1. S100A16 knockout decreases HIF-1α expression in AKI mice.
In our previous study, we reported that S100A16 promotes acute kidney injury (AKI) by activating E3 ubiquitin ligase, the HMG-CoA reductase degradation protein 1 (HRD1), induced ubiquitination and degradation of GSK3β and CK1α, two β-catenin complex members [7]. β-catenin is subsequently released in response to AKI. However, it is unknown why HRD1 is increased in AKI. Hypoxia-inducible Factor 1-alpha (HIF-1α) is a transcription factor that plays a crucial role in the regulation of cellular responses to hypoxia. In this study, the renal ischemia-reperfusion injury (IRI) model in S100A16 knockout mice (S100A16+/-) was used to determine if HIF-1α is associated with HRD1 upregulation in AKI.
The histological results in mouse kidneys by hematoxylin-eosin (HE) staining demonstrated that IRI induced kidney injury including tubular dilatation and glomerular atrophy. S100A16 knockout attenuated IRI in mice (Fig. 1A). The expressions of S100A16 and HRD1 detected by western blots were significantly increased in the IRI group compared to the sham-operated group in WT mice (Fig. 1B and 1C), which are consistent with our previous reports [7]. Interestingly, we also observed that the expression of HIF-1α was mainly expressed in renal tubular epithelial cells, and significantly increased in the kidneys in the IRI mice compared to the sham-operated group (Fig. 1D). S100A16 knockout attenuated the increased expression of HIF-1α induced by the IRI (Fig. 1B and 1C). Immunohistochemical results in HIF-1α and HRD1 expressions are also consistent with its biochemistry data (Fig. 1D). The reciprocal relationship of HIF-1α and S100A16 in IRI mice led us to hypothesize that S100A16 upregulated HRD1 level through HIF-1α activation in renal injury condition.
2. S100A16 knockout attenuates the injuries induced by hypoxic reoxygenation and TGF-β1 stimulation in NRK-52E cells.
To further determine whether S100A16 modulates HIF-1α expression in renal injury, we used CRISPR/Cas9 gene editing technology to knockout S100A16 in rat renal tubular epithelial cells (NRK-52E cells). After analyzing rat derived S100A16 gene (Gene ID: 361991) in GENEBANK, four sgRNAs pairs in the second S100A16 exon were designed using the online design tool (https://www.zlab.bio/resources) (Fig. 2A). The four pairs of sgRNAs were ligated to the PX330 plasmid to form the ligation product S100A16-Cas9/sgRNA. After sequencing and identification, the four pairs of ligation products were transfected into NRK-52E cells for T7E1 cleavage assay and sequence analysis. The sgRNA cleavage efficiency is shown in Fig. S1A. sgRNA3>sgRNA1>sgRNA4>sgRNA2, consistent with the results of website analysis (https://tider.deskgen.com/) (Fig. S1B). The sgRNAs with high cleavage efficiency (sgRNA3 and sgRNA4) and pCMV-TD-Tomato were selected for transfection into NRK-52E cells in a 5:1 ratio, and the transfected cells were treated with G418 (Fig. 2B) with 19 positive monoclonal drugs screened clones from 33 monoclonal cell lines (Fig. S1C). Three S100A16-/- cell clones were selected for validation at the protein (Fig. 2C) and mRNA levels of S100A16 (Fig. 2D). S100A16-/- cells were also confirmed by cellular immunofluorescence (Fig. S1D). Compared to WT cells, S100A16-/- cells have no difference in cell morphology (Fig. S2A) and growth viability (Fig. S2B). Western Blot analysis showed that S100A16 knockout effectively inhibited H/R-induced upregulation of pro-apoptotic genes, including BAX, Cleaved Caspase3 and Caspase3 and downregulation of Bcl-2 (Fig. 2E and 2F). S100A16 knockdown also suppressed the TGF-β1-induced upregulation of Fibronectin, α-SMA expression, an indicator of renal fibrosis (Fig. S2C and S2D). These data show that knockout of S100A16 can effectively attenuate apoptosis caused by hypoxic reoxygenation and TGF-β1-induced cell injury.
3. S100A16 modulates the expression of HIF-1α and HRD1 in NRK-52E cells.
Next, using S100A16-/- cells, we determined if S100A16 regulated HIF-1α and HRD1 with Wnt/b-catenin signaling pathway in a cellular hypoxia-reoxygenation (H/R) model. Our results revealed that H/R injuries significantly increased both protein and mRNA levels of HIF-1α, HRD1 and b-catenin, decreased GSK-3b in NRK-52E cells. However, the knockout of S100A16 significantly reversed the expressions of HIF-1α, HRD1, b-catenin and GSK-3b in S100A16-/- cells under the H/R conditions (Fig. 3A and 3B). Similarly, the transcript levels of HIF-1α and HRD1 were all upregulated in NRK-52E cells under the H/R conditions (Fig. 3C).
In contrast, S100A16 overexpression in NRK-52E cells by the transient transfection with S100A16 plasmid increased the expressions of HIF-1α, HRD1 and b-catenin, decreased GSK-3b protein level (Fig. 3D). The transcript levels of HIF-1α and HRD1 were all upregulated that consistent with their protein expressions (Fig. 3E). The changes of HRD1, b-catenin and GSK-3b in NRK-52E cells transfected with S100A16 consistent with our previous reports. Our concept is that HRD1 can ubiquitously degrade GSK-3b, thus allowing cytoplasmic b-catenin accumulation into the nucleus to initiate transcription of downstream genes. We assumed that the upregulation of HRD1 was a result of HIF-1α, which was growing in H/R-damaged NRK-52E cells.
4. HIF-1α transcriptionally upregulates HRD1 expression in HK-2 cells
To clarify the relationship between HIF-1α and HRD1, we used human kidney-2 (HK-2) cells which are human renal proximal tubular epithelial cells to perform the experiments of HIF-1α overexpression.We transfected the HIF-1α plasmid into HK-2 cells. The results revealed that overexpression of HIF-1α significantly upregulated HRD1 and b-catenin, and downregulated GSK-3b (Fig. 4A). At the mRNA level, overexpression of HIF-1α, the expression of HRD1 was significantly increased (Fig. 4B). In the cellular H/R model, inhibition of HIF-1α expression by BAY, a HIF-1α inhibitor, reversed the upregulation of HRD1 and b-catenin expression and the downregulation of GSK-3b by cellular H/R injury (Fig. 4C). The mRNA level of HRD1 was suppressed after inhibition of HIF-1α expression using BAY in the cellular H/R model (Fig. 4D). The results suggest that HIF-1α directly regulates the transcription of HRD1.
We hence hypothesized that HIF-1α is a transcription factor of SYVN1 (HRD1). To test this, we analyzed the potential binding sites of the HRD1 promoter with HIF-1α at the website https://jaspar.genereg.net/. We identified five potential binding sites in the HRD1 promoter (Supplementary Table 1.1). We designed three regions at the -1824 ~ -1458 bp (HRE1), -1439 ~ -985 bp (HRE2), -286 ~ +39 bp (HRE3) position of the HRD1 gene to determine the interaction between HRD1 promoter and HIF-1α (Fig. 4E)[16]. Chromatin immunoprecipitation (ChIP) assay revealed that there are HIF-1α binding signals on the HRD1 promoter region at the HRE3 area in the HK-2 cells (Fig. 4F).
To test if HIF-1α binds to this sequence to exert its transcriptional regulation of HRD1, we constructed the reporter plasmid consisting of the HRD1 promoter sequence HRE1, HRE2, HRE3 and full length (-2000 ~ +46 bp), followed by a luciferase coding sequence (HRD1-Luc). We transfected the HK-2 cells with HRD1-Luc or empty vector, followed by transfection of these cells with HIF-1α plasmid. The relative luciferase signals from each of the combinations in HK-2 cells were quantified. As expected, HIF-1α increased the luciferase signal levels in HK-2 cells that were transfected with HRD1(HER3)-Luc (Fig. 4G). The data indicated that the HIF-1α’s binding to the HRD1 promoter induced the level of transcription of HRD1.
5. TFAP2B and S100A16 are upregulated in injured HK-2 cells and UUO mouse kidney
In our previous studies, we reported that the expression of S100A16 is significantly increased in kidney biopsy specimens from patients with various clinical nephropathy and kidney disease mouse models including IRI model and unilateral ureteral occlusion (UUO) model [6, 12]. However, the reasons why S100A16 highly expressed in renal disease are still unclear.
To explore the mechanism of an increased S100A16 expression in kidney injury, we analyzed the differential expression of CKD samples (GSE66494) from the GEO database (Supplementary Table 2). The results revealed that the expression of TFAP2B, a transcript factor, was significantly upregulated in kidney biopsy tissues from patients with chronic kidney disease compared to healthy individuals (Fig. 5A). We examined the protein and transcript level of TFAP2B in HK-2 cells under the hypoxia-reoxygenation (H/R) injury. As shown in Fig. 5B and 5C, H/R injury significantly induced an increasing protein level of TFAP2B and S100A16 in HK-2 cells. Similarly, the transcript levels of TFAP2B and S100A16 were all upregulated in injured HK-2 cells compared to normal control (Fig. 5D). Using TGF-b1 treated cells, we also verified that TGF-b1 stimulation dramatically elevated the expression of TFAP2B and S100A16 at both protein levels (Fig. 5E) and mRNA levels (Fig. 5F). Consistently, TFAP2B protein expressions were higher in mice kidney tissues after UUO surgery detected by immunohistology staining (Fig. 5G), comparing with those in sham control subjects. These findings implied that TFAP2B regulates the transcription of S100A16 under renal injury.
6. TFAP2B transcriptionally regulates S100A16 expression in HK-2 cells
We then asked the question of whether TFAP2B directly regulates the expression of S100A16. In HK-2 cells, overexpression of TFAP2B by its plasmid transfection resulted in significantly increased levels of S100A16 proteins (Fig. 6A) and transcripts (Fig. 6B). These results suggest TFAP2B directly regulates the transcription of S100A16 under the renal injury condition.
To test whether TFAP2B is a transcription factor of S100A16, we worked to determine if TFAP2B binds to the promoter of S100A16. First, we predicted the binding sites via the JASPAR website (Supplementary Table 1.2). The position of S100A16 promoter at -200 ~ +88 bp (S1) and -1980 ~ -1734 bp (S2) were two potential binding regions with TFAP2B (Fig. 6C). ChIP assay revealed that TFAP2B is bound to S1 and S2 region of S100A16 promoter, and there is a stronger binding of TFAP2B with S1 than that with S2 (Fig. 6D). To test if TFAP2B binding these regions to transcriptional regulate S100A16, the S100A16 promoter sequence at S1 (-200 ~ +88 bp) or Full length (-2000 ~ 0 bp) were ligated to the luciferase reporter gene constructs (S1-Luc, and Full length-Luc), which were further transfected into HK-2 cells with overexpressed TFAP2B plasmid. The relative luciferase signals from each of the combinations in HK-2 cells were quantified. The data showed that TFAP2B increased the luciferase signal levels in HK-2 cells that were transfected with S100A16-Luc (Fig. 6E).
Together, these results suggest a TFAP2B-S100A16-HIF1a-HRD1 regulatory axis in renal tubular cells under the injured condition, and that TFAP2B directly regulates the transcription of S100A16.