none were detected in potassium depleted cells.

Potassium depletion impairs Smad2 phosphorylation and transferrin endocytosis. (A) Control (+KCl) or potassium-depleted (−KCl) HeLa cells were treated with 100 pM TGFβ for the indicated time. Cell lysates were analyzed by immunoblotting with antibody against phospho-Smad2 (pSmad2), and reprobed with anti–Smad2/3 (tSmad). Fluorescent transferrin on the nitrocellulose blot was visualized using the red fluorescence option the Strom 860 phosphoimager (Tf). (B) The intensity of the pSmad and Smad2/3 bands was measured and the ratio is graphed and compared with the intensity of the transferrin band. (C) Control (+KCl) or potassium-depleted (−KCl) Mv1Lu cells were incubated with 50 μg/ml Alexa-594 labeled mouse transferrin. The arrow indicates transferrin internalized after 30 min of incubation in KCl depleted cells, which represents 10–20% of that measured in the perinuclear region of control cells.

Potassium depletion impairs TGFβ- induced Smad2/3 translocation, but not TGFβ receptor activation or the interaction between SARA and Smad2. (A) Control (+KCl) or potassium-depleted (−KCl) Mv1Lu cells were incubated for the indicated time with 100 pM TGFβ. Cells were stained for Smad2/3. (B) Cos-7 cells were transfected with myc-tagged type II TGFβ receptor (RII) and HA-tagged type I receptor (RI). Cells were labeled with ³²P-labeled inorganic phosphate as described in Materials and methods. Cells were subjected to the potassium depletion protocol in the absence (−KCl) or presence (+KCl) of 10 mM KCl and the incubated with (+) or without (−) TGFβ for 15 min. Receptors were immunoprecipitated using antibodies against the epitope. (C) Control (+KCl) or potassium-depleted (−KCl) Mv1Lu cells expressing myc-SARA were fixed and stained with a monoclonal antibody against Smad2/3 (left) and polyclonal antibody against myc (middle). The overlap between the signals is displayed in yellow (right).

Effect of GFP-SARA-FYVE overexpression on SARA Localization and TGFβ induced Smad2/3 nuclear translocation. (A) Mv1Lu cells were transfected with GFP-SARA FYVE. Cells were fixed and stained for endogenous SARA (middle). The overlap of the two signals is displayed in the right-most panels. The localization of several endosomes containing GFP-SARA-FYVE is indicated by the arrows; note the colocalization with endogenous SARA at low levels of expression (top right, yellow signal), and its absence at higher levels of overexpression (bottom right, green signal). (B) HeLa cells were transfected with GFP-SARA-FYVE and treated without (left) or with 100 pM TGFβ for 30 min (middle and right). The cells were stained with antibodies against endogenous Smad2/3 (left and middle). GFP SARA FYVE (left) was viewed using a GFP filter set. The arrow indicates the transfected cell. (C) The nuclear/cytoplasmic ratio of Smad2/3 in nontransfected cells (gray bars), cells expressing GFP-SARA-FYVE (black bars) or cells expressing GFP alone (white bars) was measured in HeLa cells treated with 0, 50, or 100 pM TGFβ. Plotted are the averages of 20 cells per condition, taken from two independent experiments. Statistically significant (P < .01) differences between nontransfected and transfected cells are indicated by the asterisk.