Klotho is regulated by transcription factor Sp1 in renal tubular epithelial cells

Background Klotho is a multifunctional protein, which exists both in a membrane bound and a soluble form. In renal tubules, Klotho is involved in cell senescence, anti-oxidant response, and renal fibrosis, thus regulation of its expression is critical to understand its roles in renal diseases. Indeed, reduced expression was observed in various renal disease. However, the mechanisms underlying transcriptional regulation of the human klotho gene (KL) largely remain unknown. Results Here we demonstrated that the Klotho expression in human renal tubular epithelial cells (RTECs) was enhanced by overexpression of the transcription factor Sp1. On the contrary, Klotho expression was decreased by Sp1 knockdown. Besides, increased expression of Sp1 alleviated TGF-β1-induced fibrosis in HK-2 cells by inducing Klotho expression. Luciferase reporter assays and chromatin immunoprecipitation assays further identified the binding site of Sp1 was located in − 394 to − 289 nt of the KL promoter, which was further confirmed by mutation analysis. Conclusions These data demonstrate that KL is a transcriptional target of Sp1 and TGF-β1-induced fibrosis was alleviated by Sp1 in human RTECs by directly modulating Klotho expression, which help to further understand the transcriptional regulation of Klotho in renal disease models.


Background
Klotho (KL) is originally identified by an accidental site defect, which is related to premature multiple organ failure. Owing to alternative splicing, the human KL encodes two forms of proteins, which are predominantly expressed in human renal tubular epithelial cells (RTECs). One exists as a full-length membrane-associated form, whereas the other exists as a secreted form lacking the transmembrane segment and the intracellular domain [1]. The membrane Klotho can form a high-affinity co-receptor with fibroblast growth factor (FGF) receptors for FGF23, and thereby contributes to the signal transduction of FGF23 [2,3]. The secreted Klotho is predominantly detected in cerebrospinal fluid and circulation and is involved in the regulation of anti-oxidative capacity, growth factors pathway and ion transport [4][5][6].
As known, KL is expressed predominantly in kidney, parathyroid gland and choroid plexus [7,8]. A significantly reduced Klotho was observed in patients with either acute or chronic kidney disease (CKD) [9,10]. Moreover, KL −/− mice experience growth retardation, hypokinesis, osteoporosis, shorten lifespan, while KL transgenic mice exhibit increasing resistance to insulin and extending lifespan. Varieties of physiological and pathological factors contribute to the regulation of KL expression [4,9,10], but the transcriptional regulatory mechanism underlying the KL expression is not entirely clear.
Sp1 is a eukaryotic transcription factor highly conserved among mammalian species [11]. It is documented that more than 12,000 Sp1 binding sites have been found in the human genome [12]. Historically, Sp1 has been regarded as a ubiquitous transcription factor responsible for basal expression of housekeeping genes [13]. However, recent studies have revealed that Sp1 is involved in regulating, either inducing or inhibiting transcription of numerous cell type-specific genes [14]. Sp1 can both activate and suppress the expression of genes implicated in senescence, proliferation, differentiation and apoptosis [15], and also involved in inflammation, epigenetic modification and chromatin remodeling [11]. In renal tubular epithelial cells, Sp1 was previously reported to regulate CD2AP promoter activity and expression, suggesting Sp1 is functional in RTECs [16]. Lately, it was reported that both Sp1 and Klotho were significantly decreased in hypoxia/reoxygenation (H/R)-injured RTECs and exogenous Sp1 or Klotho could separately function as the protector during H/R injury [17,18]. Moreover, in LPS-induced inflammation injury, LPS could down-regulated Sp1-mediated gene transcription, while Klotho was significantly reduced during LPSinduced injury [19,20]. Further bioinformatics analysis showed no typical TATA or CAAT boxes were found in the human KL promoter. Instead, 5 potential Sp1 binding sites were predicted [8]. Thus, we assume that there may be some causative linkers between Klotho and Sp1. However, transcriptional regulation of KL by Sp1 has still not been reported.
As is reported, TGF-β1 has been participated in renal fibrosis through inducing epithelial-to-mesenchymal transition (EMT) in RTECs [21]. Evidences also support that EMT has been demonstrated to lead to renal fibrosis [22], the final common pathway to end-stage kidney disease (ESRD). Previous study reported that TGF-β1 could decrease Klotho protein expression, while as a therapeutic target, elevated Klotho expression could inhibit TGF-β1-induced EMT [23].
In the study, we investigated whether Sp1 could affect Klotho expression and TGF-β1-induced fibrosis in human RTECs. Moreover, we further detected the underlying mechanism of regulation of Klotho expression by Sp1.

Overexpression of Sp1 dose-dependently induced Klotho expression in human kidney cells
To examine the specific involvement of Sp1 in regulating Klotho expression, a Sp1 expression plasmid (pcDNA3-Sp1) was transfected into HK-2 cells to introduce overexpression of Sp1 and Klotho expression was measured both at protein and mRNA levels. As shown in Fig. 1a, Sp1 vector could be efficiently expressed in HK-2 cells. Concomitantly, increased Klotho expression both at protein level and mRNA level was observed in a dosedependent manner ( Fig. 1a-b). To further confirm this finding, another cell line was also transfected with pcDNA3-Sp1. Similarly, dose-dependent induction of Klotho expression by Sp1 was also found in HEK-293 cells ( Fig. 1c-d). Thus, Sp1 could positively regulate the Klotho expression in human RTECs.
Knockdown of Sp1 reduced Klotho expression and fibrosis marker in human kidney cells As overexpression of Sp1 could regulate the Klotho expression, we further wondered the effect of Sp1 knockdown on gene expression of Klotho. Consistently, Sp1 knockdown using a specific siRNA could significantly reduce endogenous Sp1 protein level (Fig. 2a). As expected, suppressed Klotho expression both at protein and mRNA levels was observed in Sp1 knockdown HK-2 cells ( Fig. 2a-b). Again, we further repeated the experiment in HEK-293 cells, and the result showed that Klotho expression was noticeably reduced by Sp1 knockdown ( Fig. 2c-d). These observations suggested that Klotho expression could be regulated by Sp1 in RTECs.
As reduced Klotho expression was observed in renal fibrosis, we wondered the effects of Sp1 knockdown on renal fibrosis markers. Interestingly, the protein levels of renal fibrosis markers, including α-SMA, fibronectin and E-cadherin, were remarkably altered by Sp1 knockdown in HK-2 cells (Fig. 2e). These demonstrated that Sp1 knockdown could induce renal fibrosis by downregulation of Klotho in RTECs.

Sp1 correlated with Klotho expression in vivo
As known, Klotho is predominantly expressed in human renal tubules. To further investigate the relationship between the Sp1 level and the Klotho expression in vivo, GSE32591 dataset, which contains the gene expression data of micro-dissected human tubulointerstitial compartments from cortical tissue segments of the healthy human kidney, was selected. Here we found that the mRNA expression level of Sp1 in human tubulointerstitial from cortical tissue of health kidney was positively correlated to mRNA expression level of Klotho (Fig. 3). These data suggested that Sp1 might regulate Klotho expression in vivo.

Overexpression of Sp1 alleviated TGF-β1-induced fibrosis in HK-2 cells
In HK-2 cells, we found that TGF-β1 treatment had no effect on Sp1 expression (Fig. 4a), whereas it dosedependently increased the expression of α-SMA and fibronectin, and decreased Klotho and E-cadherin expression (Fig. 4b). Further, we also found that short-term treatment of TGF-β1 had no effect on Sp1 expression in HK-2 cells (Fig. 4c). After treatment with TGF-β1, the morphology of HK-2 cells had changed from normal to shuttle shape. However, the morphological changes were alleviated by pre-transfection with a Sp1 expression plasmid (Fig. 4d). In addition, decreased expression of Klotho and E-cadherin, and increased expression of fibronectin and α-SMA after the TGF-β1 treatment could be partially reversed by Sp1 pre-transfection into HK-2 cells ( Fig. 4e & f). The data suggested that Sp1 overexpression could remarkably suppress TGF-β1-induced fibrosis in HK-2 cells.

Defining the Sp1-responsive region in Klotho promoter
Then we wondered whether Sp1 binding sites existed in Klotho promoter. Bioinformation analysis of human putative Klotho promoter showed that there were 5 potential binding sites (GC boxes) for Sp1 binding in this region. Reporter assay of Full KL/LUC showed that expression vector pcDNA3-Sp1 dose-dependently enhanced the relative luciferase activity (Fig. 5a), suggesting that Sp1 could regulate the transcriptional activity of KL promoter in HK-2 cells. Then, a series truncated reporter was constructed to define the Sp1responsive region (Fig. 5b). To test the inducted activity by Sp1, pcDNA3-Sp1 or control was co-transfected into HK-2 cells with these reporters. Interestingly, induced reporter activity by Sp1 was only detected in KL5/LUC construct, which contained − 430 nt to + 8 of Klotho promoter. Its induction by Sp1 was similar to that of full-length Klotho promoter. Compared with nonresponsive construct KL4/LUC, KL5/LUC construct contained one additional putative binding site of Sp1. To further confirm the involvement of this binding site, the consensus sequence was mutated by site-specific mutation. As expected, no induced reporter activity by Sp1 was observed in the mutated reporter (Fig. 5c).
Finally, ChIP analysis was performed to further confirm the binding of Sp1 in Klotho promoter. After immunoprecipitation with specific antibody against Sp1, the DNA fragments were detected by PCR with primers spanning − 394 and − 289 nt of the KL promoter (Fig. 6a). As expected, positive bands were appeared in Sp1 antibody-immunoprecipitated DNA sample, but not in IgG group. In addition, more tensive band was detected in Sp1 transfected HK-2 cells, compared with that in empty vector transfected cells (Fig. 6b). Taken together, Sp1 could bind with the Klotho promoter directly.

Discussion
In this study, we described a transcriptional mechanism underlying the gene regulation of KL, a crucial agingsuppressor gene. Klotho expression was enhanced by Sp1 overexpression, while inhibited by Sp1 knockdown. Besides, we found that increased expression of Sp1 in human RTECs attenuated TGF-β1-induced fibrosis by increased expression of Klotho. More importantly, we identified the specific Sp1 binding site within the KL promoter region in HK-2 cells.
As a general expressed transcription factor, Sp1 is regarded as a member of the C2H2-type zinc finger family that contains DNA-binding proteins and is highly conserved among mammalian species [24][25][26]. Sp1 directly bond to GC-rich motifs and is involved in the expression of genes as evidenced by more than 12,000 Sp1 binding sites within the human genome [11,27]. Sp1 also has been shown to transcriptional regulate a large number of cellular genes critically related to essential cellular functions, including differentiation, senescence, proliferation, and apoptosis [27][28][29]. In addition, Sp1 could regulate genes with opposing biological functions. For example, Sp1 regulated both oncogenes and tumor suppressor genes, as well as survival-promoting genes and apoptotic-promoting genes [30][31][32]. However, fully understanding of its roles in various promoters and under various cellular conditions is needed to elucidate contributions to tumorigenesis and other disease processes. The human KL promoter, a GC-rich promoter, has 5 potential Sp1 binding sites rather than typical TATA or CAAT boxes [8]. Our luciferase reporter and ChIP assays demonstrated that the specific Sp1-binding site is located between nt − 430 and − 380 in the KL promoter region. It is noteworthy that Klotho is predominantly expressed in RTECs where it regulates senescence, tumor growth, and renal fibrosis [1,4,6]. The regulation of Klotho expression is affected by different physiological and pathological conditions [9,33]. Patients with AKI exhibit dramatically Klotho reduced in kidneys, and scientists identified Klotho acts as an early biomarker in AKI [10]. Furthermore, renal Klotho decreased in mice ischemia reperfusion injury (IRI), while overexpression Klotho played a renoprotective role in mice with ischemia-reperfusion injury (IRI) [10,18,34]. Interestingly, scientists reported that the human-induced pluripotent stem cell-derived mesenchymal stromal cells fused with target kidney cells and then delivery Sp1 to protect rat IRI kidney and hypoxia/reoxygenation -injured HK-2 cells, meanwhile Sp1 inhibition totally eliminated the renoprotective role of hiPSC-MSCs-EVs in IRI [17]. To explore the relationship between Sp1 and Klotho, we confirmed that Klotho is positively regulated by the transcription factor Sp1 in cultured human kidney cells. Regulation of Sp1 expression by TGF-β1 is contradictory in previous research. Some reported that TGF-β1 could upregulate Sp1 expression in rat renal tubular epithelial cells (NRK-52E) [14]. In contrast, several studies observed that after treatment with TGF-β1, only a slight induction or no Sp1 response was observed. Zhou Q et al. found that Fos and Val inhibited TGF-β1 expression, thus upregulated Sp1 and Klotho expression to protect Ang II-induced renal damage in NRK-52E cells [35]. In human dermal fibroblasts, the slight induction of TGF-β1 supplementation on the Sp1 mRNA and protein expression was not statistically significant [36]. Scientists demonstrated that TGF-β1 supplementation had no effect on the expression of Sp1 mRNA or protein in human stellate cells [12]. Therefore, scientists draw the conclusion that TGF-β1-induced Sp1 expression is usually weak and dependent upon cell line. Our results showed that TGF-β1 did not affect Sp1 expression in HK-2 cells.
It has been reported that TGF-β1 promoted renal fibrosis through inducing EMT in RTECs, and this was partially through inhibiting Klotho expression. In our research, when Sp1 was transfected into HK-2 cells, the expression of Klotho was increased, whereas the TGF-β1-induced EMT was alleviated, suggesting that TGF-β1-induced fibrosis could be attenuated by elevated Sp1 expression and Sp1 contributed to the anti-EMT effect through regulating Klotho expression in HK-2 cells.
To explore the mechanisms underlying the regulation of Klotho by Sp1, our luciferase reporter and ChIP assays suggested that Sp1 may differentially modulate klotho expression under specific conditions, although the precise upstream signaling pathways remain to be elucidated. Further, as the Sp1 binding site is relatively far away from the transcription start site, other transcription factors are still needed to be investigated in the regulation of KL expression.

Conclusions
Collectively, our study demonstrated that overexpression of Sp1 alleviated TGF-β1-induced fibrosis in HK-2 cells by inducing Klotho expression and Sp1 directly modulated Klotho expression in kidney cells by binding to a specific CG-rich site between nt − 430 and − 380 in the KL promoter region, which help to further understand the transcriptional regulation of Klotho in renal disease models.

Plasmid and reporter gene construct
The coding region of Sp1 was amplified by RT-PCR (primer sequences in Table 1) from HK-2 cells and was inserted into pcDNA3.0 vector to construct recombinant plasmid pcDNA3-Sp1 which were further confirmed by sequencing. The human KL promoter region (− 1900 to + 8) was cloned into the pGL3-basic reporter to yield Full KL/LUC (a gift from Professor Gu Jun, Peking University, China). A series of truncated deletions in the 5′ region of the KL promoter were also constructed by insertion of PCR products into pGL3basic by homologous recombination, yielding -430KL5/LUC, −230KL4/LUC, −193KL3/LUC, − 82 KL2/LUC, and -50KL1/ LUC (the first number of each construct indicates the first nucleotide of the KL promoter, with the transcriptional start site defined as + 1) [17]. To test the binding specificity, the potential Sp1 binding site in KL5/LUC was mutated using a PCR-based site-directed mutagenesis kit (Sangon Biotech, Shanghai, China) (from ATCCCGCCCCC to ATCCTTAC CCC). All primers are listed in Table 1.

Cell culture and treatment
HK-2 (the human renal proximal tubular cell line) and HEK-293 (the human embryonic kidney cell line) obtained from ATCC were separately cultured in DMEM/ F12 and DMEM-basic medium with 10% fetal bovine serum (FBS), in a 37°C incubator containing 5% carbon dioxide. After planked for 24 h, HK-2 cells were treated with TGF-β1 (1 ng, 5 ng, 10 ng/mL) for 48 h.

Cell transfection
The specific vectors were transiently transfected into the cultured HK-2 or HEK-293 lines using Lipofectamine 2000, a reagent from Invitrogen. SiRNA-based strategies were used to silence endogenous Sp1 expression in HK-2 and HEK-293 cells. Sp1 siRNA sequence was 5′-AUCACUCCAUGGAU GAAAUGATT-3′ [16]. A random scrambled siRNA was used as negative control. Sp1 siRNA was purchased from Gene Pharma. HK-2 and HEK-293 cells were transfected with specific constructs. After transfection, cells were further cultivated until used in subsequent experiments. To further explore the role of Sp1 on HK-2 cells, which were treated with TGF-β1 (10 ng/mL) for 48 h following Sp1 vector were pre-transfected. In each well of 24-well plate, 375 ng for pc-DNA3-Sp1 transfection, 125 ng for full KL, KL5, KL4, KL3, KL2, KL1, PGL-3 basic transfection, and 5 ng for pRL-TK transfection was used for testing relative LUC activity.

Real-time quantitative PCR
Cell total RNAs from HK-2 and HEK-293 cells was extracted using Trizol (Invitrogen), and cDNA was conducted using a Reverse Transcription Kit (TaKaRa Bio, Japan) following the recommendations of manufacturer. The PCR reaction was then performed in a programmed thermal cycler using a Bio-Rad IQ5 Detection System as previously described. The primer pairs for RT-qPCR are shown in Table 2. The relative expression levels specific genes were based on the 2(−ΔΔCt) approach after normalizing to GAPDH expression.

Immunofluorescence assay
To visualize fibrosis marker, after staining with an antiα-SMA antibody (1:1000), the slides were incubated with fluorescent secondary antibody for 50 min. Then, DAPI was used to stain the nuclei and visualized on a ZEISS confocal microscope.

Luciferase reporter assay
HEK-293 cells were seeded and co-transfected with Klotho reporters plus the Sp1 expression plasmid using Lipofectamine 2000 reagent. An internal control Rinella luciferase plasmid driven by the TK promoter (pRL-TK) was co-transfected as an internal control. 24 h after transfection, cells were washed in precooled PBS, and samples were prepared in 1 × passive lysis buffer (PLB). The Dual Luciferase Reporter Assay Kit (Promega) was used to detect the relative luciferase activity following the manufacturer's protocol [18]. Expression was assessed as fold change over control after normalizing for transfection efficiency.

Chromatin immunoprecipitation (ChIP) assays
ChIP was performed by using a commercially available EZ-ChIPTM kit (Merck Millipore), accoding to the manufacturer's instruction. HK-2 cells were transfected with pcDNA3-Sp1 for 24 h, harvested, washed with precooled PBS, and then cross-linked by 1% formaldehyde. Crosslinking reactions were halted by addition of glycine. Diluted lysates were prepared and sonicated to yield fragments of 200-500 bp. The sheared cross-linked chromatin fragments were immunoprecipitated with Sp1 antibody or IgG (control) overnight, collected, washed, and reversed cross-fixed at 65°C. Purified DNA was collected from Spin Columns and used as the template in PCR [18]. Primer sequences designed for two different KL regions are listed in Table 2.