Figure 1.
Ferristatin II induces degradation of TfR1 in vitro.
Panel A: HeLa cells were treated for 4 hours with indicated concentrations of ferristatin II in the presence of 40 nM 55Fe-Tf. Shown are means ± SEM for triplicate values. Panel B: HeLa cell lysates were collected for Western blotting to determine Tf receptor levels. Panel C: Time course studies were carried out with cells treated with 50 µM ferristatin II for up to 6 hours.
Figure 2.
Ferristatin II does not degrade TfR2 or HFE.
Panel A: Lysates from Hep3B cells stably expressing TfR2 were collected for Western blotting after 4 hours of 50 µM ferristatin II treatment. Ratios of band density for TfR1/Actin or TfR2/Actin are normalized to vehicle control (DMSO) in the absence of ferristatin II. The bar graph summarizes data from four separate experiments (n = 11 TfR1/Actin p<0.001, determined by two-tailed Student’s t test, n = 12 TfR2/Actin). Panel B: HeLa cells were transfected to express HFE-flag as described in Materials and Method. Cells were then incubated 4 h with or without 50 µM ferristatin II. Tubulin is shown as a loading control and the indicated values were normalized to control lanes in the absence of ferristatin II. The bar graph represents multiple transfections (n = 8, p<0.001 for TfR1/Actin determined by two-tailed Student’s t test).
Figure 3.
Ferristatin II induced degradation is bafilomycin and nystatin sensitive.
Panel A: HeLa cells were treated overnight with 10 nM bafilomycin A1 prior to 4 h treatment with or without 50 µM ferristatin II in the presence of 10 nM bafilomycin A1. Blot is representative of several experiments. Panel B: HeLa cells were pre-treated for 30 minutes with 25 µg/mL nystatin or left untreated. After incubation, cells were treated with 50 µM ferristatin II for 4 hours. Shown below are the density ratios for TfR1/Actin normalized to control lanes (DMSO treated) in the absence of ferristatin II. Shown is a representative blot from several similar experiments.
Figure 4.
Ferristatin induced degradation is independent of an interaction with the clathrin endocytic machinery.
Panel A: TRVb cells stably transfected to express WT TfR1, Y20C/F23A TfR1, or Δ3–28 TfR1 were treated for 4 hours with 50 µM ferristatin II. Density ratios for TfR1/Actin normalized to control lanes (DMSO treated) in the absence of ferristatin II are indicated for each lane. Nonconsecutive lanes are separated by a white space. Individual blots are separated by a black bar. All blots were probed using a sheep-TfR1 antibody raised against the ectodomain of TfR1. Panel B: Time course of WT TfR1, Y20C/F23A TfR1 and Δ3–28 TfR1 degradation after 0–4 hour incubation with 50 µM ferristatin II. Shown are mean TfR1/Actin ratios ± SEM from 3 separate experiments performed in duplicate.
Figure 5.
Tf blocks ferristatin II induced receptor degradation.
HeLa cells were treated for 4 hours with 50 µM ferristatin II with or without 1 mg/mL Tf. Shown below are the density ratios for TfR1/Actin normalized to control lanes (DMSO treated) in the absence of ferristatin II. Shown is a representative blot with similar results observed on several separate occasions.
Figure 6.
Ferristatin II induces degradation of ligand binding mutant G647A TfR1.
Panel A: TRVb cells were transfected with 1 µg WT TfR1 or G647A TfR1, plated on poly-L-lysine coated cover slips and fixed with 4% paraformaldehyde. Cells were immunoreacted with OKT9 (α-TfR1) followed by goat anti-mouse Alexa 488 and imaged using a Zeiss Observer Z1 Axioscope microscope. Panel B: TRVb cells were transfected as in (A) in 6-well plates and incubated for 48 hours (see Experimental Procedures for details). Cells were then chilled, washed and incubated with 500 nM 125I-Tf in the absence or presence of 5 µM unlabeled Tf for 2 hours. After washing and lysis, cell associated radioactivity was measured by γ-counting. Absolute deviation for duplicate values for cells with 500 nM 125I-Tf (open bars) or 500 nM 125I-Tf +5 µM Tf (closed bars) is shown. Inset: Western blot confirms equivalent expression levels for WT and G647A TfR1. Panel C: TRVb cells, transfected with WT or G647A TfR1 were treated for 4 hours with 50 µM ferristatin II with or without 1 mg/mL Tf. Shown below are the density ratios for TfR1/Actin normalized to control lanes (DMSO treated) in the absence of ferristatin II. Nonconsecutive lanes are separated by a white space. Blot is representative of several experiments (n = 6, WT and G647A TfR1).
Figure 7.
Effects of ferristatin ll in vivo.
Rats were treated with vehicle control (saline) or 0.2, 10 or 40 mg/kg ferristatin ll for 2 or 4 days as described in Experimental Procedures. Shown are mean values ± SEM (n = 4) for body weight (*P = 0.022), liver non-heme iron, Tf saturation (*P = 0.018, 0.002 and 0.006 for 0.2, 10 and 40 mg/kg, respectively) and serum iron levels (*P = 0.016, 0.002 and 0.005 for 0.2, 10 and 40 mg/kg, respectively) measured after a 6 h fasting period. P values were determined by one-way ANOVA followed by Tukey’s test as a post hoc comparison. Monoacetylbenzidine (MAB) levels in rat urine were determined by HPLC and normalized to creatinine levels (arbitrary units). Shown are means ± SEM (AU = arbitrary units; ND = not detected. Serum ALT and AST activities were determined using ALT and AST reagents (Thermo Scientific) for rats injected for 4 days with vehicle or 40 mg/kg ferristatin II. Shown are means ± SEM (n = 3–6; *P = 0.033 determined by two-tailed Student’s t test).
Figure 8.
Ferristatin II alters iron homeostasis in vivo.
Panel A: Liver lysates from rats injected with saline or 40 mg/kg ferristatin ll for 4 days were immunoblotted to determine TfR1 levels. Actin was used as a loading control. Bar graph shows normalized TfR1/actin ratio as the mean ± SEM (n = 7; *P<0.001 determined by two-tailed Student’s t test). Panel B: Liver RNA was isolated, reverse transcribed and qPCR performed as described under Experimental Procedures. Data were normalized to levels of 36B4. Shown are means ± SEM (n = 5–7; *P<0.001 determined by two-tailed Student’s t test). Panel C: To determine intestinal iron absorption, control and treated rats were fasted for 4 h and 59Fe was administered by gavage. Blood samples were drawn at indicated times and radioactivity was determined by gamma counting. Shown are means ± SEM for control (open circles; n = 4–6) and ferristatin ll (closed circles; n = 7–8; *P = 0.01 and **P = 0.05 determined by two-tailed Student’s t test).
Figure 9.
Model of iron homeostasis under normal and ferristatin II conditions.
The contributors to iron metabolism include TfR1, TfR2, Tf and HFE. These molecules work together to maintain systemic iron homeostasis by sensing high and low iron levels. See text for details.