Chromatin remodeling factor, INO80, inhibits PMAIP1 in renal tubular cells via exchange of histone variant H2A.Z. for H2A

Epigenetic modifications such as DNA methylation, histone modifications, and chromatin structures in the kidney contribute towards the progression of chronic kidney disease (CKD). In this study, the role of chromatin remodeling factor inositol requiring 80 (INO80) was investigated. Although INO80 regulates transcription by altering the chromatin structure at the nucleosome level, its role in the kidney remains unknown. We demonstrated that the expression of INO80 in impaired kidneys decreased in rats with unilateral urethral obstruction. We investigated INO80 expression in a proximal tubular cell line and observed that its expression decreased under hypoxic condition. Additionally, INO80 knockdown promoted apoptosis, suggesting that INO80 plays a role in inhibiting tubular cell apoptosis. We identified downstream target genes of INO80 via genome-wide analysis using RNA-sequences and found that the expression of apoptosis-related genes, such as TP53 and E2F1, and pro-apoptotic genes, such as PMAIP1, increased upon INO80 knockdown. ChIP-qPCR of the loci of PMAIP1 showed that the amount of H2A.Z. increased instead of decreasing the amount of H2A when INO80 was knocked down. These results indicated that INO80 plays a role in the exchange of H2A.Z. for H2A in the promoter region of PMAIP1 in tubular cells to inhibit apoptosis during CKD progression.

Inositol requiring 80 (INO80) is an ATP-dependent chromatin-remodeling factor that plays a role in transcription, DNA repair, and replication.In the eukaryotic nucleus, genomic DNA, the carrier of genetic information, wraps around histone octamers consisting of four core histones (H2A, H2B, H3, and H4) to form nucleosomes.The N-and C-terminal regions of histones are called histone tails and have been reported to undergo various post-translational modifications such as acetylation, methylation, phosphorylation, and mono-ubiquitination 1 .These modifications are thought to be involved in the epigenetic regulation of gene expression by altering chromatin structure.The highly aggregated chromatin structure inhibits the binding of regulatory factors such as basic transcription and replication factors; thus, genomic function is inactivated by chromatin structure.Therefore, genome function is controlled by a chromatin structural transformation, known as chromatin remodeling, which slides and removes nucleosomes.An important function of the chromatin-remodeling complex is to bind histones and work with histone chaperones to remove all or part of the core histones from the nucleosome.Thus, chromatin remodeling factors can positively and negatively control DNA reactions such as transcription, replication, repair, and recombination.
The chromatin-remodeling complex is a large complex composed of multiple subunits.It is classified into four families SWI/SNF (switch/sucrose-non-fermenting); ISWI (imitation switch); CHD (chromodomain-helicase-DNA binding); and INO80-according to the domain constitution type of the Snf2 family protein contained in the complex 2 .In yeast INO80, the histone variant H2A.Z. is genome-wide and controls its dynamics.INO80 has histone exchange activity and has been shown to replace the nucleosome H2A.Z./H2B with free H2A/H2B dimers 3,4 .The H2A.Z.histone variant is located on particular on nucleosomes located on both sides of the promoter region of a gene transcribed by RNA polymerase II; it is also localized on nucleosomes on both sides of the chromatin boundary elements, centromeres, and DNA replication origins 5,6 .These nucleosomes are characterized by rapid turnover independent of DNA replication.Nucleosome turnover is a dynamic process, thought INO80 expression decreased in impaired kidneys.We evaluated the expression level of INO80 in vivo in rats with impaired kidneys.Rats in the impaired kidney model showed unilateral urethral obstruction (UUO).Ureteral obstruction results in marked renal hemodynamic and metabolic changes, followed by tubular injury and cell death by apoptosis or necrosis with interstitial macrophage infiltration.The UUO model exhibits progressive renal fibrosis.Proliferation of interstitial fibroblasts with myofibroblast transformation leads to excess deposition of the extracellular matrix and renal fibrosis 13 .The mRNA level of INO80 extracted from the whole kidney in the UUO model rat group (n = 6) was significantly reduced compared to the contralateral kidney group (n = 6) (p value < 0.01) (Fig. 2).
Experiments were also performed on 5/6 Nx rats.The 5/6 Nx model is often used for studies of CKD.In 5/6 Nx rats, INO80 mRNA expression tended to decrease compared to that in sham rats; however, the difference was not significant (Supplementary Fig. S3A).The extent of interstitial fibrosis was determined by Masson's trichrome (MT) staining.MT staining showed that interstitial cells and collagen fibers increased in UUO mice.MT staining increased the number of interstitial cells and collagen fibers in 5/6 Nx rats (Supplementary Fig. S3B).Immunohistochemistry revealed that INO80 was mainly expressed in the nuclei of the tubules (Supplementary Fig. S3C).The tubule, especially its nucleus, was strongly stained, similar to UUO rats.Nuclear staining was more intense in sham rats than in the 5/6 Nx rats.We also performed immunohistochemistry on commercially available human kidney samples using an INO80 antibody (Supplementary Fig. S4).Details of the patients' backgrounds are shown in Supplementary Table S1.Normal kidney samples had more INO80 staining in the cytoplasm compared with CKD patient samples; however, statistical analysis was not possible because only two patients were available.The expression pattern of INO80 in immunohistochemistry was consistent with that of normal kidney samples, with INO80 expression in the tubules being higher than in CKD samples.
Genome-wide RNA-seq using HK2 cells knocked down with INO80 identified 32 down-regulated genes and three up-regulated genes.To identify the downstream target genes of INO80, we performed RNA-seq using siRNA of INO80 (N = 3).A principal component analysis (PCA) plot was shown in Fig. 4A.The volcano plot shows the degree of variation in the 35 genes (Fig. 4B).The fluctuation increases with distance from the center to the left and right (horizontal axis).Gene at the top of the figure show significant differences (vertical axis).As shown in the heatmap in Fig. 4C, we identified 32 down-regulated genes (blue) and three up-regulated genes (red; DENND6B, KIAA0391, SDK2) according to the criteria we set.These genes were suspected to be INO80 downstream target candidates.The expression profiles of these genes in the heatmap are shown in Table 1.As a result of gene ontology (GO) (Fig. 4D), apoptosis-related genes such as TP53 (tumor protein p53), E2F1 (E2F transcription factor 1), and PMAIP1 (phobol-12-myristate-13-acetate-induced protein 1, also known as NOXA) were significantly associated with INO80 functions.To validate the GO results, we examined the mRNA levels of apoptosis-related genes.As shown in Fig. 4E, the expression of TP53 mRNA was significantly increased by INO80 knockdown.It has been reported that INO80 suppresses the expression of the downstream gene of the transcription factor E2F in human umbilical venous endothelial cell (HUVEC) 14 , and E2F is involved in the apoptotic pathway, which is different from the p53 pathway.We also demonstrated that the expression of E2F1 mRNA significantly increased by INO80 knockdown (Fig. 4F).
INO80 exchanges H2A.Z. for H2A in the promoter region of PMAIP1.Among the apoptosisrelated genes, the expression of PMAIP1 mRNA significantly increased by INO80 knockdown (Fig. 5A).PMAIP1 is a member of pro-apoptotic subfamily of the BCL-2 protein family and is a known target of p53.PMAIP1 is a tumor suppressor gene candidate in the pancreatic cancer cell line 15 and induces apoptosis in lung cancer 16 .PMAIP1 induces apoptosis in several cell lines 17 .
To elucidate the role of INO80, we examined whether INO80 altered chromatin modifications in the promoter region of PMAIP1.We examined the role of INO80 in the promoter region because INO80 complex occupies the nucleosome free regions of the transcriptional start sites of over 90% in the budding yeast promoters 18 .Enrichment of INO80 at the promoter regions was correlated with transcriptional activity 19 .Chromatin immunoprecipitation (ChIP) was performed on HK-2 cells with INO80 knockdown using the H2A.Z.antibody, because INO80 has been reported to remove the H2A.Z.variant and exchange it for the H2A variant 20,21 .When qPCR was performed on collected DNA samples using primers for the promoter regions of PMAIP1, INO80 knockdown significantly increased H2A.Z. in the promoter region of PMAIP1 (Fig. 5B).In contrast, H2A expression significantly decreased in the promoter region of PMAIP1 (Fig. 5C).We also confirmed that the H2A.Z. in the promoter region of TP53 also significantly increased and H2A decreased when INO80 is knocked down (Fig. S6A).INO80 knockdown significantly increased H2A.Z. in the promoter region of E2F1 while H2A has a tendency of decrease (p value = 0.07) (Fig. S6B).These results suggested that INO80 removed H2A.Z. and exchange it for H2A in the promoter region of apoptosis-related genes, resulting in the inhibition of apoptosis.

Discussion
Here, we demonstrated the functional roles of INO80 in renal tubular cells.The novel findings of this study are as follows: (1) INO80 decreases under hypoxia and kidney injury; (2) INO80 protects renal tubular cells from apoptosis; (3) INO80 has histone-exchange activity that can replace nucleosomal H2A.Z./H2B with H2A/H2B in the promoter region of PMAIP1, resulting in increased apoptosis of tubular cells via the accumulation of H2A.Z. and the reduction of H2A levels in injured kidney cells.
Chromatin remodeling complexes are evolutionarily conserved from Saccharomyces cerevisiae to humans.Several chromatin remodeling complexes contain actin and actin-related protein as essential components for their function 22 .Chromatin remodeling complexes are transported to specific chromatin regions by histone variants, histone modifications, or interactions with transcription factors.Thereafter, the removal of nucleosomes, conversion of positioning, exchange of histone variants in nucleosomes, and changes in the histone modification state are performed in an ATP-dependent manner to change the chromatin structure at the nucleosome level.H2A.Z. is a H2A histone variant that is conserved from yeast to humans, suggesting that H2A.Z. has important biological roles 23 .Previous studies have shown that H2A.Z.contributes to epigenetic regulation in various species 24,25 .For example, H2A.Z. are located in the promoter regions of genes and are involved in chromosomal distribution and DNA damage repair.In our model, H2A.Z.accumulated and H2A was reduced when INO80 was knocked down in HK-2 cells in the promoter regions of PMAIP1.In Arabidopsis, the INO80 chromatin complex has been reported to promote thermomorphogenesis via H2A.Z.eviction and transcription 26 .In yeast, INO80 is recruited to specific promoters required for H2A.Z.exchange and it promotes hyphal development of Candida albicans 27 .Previous studies indicated that INO80 removes H2A.Z. and regulates its downstream target genes.In addition, INO80 was demonstrated to translocates along the DNA at the H2A-H2B interface of nucleosomes and persistently displace DNA from the surface of H2A-H2B 28 .INO80 has a dimeric exchange mechanism, H2A.Z. for H2A in an in vitro assay.Several studies on INO80 and apoptosis have been conducted in oncology field.For example, in human colon cancer cells, INO80 knockdown increases apoptosis 29 and in cervical cancer cells, INO80 knockdown does not influence apoptosis 30 .We hypothesized that INO80 may have different mechanisms depending on the cell type.In our experimental design, INO80 inhibited apoptosis of renal tubular cells during the progression of CKD.Although INO80 has been reported to be associated with renal function in a genome-wide analysis 9 , the mechanism by which INO80 works in the kidney and affects renal function has not been elucidated.To the best of our knowledge, this is the first study to show that INO80 inhibits tubular cell apoptosis by exchanging H2A.Z.

FPKM
PMAIP1, a pro-apoptotic protein, is a member of the B-cell lymphoma 2 (Bcl-2) family and belongs to the Bcl-2 homology 3 (BH3) only protein subclass.Induction of PMAIP1 occurs via both p53-dependent and p53-independent mechanisms.P53-independent PMAIP1 regulators include hypoxic stimulation and the transcription factor E2F1.In addition to the p53 binding sites, the PMAIP1 promoter contains a Hypoxia-Responsive Element (HRE) and an E2F1 binding site.PMAIP1 was reported as a direct target of HIF1 and mediates hypoxic cell death in a p53-independent manner 31 .In our study, PMAIP1 expression was found to be dynamically regulated by histone exchange via chromatin levels.
The current research has several limitations.Although we identified 35 INO80 down-stream target genes, the functions of the remaining down-stream target genes, except PMAIP1 remain unknown.There is a possibility that more down-stream target candidates of INO80 were identified when we set different criteria in the analysis of RNA-seq.Further analysis of INO80 functions in chromatin remodeling will be needed.

Conclusion
We found that INO80 levels decreased in patients with CKD.In normal kidney tissue, INO80 removes H2A.Z. and exchanges it for H2A on the promoter region of PMAIP1, suppressing PMAIP1, and inhibiting tubular cell apoptosis.However, during the progression of CKD, tubulointerstitial hypoxia decreases the expression of INO80, resulting in an increase in H2A.Z., which increases PMAIP1, leading to the promotion of tubular cell apoptosis (Fig. 6).Western blotting.Whole kidneys and HK-2 cells were lysed using RIPA buffer (50 mM tris-HCl (pH 8.0), 150 mM sodium chloride, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate and 1.0% Triton X-100) with a protease inhibitor mixture (cOmplete ® Mini, Sigma) on ice.Whole-kidney samples were homogenized using a homogenizer (PHYSCOTRON NS-310EIII, MICROTEC Co. Ltd.).Both the whole kidney and cell samples were centrifuged at 6000 rpm for 3 min at 4 °C.The supernatants were collected as the total protein content.Protein concentrations were measured using the DC™ Protein Assay Kit (Bio-Rad Laboratories Inc.).The extracted protein solutions were denatured with incubation in sample buffer (60 mM Tris (pH 6.8), 2% sodium dodecyl sulfate, 10% glycerol, 10 mM dithiothreitol and 0.01% bromophenol blue) for 5 min at 95 °C.The lysates were separated with 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions.After electrophoresis, the protein samples were transferred onto polyvinylidene difluoride membranes (Millipore).Nonspecific protein binding was blocked with 5% skim milk in Tris-buffered saline (pH 7.4), containing 0.05% Tween 20 (TBST).The membranes were subjected to immunoblotting using polyclonal rabbit anti-INO80 (1/2000 dilution; Abcam ab118787, Cambridge, UK) and polyclonal rabbit anti-β-actin (1/2000 dilution; Bio-Rad Laboratories Inc.) overnight at 4 °C.The membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG antibody (1/10,000 dilution; Bio-Rad Laboratories Inc.) for 30 min at room temperature.The bands were developed with Pierce™ ECL Plus Western Blotting Substrate (Thermo Fisher Scientific) and detected with Image Quant LAS 4000 (GE Healthcare, Tokyo, Japan).The bands were normalized by β-actin.Quantitatively, we analyzed the average of 20 fields of image where the fluorescence signal intensity was divided by the number of nuclei and averaged using Image J.
Immunohistochemistry. The formalin-fixed and paraffin-embedded tissues were prepared with 3 μm thick sections.The sections were dewaxed with Histoclear ® (National Diagnostics) and rehydrated using a graded series of ethanol.After washing with phosphate buffered saline (PBS), tissue slides were subjected for antigen retrieval by boiling in 10 mM citrate buffer (pH 6.0) for 10 min in a microwave.Endogenous peroxidase activity was quenched with 3% H 2 O 2 in PBS for 15 min at 24 °C.Nonspecific antibody binding was blocked via incubation with Protein Block Serum-Free (Dako, Agilent Technologies) for 15 min at 24 °C and subsequently incubated overnight at 4 °C with primary antibody, polyclonal rabbit anti-INO80 (1/400 dilution; LSBio LS-B11393) (1/400 dilution; Abcam ab118787).Thereafter, the sections were incubated with a biotinylated, affinity-purified anti-immunoglobulin secondary antibody (PK-6101 VECTASTAIN Elite ABC Rabbit IgG Kit; Vector Laboratories Inc., Burlingame, CA, USA), for 40 min at 24 °C.The color was developed using ImmPACT™ DAB (Vector Laboratories Inc.).Finally, the samples were dehydrated and mounted on coverslips.A BX51 microscope (OLYMPUS, Tokyo, Japan) was used for observation, and the slides were photographed using a camera DP20-5 (OLYMPUS).
Commercially available human kidney samples were obtained from ProteoGenex (CA, USA), and an outline of the samples is shown in Supplementary Table S1.and incubated at 37 °C for 24 h.Thereafter, they were fixed with methanol and acetone 1:1 for 30 min on ice.After washing with PBS, cells were incubated with 0.3% Tween 20 in PBS for 30 min at 24 °C.Non-specific antibody binding was blocked via incubation with 5% bovine serum albumin, diluted with PBS for 30 min, and Protein Block Serum-Free (Dako by agent Technologies) for 10 min at 24 °C.They were incubated for 1 h at 24 °C with the primary polyclonal rabbit anti-INO80 (1/400 dilution; Abcam ab118787).After washing with 0.1% Tween 20, the membranes were incubated with a secondary polyclonal swine anti-rabbit immunoglobulins/FITC (1/20 dilution; Dako Agent Technologies) for 1 h at 24 °C in dark.Finally, the cells were loaded with Hoechst stain (1/2000 dilution; bisbenzimide H33342 trihydrochloride, Sigma Aldrich, St. Louis, MO, USA) for 2.5 min for nuclear staining, followed by coverslip mounting.Fluorescence microscope BZ-X710 (KEYENCE, Osaka, Japan) was used for observation.For quantitative analysis, 20 fields of view were photographed with the 1000 × lens of a fluorescence microscope, and the fluorescence signal intensity in each nucleus was measured using ImageJ software 32 .
Functional assay.Caspase 3/7 assay.We measured caspase 3/7 activity using the Caspase-Glo ® 3/7 Assay kit (Promega, Madison, WI, USA), according to the manufacturer's protocol.This is a homogeneous, luminescent assay that measures caspase 3 and caspase 7 activities.Luminescence is proportional to caspase activity.First, HK-2 cells were transfected with siRNA and reseeded in 96-well plates at 100,000 cells per well in 100 μL of FBS-free medium.Cells were cultured under normoxic or hypoxic conditions for 48 h.Thereafter, we added the 100 μL of caspase-glo reagent, which is a mixture of caspase-glo substrate and caspase-glo buffer.Luminescence was measured after 30 min of incubation using luminometer (Multimode Plate Reader EnSpire ® , PerkinElmer, Inc., MA, USA).
MTS assay.We evaluated cell viability using CellTiter ® 96 AQueous One Solution Cell Proliferation Assay (Promega), according to the manufacturer's protocol.The assays were performed by adding the MTS reagent directly to culture wells, and the quantity of formazan product, as measured by the absorbance at 490 nm, was directly proportional to the number of living cells in the culture.First, HK-2 cells were transfected with siRNA and reseeded in 96-well plates at 100,000 cells per well in 100 μL of FBS-free medium.Cells were cultured under normoxic or hypoxic conditions for 24 h.After adding 20 μL of MTS reagent and incubating at 37 °C for 1 h, absorbance at 490 nm was measured using a plate reader (Multimode Plate Reader EnSpire ® , PerkinElmer, Inc).
RNA sequencing.mRNA was isolated as described above.RNA-seq libraries were prepared and sequenced using the HiSeq platform (Illumina, San Diego, CA) according to the manufacturer's protocol.The reads per kilobase of exons per million mapped reads for each gene were calculated based on the length of the gene and read counts mapped to the gene.The sequences were aligned with human reference genome (UCSC hg38) using ELAND (Illumina) 33 .Details were described in our previous paper 34,35 .GENEWIZ (South Plainfield, NJ, USA) was used to analyze mRNA samples extracted from HK-2 cells with INO80 knockdown.RNA samples were prepared using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions.

Statistical analyses.
All measurements were expressed as the mean ± SEM.Data between groups were compared using a Student's two-tailed t test.Statistical significance was set at p < 0.05.All analyses were performed using Microsoft Excel and GraphPad Prism 9.

Figure 1 .
Figure 1.INO80 decreased under hypoxia in HK-2 cells.(A) Quantitative PCR analysis of INO80 mRNA expression under normoxia, hypoxia and anoxia for 24 h in HK-2 cells.Data are shown as the mean and the standard deviation from triplicates.*p < 0.05.(B) Western blotting of INO80 protein and β-actin extracted from HK-2 cells under normoxia, 1% hypoxia and anoxia for 24 h.The cropped blots were used in the figure and different molecular weight were separated by lines.Full-length blots are presented in Supplementary Figure S2A,B.(C) Quantitative analysis of western blot of INO80 under normoxia, 24 h hypoxia and 24 h anoxia.The numbers in the data are the INO80 bands normalized with β-actin.The data uses HK-2 cell samples collected on three different days.**p < 0.01.(D) Immunocytochemistry image of INO80 in HK-2 cells under normoxia, 24 h hypoxia, and 24 h anoxia.(E) In the immunocytochemistry image of INO80, the fluorescence signal intensity in each nucleus was measured by ImageJ software (N = 20 for each group).*p < 0.05, **p < 0.01.

Figure 3 .
Figure 3. Functional assay of INO80 in HK-2 cells.(A) Quantitative PCR analysis of INO80 mRNA expression in transfected HK-2 cells.(B) Full membrane of western blotting of INO80 when INO80 was knocked down using siRNA.The membrane was once cut when INO80 antibody and β-actin antibody were exposed.After the exposure of second antibodies, the separate membranes were lined up next to each other.Ctl means Control siRNA.(C) Immunocytochemistry image of INO80 in HK-2 cells when INO80 was knocked down using siRNA.The values in the western-blotting data show the INO80 bands normalized with β-actin.***p < 0.001.(D) The apoptotic rate of HK-2 cells was determined by caspase activation using Caspase 3/7 Glo assay after transfected with negative control or INO80 siRNA for 48 h normoxia or hypoxia.*p < 0.05.(E) HK-2 cells viability measured by MTS assay after transfected with negative control or INO80 siRNA for 24 h normoxia or hypoxia.*p < 0.05, **p < 0.01.

Figure 4 .
Figure 4. Results of RNA-seq of INO knockdown by siRNA.(A) Principal Component Analysis chart.Control siRNA (N = 3) and siRNA of INO80 (N = 3) were analyzed as RNA-seq samples.(B) Differential expression volcano plot of each sample.Blue dots were significant down-regulation by siRNA of INO80.Only three red plots were upregulated by INO80 siRNA.X-axis; log 2 folds change of gene expression.Y-axis; statistical significance of differential expression in log 10. (C) Cluster analysis of differentially expressed genes log 10 (FPKM + 1) values are used for clustering.Genes of high expressed are in red, and those with low expression are shown in blue.Thirty-five genes were significantly up-regulated (three genes) and down-regulated (thirty-two genes) by INO80 siRNA.The names of these genes have listed.Their expression levels were shown as heatmaps.(D) GO enrichment p value histogram.X axis: log 10 (p value) of each term.Y axis represents significantly enriched GO terms.The highest p value was listed in the top of the graphs in Gene Ontology.(E) Quantitative PCR analysis of TP53 mRNA expression in INO80 knockdown cells.(F) Quantitative PCR analysis of E2F1 mRNA expression in INO80 knockdown cells.

Figure 5 .
Figure 5. Histone exchange on the promoter region of PMAIP1 by INO80 knockdown.(A) Quantitative PCR analysis of PMAIP1 mRNA expression in INO80 knockdown cells.(B) Quantitative ChIP-qPCR analysis of PMAIP1 promoter using an H2A.Z.antibody upon INO80 knockdown.(C) Quantitative ChIP-qPCR analysis of PMAIP1 promoter using an H2A antibody upon INO80 knockdown.

Table 1 .
The expression profiles of downstream target gene candidates of INO80 in the heatmap.