EGFR signaling promotes TGFβ-dependent renal fibrosis

Abstract

The mechanisms by which angiotensin II (Ang II) promotes renal fibrosis remain incompletely understood. Ang II both stimulates TGFβ signaling and activates the EGF receptor (EGFR), but the relative contribution of these pathways to renal fibrogenesis is unknown. Using a murine model with EGFR-deficient proximal tubules, we demonstrate that upstream activation of EGFR-dependent ERK signaling is critical for mediating sustained TGFβ expression in renal fibrosis. Persistent activation of the Ang II receptor stimulated ROS-dependent phosphorylation of Src, leading to sustained EGFR-dependent signaling for TGFβ expression. Either genetic or pharmacologic inhibition of EGFR significantly decreased TGFβ-mediated fibrogenesis. We conclude that TGFβ-mediated tissue fibrosis relies on a persistent feed-forward mechanism of EGFR/ERK activation through an unexpected signaling pathway, highlighting EGFR as a potential therapeutic target for modulating tissue fibrogenesis.

Figure 1.

Ang II–mediated tubulointerstitial fibrosis is attenuated in mice with selective deletion of EGFR in renal proximal tubules or in response to the EGFR tyrosine kinase inhibitor, erlotinib. (A) Schematic for the generation of EGFR^ptKO mice by crossing γGT.Cre mice with EGFR^f/f mice. EGFR deletion of exon 3 and Cre expression were verified by reverse transcription PCR using kidney RNA as a template. EGFR attenuation from cortical lysates was confirmed by immunoblotting. (B) Immunohistochemistry of kidney sections stained with anti-EGFR antibody (red) indicated that EGFR was predominantly expressed in cortical tubular cells and markedly diminished in EGFR^ptKO mouse (×25 magnification). The proximal tubular marker, LTA (green) on merge with red indicates deletion of EGFR primarily in renal proximal tubular epithelium-like cells (yellow; ×200 magnification). Immunoblotting of the renal cortex lysates confirmed deletion of EGFR in the renal cortex. (C) EGFR^ptKO mice 9–10 weeks of age and their control littermates were subjected to unilateral nephrectomy followed by subcutaneous administration of saline or Ang II (1.4 mg/kg per day) through osmotic mini-pumps for 2 months. After 3 months of Ang II infusion, Masson trichrome staining to determine the extent of tubulointerstitial fibrosis indicated less tubulointerstitial fibrosis in EGFR^ptKO mice as quantitated by determining the degree of blue staining by image analysis of 25 randomly designated cortical areas for each sample (n=5–7, data were shown as mean ± SEM). *P<0.05, **P<0.001). (D) Wild-type FVB/NJ mice were also subjected to unilateral nephrectomy (UNX) followed by infusion with Ang II (1.4 mg/kg per day) or saline alone, and the Ang II–infused mice were randomly divided into erlotinib (80 mg/kg per day, 7 days/wk) and vehicle subgroups. After continuous treatment for 6 weeks, Masson trichrome staining indicated less tubulointerstitial fibrosis in erlotinib-treated mice infused with Ang II compared with treatment with Ang II alone (n=4 per group; as mean ± SEM). *P<0.05, **P<0.001. ×200 magnification.

Figure 2.

Ang II infusion induces dedifferentiation in proximal tubular epithelium-like cells, which is attenuated by selective deletion of EGFR in the proximal tubules or pharmacological inhibition of EGFR tyrosine kinase activity. (A) EGFR^ptKO mice and littermate controls were administered vehicle (saline) or Ang II for 3 months. Representative kidney cortical sections of four groups were stained with the indicated epithelium-like (E-cadherin) and mesenchymal (N-cadherin, vimentin, snail) markers (red), the proximal tubule marker, LTA (green), and Topro, indicating nuclei (blue). Ip, intermediate phenotype; Fb, fibroblast. (B) Immunoblotting of kidney cortex lysates in wild-type and EGFR^ptKO mice in response to chronic Ang II exposure. (C) Immunoblotting of kidney cortex lysates of wild-type mice with chronic Ang II exposure with or without erlotinib treatment.

Figure 3.

Ang II infusion increases TGFβ expression and activity in proximal tubules, which is attenuated by selective deletion of EGFR in the proximal tubules or pharmacological inhibition of EGFR tyrosine kinase activity. (A) Immunoreactivity in EGFR^ptKO mice and littermate controls in response to chronic Ang II administration. TGFβ immunoreactivity is red, LTA is green, and Topro is blue. (B) Immunoblotting of lysates of isolated by Percoll gradient centrifugation from wild-type mice exposed to chronic Ang II or saline control. (C) Immunoblotting of kidney cortex lysates of EGFR^ptKO mice and littermate controls after chronic Ang II administration. (D) Immunoblotting of kidney cortex lysates of wild-type mice with chronic Ang II exposure with or without erlotinib treatment. (E) In AT1R/Cl4 cells, Ang II administration increased phospho-Smad2/3 expression, which was inhibited by the TGFβ receptor kinase inhibitor, 3-(pyridin-2-yl)-4-(4-quinonyl)]-¹H-pyrazole. (F) Pretreatment with the protein synthesis inhibitor, cycloheximide, did not prevent Ang II–mediated early (10 minutes) TGFβ activation but did prevent the later (3 hours) increase in TGFβ expression and activation.

Figure 4.

Ang II–mediated ERK activation in AT1R/Cl4 cells is mediated by sustained EGFR activation. (A) ERK_1/2 phosphorylation increased in response to either Ang II or EGF, but only the Ang II–dependent activation was sustained. (B) Knockdown of EGFR expression with siRNA partially inhibited the early (10 minutes) ERK_1/2 phosphorylation and almost completely inhibited the late (3 hours) phosphorylation. (C) The HB-EGF inhibitor, CRM197, partially inhibited early ERK_1/2 phosphorylation but did not affect late activation. (D) Ang II–induced EGFR phosphorylation at both Y1173 and Y845, but only Y845 phosphorylation persisted at later time points.

Figure 5.

Ang II–mediated increases in ROS in AT1R/Cl4 cells activate Src kinase, which mediate EGFR/ERK activation and increase TGFβ expression. (A) Ang II increased ROS production, measured by fluorescence intensity of 2′,7′-dichlorodihydrofluorescin, which was inhibited by the NADPH oxidase inhibitor, apocynin. *P<0.05, **P<0.001. (B) Ang II increased Y416 phosphorylation of Src, a marker of kinase activation, which was inhibited by the SOD mimetic, Tempol. Tempol also inhibited the sustained Ang II–mediated expression of ^Y845EGFR and ERK_1/2 phosphorylation. (C) Knocking down Src expression with siRNA inhibited Ang II–mediated ^Y845EGFR expression and late (3 hours) ERK_1/2 phosphorylation, TGFβ expression, and phospho-Smad2/3 expression. (D) Knocking down EGFR with siRNA inhibited Ang II–mediated TGFβ and phospho-Smad2/3 expression. (E) The MEK inhibitor, PD98059, inhibited Ang II–mediated TGFβ and phospho-Smad2/3 expression.

Figure 6.

Inhibition of EGFR attenuates Ang II–mediated TGFβ expression in renal cortex. (A) Chronic administration of Ang II selectively increased expression of ^Y845EGFR and phospho-ERK_1/2, which were markedly attenuated in EGFR^ptKO mice. Ang II–mediated increases in TGFβ and ^pSmad2/3 expression were also inhibited in EGFR^ptKO mice. (Shown is a representative set of blots from three separate experiments with similar results; each lane represents one sample from an individual mouse out of six mice per group.) (B) Erlotinib treatment inhibited expression of ^Y845EGFR, ^pERK_1/2, TGFβ, and ^pSmad2/3. (C) Tempol or apocynin inhibited expression of ^Y416Src, ^Y845EGFR, and ^pERK_1/2. (D) Induction of streptozotocin-induced diabetes for 3 weeks increased expression of ^Y416Src, ^Y845EGFR, ^pERK_1/2, TGFβ, and ^pSmad2/3, which were attenuated in EGFR^ptKO mice.

Figure 7.

Proposed mechanism for the role of sustained ROS-mediated EGFR activation in mediation of progressive tubulointerstitial injury.