Deleting the TGF-β receptor attenuates acute proximal tubule injury

Abstract

TGF-β is a profibrotic growth factor in CKD, but its role in modulating the kidney's response to AKI is not well understood. The proximal tubule epithelial cell, which is the main cellular target of AKI, expresses high levels of both TGF-β and its receptors. To determine how TGF-β signaling in this tubular segment affects the response to AKI, we selectively deleted the TGF-β type II receptor in the proximal tubules of mice. This deletion attenuated renal impairment and reduced tubular apoptosis in mercuric chloride-induced injury. In vitro, deficiency of the TGF-β type II receptor protected proximal tubule epithelial cells from hydrogen peroxide-induced apoptosis, which was mediated in part by Smad-dependent signaling. Taken together, these results suggest that TGF-β signaling in the proximal tubule has a detrimental effect on the response to AKI as a result of its proapoptotic effects.

Figure 1.

HgCl₂ stimulated TGF-β signaling in renal cortices. (A) Renal cortices were dissected from wild-type FVB mice that were injected with HgCl₂ 3 days earlier or were untreated. The tissue lysates were immunoblotted for TβRII expression. (B) Tissue lysates of renal cortices isolated 1 day after HgCl₂ were immunoblotted for pSmad2 and total Smad2. (C) Bands from six mice (three treated, three untreated) were quantified by densitometry and reported as means ± SEMs. *P<0.05. GADPH, glyceraldehyde 3-phosphate dehydrogenase.

Figure 2.

Deletion of TβRII in the γGT-Cre;Tgfbr2^flox/flox mice. The mT/mG reporter mouse was crossed with our γGT-Cre;Tgfbr2^flox/flox mice, demonstrating membrane-bound red fluorescence (mT) in all cells except those where Cre is active and green fluorescence (mG) replaces mT (A and B). (C) The red (580–610 nm) and green (510–540 nm) wavelengths are merged, showing that Cre is primarily expressed in tubules in the cortex and corticomedullary renal tissue. (D) Cortical lysates of adult Tgfbr2^flox/flox and γGT-Cre;Tgfbr2^flox/flox mice were immunoblotted for TβRII expression with focal adhesion kinase (FAK) used as loading control.

Figure 3.

γGT-Cre;Tgfbr2^flox/flox mice were protected from HgCl₂-induced injury. (A) Plasma BUN levels (mg/dl) were measured at the time of injury and daily thereafter until euthanasia at day 7. Means are shown ± SEMs from 14 Tgfbr2^flox/flox and 12 γGT-Cre;Tgfbr2^flox/flox mice. *P<0.05. (B) Plasma creatinine 3 days after HgCl₂ injection was measured by HPLC on six Tgfbr2^flox/flox and five γGT-Cre;Tgfbr2^flox/flox mice. Means are shown ± SEMs. *P<0.05. (C–J) Tissue from mice at 1, 3, and 7 days after treatment with HgCl₂ was stained with hematoxylin and eosin. There was minimal epithelial injury 1 day after HgCl₂, but at 3 days, there was significantly more cast formation (black arrow in D) and tubular necrosis in the Tgfbr2^flox/flox than in the γGT-Cre;Tgfbr2^flox/flox mice. Both genotypes had resolving epithelial injury at 7 days, but fewer residual tubular casts were present in the γGT-Cre;Tgfbr2^flox/flox mice. (K) Injury was scored at 3 days, as discussed in the Concise Methods section. The values represent the means of six Tgfbr2^flox/flox and seven γGT-Cre;Tgfbr2^flox/flox mice. Means are shown ± SEMs. *P<0.05.

Figure 4.

γGT-Cre;Tgfbr2^flox/flox mice had less tubular apoptosis after HgCl₂. (A and B) TUNEL staining was performed at 18 hours after injury with apoptotic nuclei staining brown (arrows). (C) TUNEL-positive tubular cells from 10 hpf per mouse were counted in a blinded fashion and reported as means using five mice per genotype ± SEMs. *P<0.05. (D) Cortical kidney lysates 18 hours after HgCl₂ injections were immunoblotted for cleaved caspase 3 expression using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a loading control. Three representative mice per genotype are shown. (E) Bands were quantified by densitometry expressed as cleaved caspase 3:GAPDH means from five Tgfbr2^flox/flox and six γGT-Cre;Tgfbr2^flox/flox mice ± SEMs. *P<0.05. (F–H) Ki-67 staining was performed on five mice per genotype at 0, 3, 5, and 7 days after HgCl₂ treatment. (F and G) Representative staining at 3 days is shown. (H) Ki-67+ tubular cells were counted in 10 hpf per mouse at the different time points and expressed as means ± SEMs. (I and J) Cortical tissue lysates from three wild-type and four γGT-Cre;Tgfbr2^flox/flox mice at 3 days after HgCl₂ injection were immunoblotted for F4/80 and CD3 to detect macrophages and T cells, respectively. Representative staining of F4/80 (K and L) and CD3 (M and N) at 3 days is shown with positively staining cells in brown (arrows). F4/80 staining was performed on four mice per genotype at days 0, 3, and 7, and F4/80+ cells were counted in 10 hpf per mouse and shown as means ± SEMs (O).

Figure 5.

Smad activity is suppressed in γGT-Cre;Tgfbr2^flox/flox mice. (A–F) Cortical tissue lysates from wild-type and γGT-Cre;Tgfbr2^flox/flox mice at 0 and 1 day after HgCl₂ were immunoblotted for phosphorylated Smad2 (pSmad2), total Smad2, phosphorylated Smad3 (pSmad3), total Smad3, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) for loading. (C, F) Densitometry on phosphorylated Smads2/3 was normalized to GAPDH and reported as means ± SEMs. *P<0.05. (G) Localization of pSmad2/3 was determined by immunofluorescence on paraffin-embedded tissue at 0 and 1 day after injury.

Figure 6.

PTCs were generated and TβRII deleted in vitro. (A and B) Cell lysates of PTCs, interstitial cells, and collecting duct (CD) cells were immunoblotted for expression of zona occludens-1 (ZO-1), E.cadherin, α-smooth muscle actin (α-SMA), claudin 2, and γ−glutamyltransferase (γGT) with focal adhesion kinase (FAK), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and α tubulin as loading controls. Vertical white lines separate samples rearranged from the same immunoblot. (C) Deletion of TβRII was confirmed by immunoblots of TβRII^flox/flox and TβRII^−/− PTCs. (D) Cell lysates of TβRII^flox/flox and TβRII^−/− PTC treated with TGF-β1 (2.5 ng/ml) for various time points were immunoblotted for phosphorylated and total levels of Smad2.

Figure 7.

TβRII^−/− PTC are resistant to apoptosis. (A) TβRII^flox/flox and TβRII^−/− PTCs were treated with various concentrations of H₂O₂ for 12 hours, and then immunoblots for cleaved caspase 3 were performed with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a loading control. (B) PTCs were treated with H₂O₂ (1mM) or TGF-β1 (2.5 ng/ml) for 12 hours and immunoblotted for cleaved caspase 3 and GAPDH. (C) TβRII^flox/flox and TβRII^−/− PTCs were plated on PolyHEMA-coated plates, as described in the Concise Methods section, for 18 hours to induce anoikis and then immunoblotted for cleaved caspase 3 and GAPDH. (D) The Roche Cell Death ELISA was used to quantify apoptosis in PTCs both treated with H₂O₂ and plated on PolyHEMA. Experiments were repeated three times, and ratios of cell death to protein concentration were normalized to 1 for TβRII^−/− PTCs. The values for untreated cells (negative controls) are expressed relative to TβRII^−/− H₂O₂-treated PTCs as means ± SEMs.

Figure 8.

Increased apoptosis in TβRII^flox/flox PTCs is partly Smad dependent. (A) Smad2/3 phosphorylation was measured in PTCs at basal conditions and after 30 minutes of H₂O₂ (1 mM). Vertical white lines separate samples rearranged from the same blot. (B and C) PTCs were transfected with 25 nM Smad2 or scramble siRNA as discussed in the Concise Methods, and knockdown of protein expression was confirmed 4 days later by immunoblots. (C) Smad2 protein expression was quantified by densitometry, normalized to loading controls, and expressed as a percentage of scramble siRNA-treated cells based on the average from three different experiments ± SEM. (D) Smad3 and Smad1 protein expression was measured after transfection with Smad2 (Ambion) with our without Smad3 (Dharmacon) siRNA. (E) Smad2 siRNA reduced expression of Smad3 as quantified by densitometry on three experiments, normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and expressed as a percentage of scramble siRNA ± SEM. (F and G) Lysates of TβRII^flox/flox and TβRII^−/− PTCs transfected with Smad or scramble siRNA and with or without H₂O₂ (1 mM) for 12 hours were immunoblotted for cleaved caspase 3 (Caspase 3) and GAPDH as a loading control. (G) Densitometry was performed on three different experiments with the average OD readings shown for H₂O₂-treated PTCs ± SEMs. *P<0.01 for apoptosis of scramble- versus Smad-siRNA–treated TβRII^flox/flox PTCs. **P<0.01 for the difference in apoptosis between TβRII^flox/flox and TβRII^−/− PTCs after H₂O₂. (H and I) PTCs with or without 2 mM H₂O₂ were immunoblotted for total and phosphorylated p38 and for ERK. (J) TβRII^flox/flox and TβRII^−/− PTCs treated with no inhibitors, 10 μM p38 inhibitor (SB203582), or 10 μM ERK inhibitor (UO126) were incubated with H₂O₂ (2 mM) for 12 hours and immunoblotted for cleaved caspase 3 (Caspase 3). The vertical white lines separate samples rearranged in order from the same immunoblot.