PON1 hypermethylation is associated with progression of renal cell carcinoma

Abstract In this study, our aim was to exploring the influences of DNA methylation of PON1 on cell proliferation, migration and apoptosis of renal cancer cells. The genome‐wide methylation array of renal cell carcinoma samples and adjacent tissues were obtained from the cancer genome atlas (TCGA) database. By analysing the DNA methylation and conducting the CpG islands array, methylation status expressed in renal tumour samples and normal renal tissue samples were detected. Methylation‐specific PCR (MS‐PCR) and qRT‐PCR were employed to detect the methylation level and mRNA expression of PON1. Wound‐healing assay, transwell assay and MTT assay were utilized to detecting the migration, invasion and proliferation abilities, respectively. The cell apoptosis was testified by Tunnel assay. In addition, the effect of PON1 on renal cancer cells was verified by experiments in vivo. The methylation status of different genes in renal cell carcinoma samples was obtained by CpG islands arrays and hypermethylated PON1 was selected for further study. PON1 was down‐regulated in renal cell carcinoma tissues detected by qRT‐PCR and Western blot. Both in vitro and vivo experiments indicated that the sunitinib‐resistant in renal cancer cells could be suppressed by treat with 5‐Aza‐dC or TSA, and the effect came out more obvious after 5‐Aza‐dC and TSA co‐treatment. In detail, the demethylation of PON1 inhibited the migration, invasion and proliferation of renal cancer cells and also arrested more cells in G0/G1 phase. The vivo experiment indicated that demethylated PON1 suppressed the growth of tumour. Hypermethylated PON1 promoted migration, invasion and proliferation of sunitinib‐resistance renal cancer cells and arrested more cells in G0/G1 phase.

disease. In current study, we tried to revealing the internal mechanism between DNA methylation and RCC.
DNA methylation was heritable and reversible, belonged to epigenetic changes, which represent in the interface between the genome and the environment. 4 In the procession of DNA methylation, cytosine was covalently modified by adding a methyl group to their backbone, forming a 5-methylcytosine nucleotide (5-mC). 5 In addition to play an important epigenetic maker in gene silencing, DNA methylation was also involved in regulating normal growth and developmental processes such as cell differentiation, genomic imprinting and suppression of repetitive elements. [6][7][8][9] There have been some studies about DNA methylation and renal cancers. Malouf et al defined the epigenetic basis for proximal versus distal tubule-derived kidney tumours. 10 Besides, methylation-associated genes such as SETD2, KRT19 and SFRP1 have been studied in renal cancers. [11][12][13] However, the mechanism of DNA methylation in RCC remained plenty of unclear regions.
Human serum paraoxonase-1 (PON1) is a Ca 2+ dependent high-density lipoprotein (HDL) associated lactonase capable of hydrolysing a wide variety of lactones, thiolactones, arylesters and cyclic carbonates. 14 PON1 is a kind of glycoprotein, which composed of 354 amino acids and approximate molecular mass of 43 KDa, and it retains its hydrophobic signal sequence in the N-terminal region which enables its association with HDL. 15 PON1 encodes a member of the paraoxonase family of enzymes and exhibits lactonase and ester hydrolase activity. In regard to the methylation of PON1, hypomethylated CpGs in the promoter of PON1 was predicted to be an underlying risk of bleeding after dual antiplatelet therapy. 16 Moreover, recent investigations indicated that PON1 had considerable effect on molecular disorders connected with cancer. [17][18][19] Some researchers found that measurement of serum PON1 concentration post-radiotherapy could be an efficient prognostic biomarker and an index of the efficacy of the radiotherapy. 20 And in this paper, we devoted to explaining the mechanism of PON1 in KIRP.
5-aza-2′-deoxycytidine (5-Aza-dC) was used in many DNA methylation studies. For example, Gao et al used 5-Aza-dC to investigating the influences of hypermethylated MEG3 in retinoblastoma. 21 Yan et al also used 5-Aza-dC, which was down-regulated the DNA methylation of SPARC, to study the progression of T-cell lymphoma. 22 As a kind of histone deacetylase (HDAC) inhibitors, trichostatin A (TSA) could retard the growth of carcinomas of cervix, colon, rectum and other cancers in vitro. 23 What's more, the co-treatment of 5-Aza-dC and TSA showed a better effect in human gastric cancer cells. 24 In addition, sunitinib, a multitargeted tyrosine kinase inhibitor (TKI), which currently was the standard of care for patients suffering from metastatic renal cell cancer. 25 On the whole, we designed to exploring the impact of 5-Aza-dC and TSA co-treatment in sunitinib-resistant RCC cell.
To sum up, the aim of our study was to find the target gene affecting the progression of KIRP and to study its mechanism.
Furthermore, we also testified the combination therapy in sunitinibresistant RCC to find a better treatment.

| Clinical samples
We obtained 15 pairs of RCC and corresponding para-carcinoma tissues randomly from patients undergoing surgical treatment from May 2016 to June 2017 at China-Japan Union Hospital of Jilin University. We got the approval from the Ethics Committee of China-Japan Union Hospital of Jilin University to collect the samples, and we obtained the informed consent from all the patients. We took the tissues from the patients and stored them in a liquid nitrogen freezer at −80°C, preparing for subsequent experiments. were grown in a 1:1 mixture of DMEM/Ham's F 12 nutrient medium (F12). All media needed to add 10% foetal bovine serum (FBS) (GIBCO Invitrogen). Culture plates were placed at the condition of 37°C and 5% CO 2 in an incubator.

| Genome-wide methylation analysis
The genome-wide methylation array of renal cell carcinoma samples and adjacent tissues were obtained from The Cancer Genome Atlas (TCGA) database to perform the unsupervised hierarchical clustering. The ChAMP R package (http://www.bioco nduct or.org/packa ges/devel/ bioc/html/ChAMP.html) was employed for methylation analysis, which contained limma-based differential methylation protein (DMP) and Probe Lasso-based differential methylation region (DMR) analysis function. By using the Illumina Infinium Human

| Methylation-specific polymerase chain reaction (MS-PCR)
To perform MS-PCR, a DNeasy Tissue Kit (Qiagen) was used to extract DNA from the tissue samples following the protocol described  Table S1 showed the primers used.

| Treatment of RCC cells with 5-Aza-dC, TSA and sunitinib
The human RCC cells was cultured in 6-cm dishes and incubated over-   Table S2. The results were recorded after the cycle, and the relative gene expression was analysed by 2 −ΔΔC t method.

| Western blot
The total protein was separated by RIPA lysate (Thermo Fisher Scientific). Then use BCA Kit (Sigma-Aldrich) to quantify the protein concentration. Separate the proteins by SDS-polyacrylamide gelelectrophoresis (SDS-PAGE) and transfer them to a polyvinylidene fluoride (PVDF) membrane (Ruiqi). Then block the membrane with 5% (w/v) bovine serum albumin in Tris-buffered saline (TBS) buffer at room temperature for 45 minutes and incubated first antibody anti-PON1 (ab24261, Abcam) and anti-GAPDH (ab8245, Abcam) at 4°C overnight. GAPDH was used for normalization. Wash the PVDF membrane three times by TBS and incubate them with the peroxidase-conjugated mouse anti-goat IgG antibody (Abcam) for 4 hours at room temperature. Immunologically active proteins were visualized with an enhanced chemiluminescence system and analyse the results by using ImageJ software.

| Wound-healing assay
1 × 10 6 SKRC39 cells were seeded in 6-well plates. Then use a pipette tip to draw a straight scratch on the bottom of the plates. Wash the suspension cells twice or three times gently by PBS and take images of the scratch as baseline by the microscope. After that, the medium was replaced to the serum-free RPMI-1640 medium. Add different drugs into each group and take pictures of the same location again after the cells were cultured for 24 hours.

| Tunnel assay
SKRC39 cells treated with four methods (sunitinib group (control group), 5-Aza-dC group, TSA group and 5-Aza-dC + TSA + sunitinib group) were stably transfected, dealt with hunger and washed twice in PBS (Thermo Fisher Scientific) at 4°C, and re-suspended in 250 μL labelling buffer (Haoranbio). Cells from each group were stained with 5 μL annexin V/FITC and 10 μL of 20 μg/mL propidium iodine (Sigma-Aldrich) and were incubated for 15 minutes at 37°C in the dark. The results can be observed by fluorescence microscope (ZEISS).

| Tumour xenograft
Five-week nude mice at the same condition were arranged as seven groups randomly, with five in each group, including normal group, sunitinib group, 5-Aza-dC group, TSA group, 5-Aza-dC + sunitinib group, TSA + sunitinib group and 5-Aza-dC + TSA + sunitinib group.

| Immunohistochemistry
Deparaffinize the tissue samples with xylene and then rehydrate the sections (5 mm) from tissues with ethanol. After that, to block endogenous peroxidase activity, we should immerse them in 3%

| Statistical analysis
We used SPSS standard version 19.0 (SPSS Inc) and GraphPad Prism 6.0 to analyse all results which came from the averages of at least three independent experiments. Statistical analyses were performed using a paired sample t-test and one-way analysis of variance (ANOVA). Data from all quantitative assays were shown as the mean ± standard and P < 0.05 was considered to indicate a statistically significant difference.

| PON1 presented hypermethylated and lower mRNA expressed in RCC
As shown in Figure S1, the hierarchical clustering analysis screened the top 1000 differential CpG islands from 395 412 probes. In addition, Figure S2 Figure 1A). The relative heatmap showed the methylation data as a standard control (from low to high methylated) in a blue-red scale (from low to high methylation level). Analogously, differential mRNA analysis of paired tumour/normal tissues identified the top 50 differentially mRNA-expression-level genes in a green-red scale (from low to high mRNA expression level) ( Figure 2B). The above two heat maps revealed that PON1 presented high methylated and low mRNA expressed in the RCC tissues.

| Methylation status of PON1
The methylation status of 660 genes were detected by CpGs island analysis which revealed that PON1 was hypermethylated among them (Figure 2A). The region of DMR_431 which was one of the differentially methylated regions of PON1 showed that most CpG islands on the PON1 gene were in high methylation level ( Figure 2B). Among the 9 differentially methylated imprinted sites including cg01874867, cg04155289, cg05342682, cg07404485, cg17330251, cg19678392, cg24062571, cg22798737 and cg21856205, these box plots revealed that the methylation level was high in the RCC cells than in the normal cells except cg 24062571 and cg 22798737 ( Figure S3). Kaplan-Meier analysis showed that hypermethylated PON1 had shorter lifespan generally. Calculate the P value by the log-rank test ( Figure 2C). Taken together, the above results showed that the high DNA methylation level of PON1 in RCC tissues.

| 5-Aza-dC and TSA co-treatment in SKRC39/ sunitinib cells overexpressed PON1
MTT assay was employed to evaluate the different treatments on RCC and determine the concentration of reagents in the experiment. The results showed that most suitable dose of 5-Aza-dC and TSA were 1 μmol/L and 100 nmol/L respectively, and the concentration of 1 μg/mL sunitinib was selected for further studies ( Figure 4A-C). Based on the results of MSP and qRT-PCR, the methylation and mRNA expression of PON1 were verified and the results showed that the treatments of 1 μmol/L 5-Aza-dC and 100 nmol/L TSA in SKRC39/sunitinib cells had relative lower methylation level and higher expression level compare with the other two RCC cells ( Figure 4D-E). Collectively, these data indicated that after treated with 5-Aza-dC and TSA, the methylation level was decreased and expression level was increased in RCC cells. Besides that, re-expression of PON1 in co-treatment 5-Aza-dC + TSA group showed a significant increase in SKRC39/sunitinib cells. In other words, the co-treatment of 5-Aza-dC + TSA on SKRC39/sunitinib cells make PON1 de-methylated and re-expressed significantly.

| Impact of PON1 in cell migration, invasion, proliferation and cell cycle in vitro
To exploring the role and mechanism of hypermethylated PON1 in RCC cells, the migration, invasion, cell cycle and cell proliferation in different groups of SKRC39/sunitinib cells were analysed.
The experiment groups were divided into four groups, including sunitinib group, 5-Aza-dC + sunitinib group, TSA + sunitinib group and 5-Aza-dC + TSA + sunitinib group. The cell migration abilities were detected by wound-healing assay, and the results revealed that the migration distance in both 5-Aza-dC and TSA groups was inhibited better than that in the control group, but the group of 5-Aza-dC + TSA in SKRC39/sunitinib cells showed shorter migration distance ( Figure 5A,B). Additionally, the transwell assay was per- Cell cycle was detected by flow cytometry assay, and the results revealed that treated with 5-Aza-dC/TSA could arrested more cells in the G0/G1 phase and co-treatment 5-Aza-dC + TSA group displayed more obvious trend ( Figure 5D,F). And then, cell proliferation was F I G U R E 3 PON1 was hypermethylated in renal tumour cells. A, PON1 was hypermethylated in tumour tissues using methylation-specific PCR. 'Pos' represented positive control; 'Neg' represented negative control. M: methylated; U: unmethylated. B, PON1 was confirmed to be hypermethylated in 786-O, Caki-2, and SKRC39 cell lines compared with HK-2cells. The change of DNA methylation level was maximal in SKRC39 cell line. ** P < 0.01, *** P < 0.001, compared with the HK-2 cell lines. C, The mRNA levels of PON1 in tumour cells and normal cells were analysed by real-time PCR. ** P < 0.01, *** P < 0.001, compared with the HK-2 cell lines. Each data represented mean value ± standard deviation (SD). D-E, The protein levels of PON1 in tumour cells and normal cells were determined using Western blotting. ** P < 0.01, compared with the HK-2 cell lines verified by MTT assay, proved that the cell proliferation curve of 5-Aza-dC + TSA in SKRC39/sunitinib cells was inhibited ( Figure 5G).
In addition, cell apoptosis was determined by Tunnel assay and the results confirmed that 5-Aza-dC and TSA could promote cell apoptosis of RCC cells ( Figure 5H). These in vitro experiments above showed that re-expression of PON1 could restrain the migration, invasion and proliferation abilities and promote cell apoptosis in RCC cells.

| 5-Aza-dC + TSA + sunitinib co-treatment group restrained RCC tumour growth in vivo
To verify the curative effect of different treatment in RCC, tumour xenograft experiment and immunohistochemistry were employed for in vivo detection. The results were showed in Figure 6A-C, and the tumour size and weight were suppressed in 5-Aza-dC group and TSA group. The suppression effect of co-treatment 5-Aza-dC + TSA was inhibited more obvious. Immunochemical staining was conducted to verify that the index of ki67 proteins was down-regulate significantly in the co-treatment group of ( Figure 6D-E). These suggested that the co-treatment group was a promising strategy for RCC therapy.

| D ISCUSS I ON
In this research, we demonstrated hypermethylated PON1 affected on-cogenesis of RCC. Through vivo and vitro experiments, the results F I G U R E 4 5-Aza-dC and TSA co-treatment induced demethylation and re-expression of PON1. A, Minimum effective dose of 5-Aza-dC was determined by MTT and 1 μmol/L showed difference. * P < 0.05, ** P < 0.01, compared with 0 μmol/L group. B, Minimum effective dose of TSA was determined by MTT and 100 nmol/L showed difference. * P < 0.05, ** P < 0.01, compared with the 0 nmol/L group. C, Minimum effective dose of sunitinib was determined by MTT and 1 μg/mL showed difference. * P < 0.05, ** P < 0.01, compared with the 0 μg/mL group. D, 5-Aza-dC and TSA could decrease the methylation level of PON1 compared with the sunitinib group and co-treatment group had more obvious trend. ** P < 0.01 compared with the sunitinib group. E, 5-Aza-dC and TSA could increase the expression level of PON1 compared with the sunitinib group and co-treatment group had more obvious trend. ** P < 0.01 compared with the sunitinib group. Each data represented mean value ± standard deviation (SD) F I G U R E 5 PON1 inhibited cell migration, invasion and proliferation, and arrested more cell at the G0/G1 phase. A,B, The wound-healing assay showed that the migration distance of 5-Aza-dC treatment and TSA treatment were inhibited and the co-treatment group displayed more obvious trend. * P < 0.05, ** P < 0.01, compared with the sunitinib group. C&E, The number of invasive cells in 5-Aza-dC treatment and TSA treatment were decreased and co-treatment group displayed more obvious trend. * P < 0.05, ** P < 0.01, compared with the sunitinib group. D&F, More cells were arrested at the G0/G1 phase in 5-Aza-dC treatment group and TSA treatment group and co-treatment group displayed more obvious trend. * P < 0.05, ** P < 0.01, compared with the sunitinib group. G, Cell proliferation was inhibited in 5-Aza-dC treatment group and TSA treatment group and co-treatment group displayed more obvious trend. * P < 0.05, ** P < 0.01, compared with the sunitinib group. H, The apoptosis cells were increased in 5-Aza-dC treatment group and TSA treatment group and co-treatment group displayed more obvious trend (×40) | 6655

LI and YU
showed that 5-Aza-dC and TSA could better block the tumour growth.
The co-treatment also made PON1 lower methylation level and higher mRNA expression at the same time. Collectively, 5-Aza-dC and TSA could synergize the effect of inhibition of sunitinib-resistant RCC tumour growth by inducing PON1 re-expressed.
Nowadays, DNA methylation was a research hotspot about cancer including renal cancers. The heterogeneity of DNA methylation was observed by Dugué et al, with stronger associations for risk of kidney cancer. 26 Additionally, Kumar et al put forward the hypothesis that IQGAP2 and IQGAP3 were promised to be prognosis therapeutic target of specific cancers including renal cancer for the close connection of their methylation and cancers. 27 It also reports that RPS6KA4/MIR1237 and AURKC promoter regions are differentially methylated in Wilms' tumour. 28 Analogously, we found that the methylation of PON1 was associated with the procession of RCC which is also the base of our research.
As forPON1 and DNA methylation, Fiorito et al found a general inverse relationship between B-vitamins intake and DNA methylation of PON1. 29 The methylation of PON1 was also linked to vascular dementia which confirmed by Bednarska-Makaruk et al. 30 Interestingly, the experimental group suffered from the disease also had lower intake of vitamin B than the normal group.
Although the relative studies about PON1 and DNA methylation are not so much and it has been a forward-looking subject due to the non-negligible role of PON1 played in the development of cancers including RCC.
Furthermore, the purpose of combination therapy in tumour treatment was to enhance the curative effect and reduce the occurrence of adverse reactions. According to the mechanism of anti-tumour drugs and tumour cell proliferation kinetics, carrying out reasonable combination of drugs has been a hot area of cancer treatment in recent years. For instance, a study found that the inhibition of CCDC69 was hypermethylated in ovarian cancer cells and might interfere with the effectiveness of combination therapy with platinum drugs, which means the combination of 5-Aza-dC and other anti-cancer drugs may have better effect. 31 To overcome drug resistance, some researchers explored a promising approach using 5-Aza-dC and the mTOR inhibitor everolimus in Medullary thyroid cancer cells, which showed a strong synergistic antiproliferative activity through the induction of apoptosis. 32 Certainly, the studies of co-treatment in renal cancer are not rare.
The researchers studied how 5-aza-dc and paclitaxel (PTX) synergized against renal cell carcinoma (RCC). 33 In our study, the drug combination of 5-Aza-dC and TSA combined with sunitinib also showed a better effect of suppressing tumorigenesis of RCC from the experimental results. Certainly, the clinical trials are needed to verify the effect.
Hypermethylation of PON1 was present in the progression of RCC. Nonetheless, there was insufficient in the vitro experiments as we only used one RCC cell line among the three renal cell lines we researched. However, we believe this trend will be similar and further investigation will be conducted in the future. Overall, our F I G U R E 6 5-Aza-dC + TSA + sunitinib co-treatment group suppressed tumour growth in vivo. A-C, Tumour growth was inhibited in 5-Aza-dC treatment group and TSA treatment group and co-treatment group displayed more obvious trend. * P < 0.05, ** P < 0.01, compared with the sunitinib group. D-E, The concentration of ki67 was decreased in 5-Aza-dC treatment group and TSA treatment group and cotreatment group displayed more obvious trend. * P < 0.05, ** P < 0.01, compared with the sunitinib group research showed that the DNA methylation of PON1, which is hopeful to be a targeted biomarker in RCC, could influence the development of RCC.

| CON CLUS ION
In short, this is the first study on the hypermethylated PON1 involved in the development of RCC, and it showed DNA methylation was one of the key parts in the proliferation of renal cancer cells especially the RCC cells. What's more, we have clarified the mechanism of how the DNA methylation of PON1, which may be a promising target for gene treatment, involved in the tumorigenesis of RCC. In addition, 5-Aza-dC and TSA co-treatment for sunitinib-resistant RCC is hopeful clinical therapy.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
XL and QY contributed to research design and the acquisition; XL analysis and interpretation of data; QY drafting the paper and revising it critically; XL and QY approval of the submission and final versions.

ACK N OWLED G EM ENT
We'd like to express our deep appreciation to Science and Technology Development Plan Project of Jilin Provincial Science and Technology Department (No.: 20160101032JC), which offered great help in our research process.

E TH I C A L A PPROVA L
The research was carried out according to the World Medical Association Declaration of Helsinki. Written informed consents were obtained from all the participants. This study approved by the Ethics Committee of China-Japan Union Hospital of Jilin University.

DATA AVA I L A B I L I T Y
The data that support the findings of this study are available from the corresponding author upon reasonable request.