The Hippo-Salvador signaling pathway regulates renal tubulointerstitial fibrosis

Renal tubulointerstitial fibrosis (TIF) is the final pathway of various renal injuries that result in chronic kidney disease. The mammalian Hippo-Salvador signaling pathway has been implicated in the regulation of cell proliferation, cell death, tissue regeneration, and tumorigenesis. Here, we report that the Hippo-Salvador pathway plays a role in disease development in patients with TIF and in a mouse model of TIF. Mice with tubular epithelial cell (TEC)-specific deletions of Sav1 (Salvador homolog 1) exhibited aggravated renal TIF, enhanced epithelial-mesenchymal transition-like phenotypic changes, apoptosis, and proliferation after unilateral ureteral obstruction (UUO). Moreover, Sav1 depletion in TECs increased transforming growth factor (TGF)-β and activated β-catenin expression after UUO, which likely accounts for the abovementioned enhanced TEC fibrotic phenotype. In addition, TAZ (transcriptional coactivator with PDZ-binding motif), a major downstream effector of the Hippo pathway, was significantly activated in Sav1-knockout mice in vivo. An in vitro study showed that TAZ directly regulates TGF-β and TGF-β receptor II expression. Collectively, our data indicate that the Hippo-Salvador pathway plays a role in the pathogenesis of TIF and that regulating this pathway may be a therapeutic strategy for reducing TIF.

Notably, the Hippo signaling pathway plays an important role in tissue regeneration after injury. A recent study demonstrated that Yap is essential for maintaining glomerular filtration barrier integrity [34][35][36] .
In the current study, we used genetic in vivo and in vitro approaches to demonstrate the role of the Hippo signaling pathway in renal tubules in progressive TIF. We found that genetic deletion of Sav1 in TECs in vitro and in vivo substantially increased TIF severity through TGF-β and Wnt/β -catenin signaling activation.

TEC-specific Sav1 deletions enhance TIF after UUO. To understand the implications of Hippo-
Salvador signaling in TIF, we generated TEC-specific Sav1-knockout mice (Sav1 fl/fl ;Ksp-Cre), in which Cre expression was limited to TECs in the distal tubular segments of the kidney 37,38 . Sav1 fl/fl ;Ksp-Cre mice were born at the expected Mendelian frequencies. No overt renal histological or functional abnormalities were observed in these animals. We assessed the wild-type and floxed-alleles of the Sav1 gene using genomic PCR (Supplementary Figure 1a). On a whole-kidney homogenate level, the level of Sav1 protein expression was greatly reduced, which indicated successful generation of conditional knockout animals (Fig. 1a). The level of Sav1 protein expression was slightly decreased in wild-type kidneys after UUO, and Sav1 protein expression was also decreased in Sav1depleted kidneys (Fig. 1a). After UUO, the progression of TIF was substantially enhanced in the kidneys of Sav1 fl/fl ;Ksp-Cre mice compared with those of wild-type (WT) mice. Masson's trichrome staining demonstrated increased ECM deposition within the tubulointerstitium in WT mice at 7 days after UUO; this deposition was more severe in Sav1 fl/fl ;Ksp-Cre mice (Fig. 1b). Immunohistochemical staining for collagen IV confirmed the presence of exacerbated TIF in Sav1-null kidneys (Fig. 1c). These findings suggest that Sav1-deficient tubular epithelial cells are prone to developing more severe TIF after UUO. We investigated whether the Hippo-Salvador pathway regulates cell apoptosis and proliferation during progressive TIF by performing TUNEL (terminal deoxynucleotidyl transferase dUTP nick-end labeling) assays and immunostaining for PCNA (proliferating cell nuclear antigen), respectively. Some TUNEL-positive cells were detected in Sav1-null kidneys in sham-operation mice, but very few TUNEL-positive cells were observed in WT mice kidneys (Fig. 1d). After UUO, WT mice kidneys exhibited increased numbers of TUNEL-positive cells, an effect that was enhanced in Sav1 fl/fl ;Ksp-Cre mice kidneys (Fig. 1d). Similarly, immunohistochemical staining for PCNA demonstrated increased cell proliferation in Sav1-null kidneys (Fig. 1e). PCNA-positive cells were co-localized with the TEC markers NCCT (Na-Cl co-transporter) and calbindin in Sav1 fl/fl ;Ksp-Cre mice after UUO (Supplementary Figure S2a). Taken together, these results indicate that severe TIF induced by Sav1 deficiency is accompanied by TEC apoptosis and proliferation.

TEC-specific Sav1 deletions increase EMT-like phenotypic changes. To determine whether
EMT-like phenotypic changes contribute to TIF development in Sav1-null kidneys after UUO, we observed EMT marker expression. The expression of α -SMA, a marker of myofibroblasts 3 , was markedly upregulated in the obstructed kidneys of Sav1 fl/fl ;Ksp-Cre mice compared with the obstructed kidneys of WT mice ( Fig. 2c and Supplementary Figure S2b). Immunohistochemical staining for vimentin (VIM), a cytoskeleton protein and a specific marker of mesenchymal cells, demonstrated significant numbers of VIM-positive TECs in obstructed Sav1-null kidneys after UUO (Fig. 2a). VIM transcripts were substantially increased in Sav1 fl/fl ;Ksp-Cre mice compared with control mice after UUO (Supplementary Figure S2b). Additionally, the numbers of interstitial cells expressing fibroblast-specific protein-1 (FSP-1) were also increased in Sav1 fl/fl ;Ksp-Cre mice compared with control mice after UUO (Fig. 2b). The mRNA expression levels of CDH1 (E-cadherin), an epithelial cell marker 39 , were decreased in Sav1-null mice compared with controls after UUO (Supplementary Figure S2b). These data suggest that Hippo-Salvador pathway dysfunction induces tubular EMT-like phenotypic changes after UUO.
To determine whether Sav1 depletion induces TGF-β -induced EMT-like phenotypic changes in vitro, we knocked down Sav1 in HK2 cells via lentiviral delivery of small hairpin (interfering) RNA (shRNA; #1 and #2) molecules against Sav1 and then treated the cells with TGF-β 1 for 12 hours (Fig. 3a). Treatment of Sav 1-depleted HK2 cells with TGF-β 1 caused marked increases in Col1a (alpha-1 type I collagen), Col3a, α -SMA, VIM, and SNAI2 (snail family zinc finger 2) mRNA expression (Fig. 3b). By contrast, exposure to TGF-β 1 reduced CDH1 mRNA levels by ~50% in Sav11-depleted HK2 cells compared with untreated Sav1 cells. These results suggest that Sav1 deficiency results in tubular EMT-like phenotypic changes after TGF-β 1 treatment in an in vitro cell culture system.  Our results suggest that Sav1 deficiency induces TGF-β -induced EMT-like phenotypic changes. Furthermore, recent studies have shown that YAP1 and TAZ cooperate with Smad proteins to promote TGF-β signaling. For example, a study reported that YAP1/TAZ act as mechanoregulators 40 and contribute to the development of renal fibrosis via promotion of TGF-β signaling 41 . We first tested whether TAZ or YAP1, which are major targets of the Hippo pathway, plays a role in TGF-β signaling in renal TIF. We examined time courses of the changes in TAZ and YAP1 protein expression during incubation with TGF-β in vitro. TGF-β 1 treatment increased TAZ protein expression in a time-dependent manner but did not affect YAP1 protein expression (Fig. 4a). CTGF is a well-characterized YAP/TAZ target gene. TGF-β 1 induced CTGF expression, which means that YAP and TAZ were activated by TGF-β 1 (Fig. 4a). We also examined the levels of proteins involved in the Hippo pathway in Sav1-depeleted cells. As shown in Fig. 4b, TGF-β 1 treatment clearly activated SMAD3 in HK2 cells. Furthermore, TAZ protein levels increased in Sav1-depleted cells in response to TGF-β treatment (Fig. 4b). However, YAP1 protein levels were unchanged between the control and Sav1-depeleted cells, irrespective of the presence of TGF-β 1 (Fig. 4b).
To confirm that increases in TAZ protein expression affect Hippo pathway target gene expression, we examined the effects of Sav1 deletion on the expression of the following YAP1/TAZ target genes: ANKRD1 (ankyrin repeat domain 1), NEGR1 (neuronal growth regulator 1), and CTGF (connective tissue growth factor). The expression levels of ANKRD1, NEGR1, and CTGF mRNA were elevated in Sav1-deficient cells but were not increased by TGFβ 1 treatment (Fig. 4c). These data indicate that increases in TAZ protein expression induced by Sav1 depletion lead to enhanced TAZ-mediated transcriptional activation of ANKRD1, NEGR1, and CTGF; however, Sav1 does not appear to significantly alter the levels of YAP1 protein expression. It has been reported that YAP1 positively regulates TGF-β 2 expression 42 .
We therefore tested whether increased TAZ activation in Sav1-deficient cells is accompanied by enhanced TGF-β 2 expression. As shown Fig. 4d, TGF-β 2 mRNA levels were increased in Sav1-depleted cells that were preincubated with TGF-β 1 compared with control cells. We also found that YAP1 and TAZ overexpression stimulated TGF-β 2 expression (Fig. 4e). Furthermore, knockdown of YAP1 or TAZ individually did not significantly reduce TGF-β 2 expression; however, depletion of both YAP1 and TAZ significantly reduced TGF-β 2 expression (Fig. 4f). These data suggest that YAP1 activation or increased TAZ protein expression in the setting of Sav1 depletion upregulates TGF-β 2 expression, thereby contributing to TGF-β pathway activation; however, YAP1 protein expression was not altered in Sav1-depleted cells.
We previously noted that YAP1 protein was enriched in the genomic region of Tβ RII (TGF-β receptor II) in human MCF10A cells (unpublished data) and was potentially bound to the genomic region of TβRII in embryonic stem cells 43 . TAZ, a paralog of YAP1, also appeared to be enriched in the same region. These observations prompted us to investigate whether TβRII expression is directly regulated by YAP1/TAZ. TβRII mRNA expression was increased to greater extent in Sav1-depleted cells than in control cells and was increased by TGF-β treatment (Fig. 4g). Notably, knockdown of both YAP1 and TAZ reduced TβRII expression (Fig. 4h). To investigate the direct regulation of TβRII expression by YAP1/TAZ, we searched the TβRII genomic region based on an analysis of our unpublished YAP1-Chip sequencing profile data, which we integrated with TEAD4-Chip sequencing data provided by the UCSC browser. We identified one conserved genomic region upstream of the transcription start site of TβRII that may be bound by YAP1/TAZ and TEAD4 (Fig. 4i, yellow box). To verify this genomic region, we tested whether this region-driven luciferase reporter is activated by YAP or TAZ overexpression. We found that YAP1 and TAZ expression dramatically enhanced the luciferase activity of the promoter-reporter construct (Fig. 4j). We also confirmed that TAZ bound directly to the TβRII genome by performing ChIP (chromatin immunoprecipitation) assays using an anti-FLAG antibody in HEK293 cells expressing Flag-tagged TAZ. These assays showed that TAZ was highly enriched in the conserved region upstream from the TβRII transcription start site (Fig. 4k). Taken together, these results suggest that activation of TGF-β signaling by Sav1 deficiency is associated with increases in TGF-β 2 and TβRII expression mediated by YAP1/TAZ activation.
TAZ and YAP1 are implicated in renal TIF in vivo. Given that TGF-β 1 treatment increased TAZ protein levels, but not YAP1 protein levels (Fig. 4a,b), and that Sav1 depletion activated YAP1/TAZ target gene transcription (Fig. 4c), we examined TAZ mRNA expression in kidney tissues and assessed the functional role of TAZ in renal fibrosis. TAZ mRNA expression was increased in Sav1-null kidneys after UUO (Fig. 5b). Consistent with this finding, TAZ was present at high levels in the kidneys of CKD patients (Fig. 5a). We also tested whether TGF-β II and TβRII transcript levels are altered in Sav1-null mice exhibiting high TAZ expression. Sav1 depletion resulted in high TGF-β II and TβRII mRNA expression after UUO (Fig. 5b). As TGF-β II and TβRII gene expression is upregulated in Sav1-null kidneys, we investigated the localization of Smad4 as a marker of TGF-β qRT-PCR analysis of EMT markers from cells confirmed in (a). Col1a, Col3a, α -SMA, VIM, and SNAI2 expression was enhanced in Sav1-depleted HK2 cells exposed to TGF-β . CDH1 expression was decreased in the same cells. Error bars represent SEMs. * P < 0.05; ** P < 0.01; *** P < 0.005 by paired t-test (one-tailed).  Figure 3). Previous results demonstrated that knockdown of both YAP1 and TAZ downregulated TGF-β 2 and Tβ RII mRNA expression, even though YAP1 total protein expression was not affected in Sav1-depleted HK2 cells (Fig. 4f,h), indicating that YAP1 activation is required for TGFβ -signaling. We therefore surmised that the cellular localization of YAP1 is altered, as opposed to its protein levels. To determine whether YAP1 cellular localization is altered in Sav1-depleted kidneys after UUO, we observed YAP1 cellular localization in the TECs of control kidneys and Sav1-null kidneys after UUO (Fig. 5c). Irrespective of UUO, the basal levels of YAP1 protein expression were low, and YAP1 cellular localization was not altered in control kidneys. However, YAP1 accumulated in the nucleus in Sav1-null kidneys (Fig. 5c). suggesting that Sav 1 deficiency leads to nuclear accumulation of YAP1, which alters gene expression. We next determined whether TAZ plays a role in renal TIF via regulation of EMT-like phenotypic changes in vitro. Interestingly, TGF-β 1-induced increases in VIM expression were remarkably reduced in the absence of TAZ (Fig. 5d). Similarly, knockdown of TAZ led to strong reductions in SNAI2 and α -SMA expression, which persisted even during stimulation with TGF-β 1 (Fig. 5d). Taken together, these results suggest that activation of TAZ in Sav1 deficiency plays a role in renal TIF by strongly triggering EMT-like phenotypic changes.

Sav1 deficiency increases β-catenin activation in vitro and in vivo. Aberrant activation of Wnt/β -catenin
signaling is a common pathologic finding in renal fibrosis. Therefore, we investigated whether Wnt/β -catenin signaling is activated in Sav1-depleted cells. As shown in Fig. 6a, Ser552-phosphorylated β -catenin and active β -catenin levels were increased in Sav1-depleted cells compared with control cells and were increased to a greater extent in Sav1-depleted cells than in control cells by TGF-β 1 treatment (Fig. 6a). A functional consequence of increased β -catenin activity in Sav1-deficient cells (shSav #1-treated or #2-treated) was increased expression of the Wnt/β -catenin target genes c-MYC and AXIN2 at the mRNA and protein level; these increases, especially that of AXIN2, were enhanced by TGF-β treatment (Fig. 6b). Conversely, Sav1 overexpression reduced active β -catenin protein levels in HEK293 and HK2 cells, suggesting that Sav1 protein negatively regulates Wnt/β -catenin signaling (Fig. 6c). Consistent with these in vitro data, we observed that active β -catenin expression was increased after UUO in the obstructed kidneys of WT mice compared with the kidneys of sham-operated mice and was substantially enhanced in the obstructed kidneys of Sav1 fl/fl ;Ksp-Cre mice (Fig. 6d,e). These results suggest that Sav1 deficiency is associated with increased β -catenin activity in renal fibrosis.

Discussion
In this study, we demonstrated that the Hippo signaling pathway plays a pivotal role in the development of renal fibrosis. We focused on Sav1 deletion and subsequent TAZ activation in the Hippo signaling pathway as regulators of fibrosis in CKD because the expression of TAZ protein was markedly increased in mouse Sav1-depleted kidneys after UUO and in patients with TIF/CKD. In addition, we found that Sav1 depletion leads to YAP1 nuclear accumulation after UUO. UUO-induced TIF in renal TECs lacking Sav1 is aggravated by increased TAZ protein expression and YAP1 nuclear localization, resulting in TGF-β and TβRII expression upregulation and aberrant activation of Wnt/β -catenin signaling activation.
Several studies have reported the occurrence of crosstalk between the Hippo pathway and TGF-β signaling. For example, TGF-β stimulates the binding of TAZ to activated SMAD complexes in the nucleus, thereby enhancing responsiveness to TGF-β signaling 30,31 . However, we did not observe strong interactions between TAZ and SMADs in Sav1-deleted cells, even in the presence of TGF-β . Accordingly, other proteins may be involved in the crosstalk between the Hippo pathway and TGF-β signaling. In fact, YAP1 has been shown to regulate TGF-β 43,44 . We also confirmed that YAP1 and TAZ overexpression increased TGF-β 2 expression in vitro (Fig. 4e) and that YAP1 and TAZ knockdown reduced TGF-β 2 expression. Sav1 deficiency elevated TGF-β 2 and TGF-β RII mRNA expression in vivo and in vitro by increasing TAZ protein levels and YAP1 activity (Fig. 4d,g and Fig. 5b). Moreover, we demonstrated that TAZ directly induces TβRII mRNA (Fig. 4k). Conversely, TAZ is known to be transcriptionally induced by TGF-β 1 30 . Consistent with this finding, TAZ was upregulated over time by TGF-β 1 in vitro (Fig. 4a), and its mRNA expression remained at a high level in vivo after UUO (Fig. 5b). However, neither TGF-β 1 treatment nor Sav1 depletion affected cellular YAP1 protein levels. Interestingly, subcellular mRNA expression in HEK293 cells transduced with (d) Sav1-depleting shRNA vectors or control vectors and incubated with (dark gray bars) or without (light gray bars) TGF-β 1 for 12 hours. P = 0.03522 (comparison between the 3rd bar and 4th bar). (e) TAZ or YAP1 expression plasmids and (f) TAZ-, YAP1-or TAZ/YAP1depleting shRNA vectors. P = 0.00093 (comparison between the 1st bar and 4th bar). (g) The increase in TβRII mRNA expression elicited by TGF-β 1 treatment in the Sav1 deficiency. qRT-PCR analysis of TβRII mRNA expression in Sav1 depleted-HK2 and control cells treated with or without TGF-β 1. P = 0.03204 (comparison between the 3rd bar and 4th bar). (h) Downregulation of TβRII mRNA expression by deletion of both YAP1 and TAZ in HEK293 cells.. P = 0.0389 (comparison between the 1st bar and 4th bar). (i) Schematic of part of the TβRII genomic region containing a YAP1/TAZ-enriched site showing TEAD binding sequences (GGAATG/ CATTCC) located between 30638606 and 30639826 in chromosome 3, a YAP1 or TAZ-enriched region (yellow box), and the TβRII transcription start site (TSS). (j) Luciferase activity assays of the TβRII genomic upstream region shown in (i). 293T cells transfected with an pGL3 vector containing a genomic fragment including the YAP1/TAZ-enriched region were co-transfected with YAP1 or TAZ expression plasmids or control vectors. (k) ChIP assay for TAZ-enriched regions. immunoprecipitated DNA with an anti-Flag antibody was amplified with specific primers covering potential TEAD binding sequences. All error bars represent SEMs. * P < 0.05; ** Relative TAZ and TGF-β 2 expression and TβRII transcript levels were measured in the kidneys of WT and TEC-specific Sav1-null mice after UUO using qRT-PCR. Sav1-null mice kidneys exhibited increased TAZ, TGF-β II, and TβRII mRNA expression. Error bars represent SEMs. * P < 0.05; ** P < 0.01; *** P < 0.005 by paired t-test (one-tailed). P = 0.00043, 0.03991, 0.00254 (comparison between the 1st bar and 3rd bar) and 0.00048, 0.01366, 0.00024 (comparison between the 1st bar and 4th bar) for ANKRD1, CTGF and NEGR1, respectively. (c) Sav1 deficiency after UUO resulted in nuclear accumulation of YAP1. Immunofluorescence staining for YAP1 (Red) with anti-YAP1 antibodies in control and TEC-specific Sav1-null mice. YAP1 was strongly localized in the nuclei of SAV1-nul kidneys after UUO. (d) qRT-PCR analysis of the EMT markers VIM, SNAI2, and α -SMA in control and TAZ-knockdown HK2 cells during incubation with TGF-β 1. TAZ knockdown reduced TGF-β 1-induced EMT marker expression. Error bars represent SEMs. n.s., not significant; * P < 0.05; ** P < 0.01 by paired t-test (one-tailed). Scale bar in a = 50 μ m and c = 20 μ m.
Scientific RepoRts | 6:31931 | DOI: 10.1038/srep31931 YAP1 localization was strongly maintained in renal TECs lacking Sav1 after UUO. Given these observations, we assumed that the UUO procedure itself stimulates TGF-β 1 secretion, resulting in increased TAZ activation and YAP1 nuclear accumulation in the setting of Sav1 depletion, as well as subsequent TGF-β signaling activation by TGF-β 2 and TβRII induction through TAZ activation. In line with these results, TAZ depletion prevented induction of the expression of the EMT markers VIM, SNAI2, and α -SMA in response to TGF-β treatment in vitro (Fig. 5d). Based on these results, we propose that YAP1/TAZ activation after UUO in TEC-specific Sav1-null mice promotes TGF-β -induced EMT-like phenotypic changes that contribute to renal TIF. In addition, CTGF, which is encoded by a known direct target gene of TGF-β and YAP1/TAZ, induces sustained fibrosis along with TGF-β 37,45 .
Our observations indicate that CTGF mRNA expression was enhanced by YAP1/TAZ activation, which may exacerbate renal fibrosis.
Wnt/β -catenin signaling activation is a common feature in a variety of CKDs. Notably, β -catenin plays a critical role in renal fibrosis. The current study also showed that Sav1 depletion resulted in increased active β -catenin expression and revealed that TGF-β treatment and UUO enhanced Wnt/β -catenin signaling activity. Unfortunately, the mechanism linking the loss of Sav1 to the increase in active β -catenin expression remains unclear. It is reasonable to assume that high levels of active β -catenin protein expression are not directly induced by Sav1 depletion but instead may be a secondary effect of intrinsic TGF-β signaling activation by Sav1 deletion. This idea is supported by the previous finding that TGF-β signaling itself upregulates Wnt/β -catenin signaling through suppression of Dkk1 (Dickkopf WNT signaling pathway inhibitor 1) or Klotho mRNA expression 46,47 .
In summary, our data show for the first time that the Hippo-Salvador pathway regulates renal TIF through TGF-β and Wnt/β -catenin signaling (Fig. 7). Our results indicate that the Hippo-Salvador pathway is a new mechanism underlying the pathogenesis of TIF development and that regulation of Hippo signaling may represent a therapeutic strategy for mitigating TIF.

Materials and Methods
Human kidney specimens. Human kidney samples were obtained from nine archived kidney biopsy specimens from patients with chronic renal TIF (three IgA nephropathy and two membranous nephropathy) and from normal kidney tissue specimens obtained via nephrectomy at the Bucheon St. Mary's Hospital Department of Pathology. The study was approved by the Medical Ethics Committee of Bucheon St. Mary's Hospital.
Animals. Floxed Sav1 mice (Sav1 fl/fl ; previously designated WW45 fl/fl ) were generated as described previously 27 . To generate mice with a Sav1 deletion specifically in the tubules, we crossed Sav1 fl/fl mice with transgenic Ksp-Cre mice, which were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). The genotypes of the offspring were determined by polymerase chain reaction (PCR) analysis using genomic DNA obtained from the tails of mice and transgene-specific primers. All mouse lines were bred onto a C57BL/6 background. Only male mice (20-25 g, 8 weeks old) were used in this study. UUO was performed using an established procedure. All animal experiments were approved by the Ethics Committee of Bucheon St. Mary's Hospital.
Immunohistochemistry. Immunohistochemical staining was performed according to a previously established protocol 28  Immunofluorescence staining and confocal microscopy. Immunofluorescence procedures were performed as previously described 28 . Cell proliferation was detected using a primary polyclonal antibody against Figure 7. Schematic representation of the proposed role of the Hippo-Salvador signaling pathway in tubular epithelial cells in renal tubulointerstitial fibrosis. YAP1/TAZ activation in Sav1-depleted cells induces TGF-β and TβRII mRNA expression, thereby synergistically enhancing responsiveness to TGF-β . This mechanism promotes TGF-β -induced EMT-like phenotypic changes and contributes to progressive renal TIF. Sav1 depletion also enhances Wnt/β -catenin activation, which exacerbates TIF.

Cell culture and viral infection. 293T cells and HEK293 cells were cultured in Dulbecco's modified Eagle
medium (DMEM) supplemented with 10% fetal bovine serum (FBS). HK-2 cells were cultured in DMEM/F12 containing 10% FBS. For TGF-β 1 treatment, cells were cultured in serum-free medium for 18 h and then treated with or without TFG-β 1 (2 ng/ml) for 12 h. Flag-YAP1, Flag-TAZ, and Flag-Sav1 were cloned into a retroviral pMSCV plasmid. shYAP1 and shTAZ were cloned into Super Retro plasmids. 293T cells were transfected with retroviralvectors (5:5:2 ratio of retroviral vector:Gag-pol:VSVG) using the calcium precipitation method. Viral supernatants were collected 1 day after transfection. Three days after transfection, viral supernatants were cleared by centrifugation at 3000 rpm for 20 minutes and filtered using a 0.45-μ m filter (Millipore). Target cells were infected with the resulting viruses using 6 μ g/ml polybrene (Sigma, 107689) and were selected 2 days later by culturing in the presence of 3-6 μ g/ml puromycin. shSav1 was cloned into the pLKO.1-puro lentiviral vector. 293T cells were transfected with lentiviruses (6:4.5:4.5:3 ratio of lentiviral vector:P1:P2:VSVG) using the calcium phosphatase method. Virus collection, infection and selection were performed using the same methods as those used for the retroviruses.

Analysis of gene expression by quantitative real-time PCR.
Total RNA was prepared using the TRIzol reagent (Invitrogen), as directed by the manufacturer. Quantitative real-time PCR analyses were carried out as previously described 48 using primers described in Supplemental Information.
Luciferase assays. Genomic fragments of Tβ RII, which were generated by PCR, were subcloned into an FGF4 minimal promoter-linked pGL3 vector expressing Firefly luciferase and co-transfected into cells with YAP1 or TAZ expression plasmids or control vectors, as well as a Renilla luciferase expression plasmid, using a Dual-Luciferase Reporter Assay (Promega), as directed by the manufacturer. Luciferase activity was measured and normalized to Renilla activity to control for transfection efficiency.
ChIP-Seq and data analysis. ChIP-Seq procedures and data analysis were performed as previously described 46 . The sequences of the primers are described in Supplemental Figure S3.

Statistical analyses.
Results are presented as means ± SDs. A t-test was used for comparisons between two groups. Differences among two or more groups were compared using one-way analysis of variance (ANOVA). P-values less than 0.05 were considered significant. Study approval. Human kidney samples were obtained from patients with chronic renal TIF and from normal kidney tissue specimens obtained via nephrectomy at the Bucheon St. Mary's Hospital Department of Pathology. The study was approved by the Medical Ethics Committee of Bucheon St. Mary's Hospital (HC14SISI0048). All human data and samples used in this study were anonymized. All identifiers of human data or samples were irreversibly stripped via an arbitrary alphanumeric code, making it impossible for anyone to link the samples to their sources. Therefore, no informed consent was obtained from the research participants. All experimental procedures were performed according to animal care and ethics legislation, and the study was approved by the Animal Care Committee of Bucheon Saint Mary's Hospital.