Dragon is localized in various tubules in (A) the whole mouse kidney and (B through E) is localized in glomeruli and proximal tubules in the cortex. Adult mouse kidney sections were co-immunostained with (A, B) rabbit anti-Dragon antibody and (C, D) chicken anti-V-ATPase E1 subunit antibody. The Dragon and V-ATPase were visualized using Cy3-donkey anti-rabbit and FITC-donkey anti-chicken secondary antibodies, respectively. Photographs were taken at (A, 4×) low and (B through E, 40×) high magnifications. (D) The staining of Dragon in the cortex was completely blocked in the presence of its competing immunizing peptide (Dra peptide, 200 ng/μl), whereas (E) V-ATPase staining was not affected by the Dragon peptide. Dragon staining was seen in glomeruli, Bowman's capsules, and in epithelial cells of proximal tubules, which show moderate apical V-ATPase staining. (B) The strongest Dragon staining was found in cortical tubules that contain cells with abundant V-ATPase staining, suggesting that collecting ducts express the Dragon protein. G, glomerulus; PT, proximal tubules; CT, collecting ducts. Scale bars = 80 μm for (A) low magnification and 20 μm for (B through E) high magnification.

Dragon is localized in collecting ducts, distal convoluted tubules, connecting segments, and thick ascending limbs in the (A through D) cortex and (E, F) outer stripe of the mouse kidney. Mouse kidney sections were co-immunostained with (A, C and E) rabbit anti-Dragon antibody and (B) goat anti-AQP2, (D) goat anti-calbindin 28, or (F) goat anti-Tamm-Horsfall glycoprotein. Sections were then stained with Cy3-donkey anti-rabbit (red) and FITC-donkey anti-goat (green) secondary antibodies to visualize Dragon, AQP2, calbindin 28, or THP. The merged images for Dragon and AQP2, calbindin 28, or THP are shown in the respective insets. Specific staining for Dragon was seen in (A, B) principal cells in collecting ducts, (C, D) epithelial cells in distal convoluted tubules and connecting segments, and (E, F) thick ascending limbs. DCT, distal convoluted tubules; CS, connecting segments; anti-Tamm-Horsfall glycoprotein (α-THP); THP, Tamm-Horsfall glycoprotein. Scale bar = 20 μm.

Dragon is localized in collecting ducts and thick ascending limbs in the (A through C) outer and (D, E) inner stripes and (F, G) collecting ducts and thin loops of Henle in the papilla of the mouse kidney. Mouse kidney sections were co-immunostained with (A, D, E) rabbit anti-Dragon antibody and (B, E, F) chicken anti-V-ATPase E1 antibody. The Dragon and V-ATPase were visualized using Cy3-donkey anti-rabbit (red) and FITC-donkey anti-chicken secondary antibodies (green), respectively. The merged images for Dragon and V-ATPase are shown in panel C for the outer stripe or in the insets in panels E and F for the inner stripe and papilla. Strong staining for Dragon was seen in intercalated cells of collecting ducts (panel C and the inset in panel E), but adjacent principal cells were also positive. (A, B) Thick ascending limbs (TAL) of the outer stripe were also strongly staining for Dragon. Staining for Dragon was seen in collecting ducts and thin limbs of Henle's loop (inset, panel G) in the papilla. Scale bars = 30 μm for the outer stripe (A through C) and 20 μm for the inner stripe and papilla (D through G).

Dragon increases BMP but not TGF-β signaling in kidney tubule cells. (A) Dragon increases BMP signaling. mIMCD3 cells were transfected with BRE-Luc construct in the absence (bars 1 and 2), or presence (bar 3) of Dragon cDNA. Cells were then starved and incubated for 16 hours in the absence (bars 1 and 3) or presence of 40 ng/ml BMP4 (bar 2), and luciferase activity was measured the next day from cell extracts. ***P < 0.001 versus the control. (B) Dragon does not affect TGF-β signaling. mIMCD3 cells were transfected with the (CAGA)₁₂MLP-Luc construct in the absence (bars 1 and 2) or presence (bar 3) of Dragon cDNA. Cells were incubated for 16 hours in the absence (bars 1 and 3) or presence of 2 ng/ml TGF-β1 (bar 2), and luciferase activity was measured the next day. ***P < 0.001 versus the control. (C) Dragon increases BMP signaling in a dose dependent manner. mIMCD3 cells were transfected with BRE-Luc construct with increasing amounts of Dragon cDNA (0, 2, 20, 200, and 500 ng). Relative luciferase activity was measured from cell extracts. *P < 0.05 and ***P < 0.001 versus the control. (D) BMP4 and Dragon signals are additive. mIMCD3 cells were transfected with the BRE-Luc construct in the absence or presence of Dragon cDNA. Cells were then incubated without (open bars) or with BMP4 (filled bars) and relative luciferase activity was measured from cells extracts. ***P < 0.001, bar 3 versus bar 1; ###P < 0.05, bar 4 versus bar 2.

BMP4 is the endogenous ligand for Dragon in mIMCD3 cells. (A) Expression of BMP ligands in mIMCD3 cells. Total RNA from IMCD3 cells was extracted for RT-PCR to determine the expression of BMP2 and BMP4 through BMP7. Total RNA from mouse kidney was used in PCR analyses as positive controls. BMP4 and BMP6 mRNAs were readily detected, BMP2 and BMP7 mRNAs were weak, and BMP5 mRNA was undetectable in mIMCD3 cells. (B) Effect of siRNA targeting of BMP2, BMP4, BMP6, and BMP7 on Dragon-mediated BMP signaling. mIMCD3 cells were transfected with BRE luciferase reporter alone or with increasing amounts of Dragon cDNA and in combination with control, BMP2 (siBMP2), BMP4 (siBMP4), BMP6 (siBMP6), or BMP7 (siBMP7) siRNA (60 nM) for 46 hours before measurement of luciferase activity. Inhibition of BMP4 expression significantly reduced Dragon-mediated BMP signaling, whereas inhibition of BMP2, BMP6, or BMP7 had little or no effect. *P < 0.05, siBMP4 versus control at 32 ng/ml Dragon cDNA. The exact P value was shown for siBMP4 versus control at 8 ng/ml Dragon cDNA.

Dragon enhances utilization of ActRIIA by BMP4. (A) Effect of siRNA targeting of BMPRII, ActRIIA, and ActRIIB on BMP4 signaling. mIMCD3 cells were transfected with the BRE luciferase reporter in combination with control siRNA or siRNA specific for BMPRII (siBMPRII), ActRIIA (siActRIIA), or ActRIIB (siActRIIB) (60 nM). Transfected cells were then incubated in the absence or presence of 20 ng/ml BMP4. Luciferase activity was measured in cell extracts. Only inhibition of BMPRII expression reduced BMP4 signaling. (B) Effect of siRNA targeting of BMPRII, ActRIIA, and ActRIIB on Dragon-mediated BMP signaling. mIMCD3 cells were transfected with BRE luciferase reporter in combination with control siRNA or siRNA specific for BMPRII, ActRIIA, or ActRIIB (60 nM) and in the absence or presence of Dragon cDNA for 46 hours before measurement of luciferase activity. In the presence of Dragon, inhibition of BMPRII and ActRIIA reduced endogenous BMP4 signaling. *P < 0.05 versus controls with treatments.

Expression of mRNAs for BMP2, BMP4, Dragon, ActRIIA, BMPRII, and tight junction proteins MUPP-1 and JAM-A in isolated proximal tubule cells and intercalated cells. RT-PCR analyses were performed for (A) isolated proximal tubule cells and (B) intercalated cells. cDNA from the adult kidney was used as positive control and replacement of the template with water was used as negative control.

BMP4 and Dragon increase TER in mIMCD3 cells. (A) Exogenous BMP4 increased TER. mIMCD3 cells were cultured in Transwells and different concentrations of BMP4 (0, 0.5, 5, and 50 ng/ml) were added. TER was measured daily. *P < 0.05, **P < 0.01, ***P < 0.001, 50 ng/ml BMP4 versus controls; #P < 0.05, 5 ng/ml BMP4 versus controls. (B) LDN-193189 inhibition of BMP4 signaling reduced its stimulatory effect on TER. mIMCD3 cells were cultured in Tranwells and were incubated with BMP4 (40 ng/ml) in the presence or absence of LDN-193189 (40 nM). TER was measure daily. *P < 0.05, ***P < 0.001, BMP4 versus BMP4 + LDN-193189. (C) Transfection with Dragon cDNA increased TER. mIMCD3 cells cultured in Transwells were transfected with Dragon cDNA and TER was measured daily. **P < 0.01 versus controls. (D) Stimulatory effects of BMP4 and Dragon on TER are additive. mIMCD3 cells cultured in Transwells were transfected with (bars 2 and 4) or without (bars 1 and 3) Dragon cDNA and were incubated with (bars 3 and 4) or without (bars 1 and 2) 10 ng/ml BMP4. TER was measured on day 3 after transfection. **P < 0.01.

BMP4 and Dragon increase expression of MUPP-1 and JAM-A mRNAs. mIMCD3 cells cultured in Transwells were transfected with Dragon cDNA for 46 hours or treated with BMP4 for 18 hours. MUPP-1 or JAM-A relative to RPL19 mRNA levels were quantified by real-time RT-PCR. *P < 0.05.