SUV39H1-mediated DNMT3A is participated in the epigenetic regulation of Tim-3 and galectin-9 in cervical cancer

Background: Methylation of histone 3 at lysine 9 (H3K9) and DNA methylation are among the most highly conserved epigenetic marks that correlate well with gene silencing. The tumor microenvironment significantly influences therapeutic responses and clinical outcomes. The epigenetic-regulation mechanism of the costimulatory factors Tim-3 and galectin-9 in cervical cancer remains unknown. Methods: The methylation status of HAVCR2 and LGALS9 was detected by MS-PCR in cervical cancer tissues and cell lines. The underlying molecular mechanisms of SUV39H1-DNMT3A-Tim-3/galectin-9 regulation was elucidated using cervical cancer cell lines containing siRNA or/and over-expression system. Confirmation of the regulation of DNMT3A by SUV39H1 used ChIP-qPCR. Results: Here, we show that SUV39H1 up-regulates H3K9me3 expression in DNMT3A promoter region, which in turn induced expression of DNMT3A. In addition, our mechanistic studies indicate that DNMT3A mediates the epigenetic modulation of the HAVCR2 and LGALS9 genes by directly binding to their promoter regions in vitro . Moreover, in an in vivo assay, the expression profile of SUV39H1 up-regulates the level of H3K9me3 in the DNMT3A promoter region was found to correlate with Tim-3 and galectin-9 expression at the cellular levelindicating that SUV39H1-H3K9me3-DNMT3A is a crucial regulatory axis in cervical cancer. Conclusion: These results indicate that SUV39H1-DNMT3A is a crucial Tim-3 and galectin-9 regulatory axis in cervical cancer.

its ligand galectin-9, Tim-3 inhibits Th1 and Th17 responses by hampering their expansion, its mediating immune exhaustion in tumor microenvironment [5][6][7]. Tim-3-expressing CD4 + T cells in human cervical cancer could represent the functional regulatory T cells which contribute to the formation of the immune-suppressive tumor micromilieu [8].
The epigenetic regulation of genes, is critical for gene's transcription [9]. DNA methylation is often associated with gene expression changes that occur in cervical cancer [10]. Our previous study revealed that EZH2, H3K27me3 and DNMT3A mediate the epigenetic regulation of the negative stimulatory molecules, Tim-3 and galectin-9 in cervical cancer which is associated with HPV18 infection [11]. Trimethylation of histone 3 lysine 9 (H3K9me3) at gene promoters is a major epigenetic mechanism that silences gene expression [12,13] and SUV39H1 is H3K9me3-specific histone methyltransferase [14].
In this study, we identified a critical role of histone and DNA methylation marks in regulating costimulatory factors Tim-3 and galectin-9 expression in cervical cancer. The underlying mechanism is mediated by the repression of Tim-3 and galectin-9 through recruitment of DNMT3A to their promoters. SUV39H1 targeted DNMT3A by increasing the level of H3K9me3 in DNMT3A promoter region so that regulated Tim-3 and galectin-9 expression by DNA methylation in cervical cancer.
These results represent a significant step forward in understanding the contribution of SUV39H1 and DNMT3A to cancer progression and in providing a potential target for epigenetic-based cervical cancer therapy.

Patients and samples
24 cervical cancer tissues, accordingly matched peri-carcinomatous tissues and 16 normal cervical tissues were obtained from the First Affiliated Hospital of Xi'an Jiaotong University between January 2014 and December 2017. All patients were diagnosed by two senior pathologists and none had received chemotherapy or radiotherapy prior to surgery. The cervical cancer samples were collected as previous described [15]. After the tissues were dissected, each sample was washed with sterilized PBS and stored at -80°C. All procedures were performed on ice.

Data mining
Oncomine database (www.oncomine.org) was used to detect the HAVCR2 and LGALS9 mRNA expression levels in cervical cancer and normal cervix tissues. The correlation among HAVCR2 and LGALS9 expression were studied by the data obtained from the GEPIA (http://gepia.cancer-pku.cn/).

Cell lines and culture conditions
The cervical cancer cell lines SiHa, HeLa and C33A were obtained from Cell Bank, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai. All these cells were cultured in the high glucose Dulbecco's Modified Eagle's Medium (DMEM) (HyClone, USA) supplemented with 10% fetal bovine serum (FBS) (Biological Industries, Israel) at 37°C in an atmosphere of 5% CO 2 .

Lentivirus vectors and stable expression cell lines construction
Lentiviral vector preparation of Plenti-CMV-puro-Dest vector containing SUV39H1 fragment. The SUV39H1 fragment was cloned from the genomic DNA of the cell line SiHa. DNA fragment treated with Kpn1 and Xho1 and then the target gene was linked to entry vector pENTR-MCS. Two-plasmid of Plenti-CMV-puro-Dest and pENTR-MCS recombination reactions were performed using LR Clonase II (Invitrogen, USA). Use Lip2000 (Invitrogen, USA) transfection plasmid into SiHa and HeLa cell lines.
The transduced cells were then selected with puromycin. Stably transduced cells were maintained in culture in the presence of puromycin. The cell lines were named SiHa-SUV39H1, HeLa-SUV39H1, successively. The expression of SUV39H1 was determined by western blotting.

RNA interference
SiHa and HeLa were transfected with scramble and SUV39H1 and DNMT3A specific siRNA (GenePharma, Shanghai), the following siRNA oligos for SUV39H1 and DNMT3A are listed in table 1.
The siRNAs respectively using X-tremeGENE siRNA Transfection Reagent (Roche) and analyzed for SUV39H1 and DNMT3A expression levels by western blotting. All cell lines were named SUV39H1-siRNA and DNMT3A-siRNA, successively.

DNA extraction, bisulfite modification and methylation-specific PCR (MS-PCR)
Genomic DNA was isolated from cells and tissues using a TaKaRa Mini BEST Universal Genomic DNA Extraction Kit (TaKaRa, China) according to the manufacturer's instructions. DNA modification was done as previous [15], in briefly, 500 ng of genomic DNA was bisulfite-modified by a EZ DNA Methylation-Gold™ Kit (Zymo Research, USA). The primers used in the MS-PCR are listed in table 1.
The annealing temperature for the methylated primers of HAVCR2 and LGALS9 were 60°C and 60℃ while that for the unmethylated primers were 55°C and 56.3℃. The MS-PCR products were separated on a 2% agarose gel, stained with Gelview and visualized under ultraviolet illumination (Bio-Rad, USA). Methylation level was calculated by the ratio of methylated and unmethylated levels. Grey value of each band represented its relative expression and was measured by Image J Software. Each reaction was performed in triplicate.

Western blotting analysis
Cells were harvested in RIPA Lysis Buffer which containing 1mM PIC and 1mM PMSF. Proteins were resolved by SDS-PAGE and electroblotted onto PVDF membrane (Millipore, Billerica, USA) was blocked 1 hour at room temperature in 5% skim milk and incubated with primary antibodies for overnight at 4℃ followed by HRP conjugated secondary antibodies for 1 h at room temperature.
Chemiluminescence signal was detected following incubation with enhanced chemiluminescence reagent (Millipore, Billerica, Mass). Grey value of each band was measured with Image J Software. The antibodies are listed in table 2.

Chromatin immunoprecipitation (ChIP) assay and ChIP-qPCR
ChIP assays were carried out using the Simple ChIP Enzymatic Chromatin IP Kit (Cell Signaling Technology, USA) according to the manufacturer's instructions. The DNMT3A, HAVCR2 and LGALS9 promoters were detected by qPCR using promoter DNA-specific primers are listed in table 1. The qPCR with the former methods [16]. We used the cycle threshold (CT) as the representative point. The relative expression of genes in each group (fold-change compared with control) was calculated using the formula: RQ = 2 -△△Ct . Each reaction was performed in triplicate. The antibodies are listed in table 2.

Immunofluorescence staining
SiHa and HeLa cells were incubated with the primary antibody overnight at 4 °C. After thorough washing, the cells were incubated with Cy3-conjugated goat anti-rabbit IgG and FITC-conjugated Donkey anti-goat IgG for 1 h at room temperature. Finally, DAPI Fluoromount-G (SouthernBiotech) was used to counterstain the cell nuclei. The fluorescent was detected, and images were taken by Leica inverted fluorescence microscope. The antibodies are listed in table 2.

Immunohistochemical staining
Human cervical cancer specimens were fixed with neutral formalin, embedded in paraffin, and sectioned at a thickness of 4 μm. Sections were deparaffinized in xylene and rehydrated in a graded alcohol series. Antigen retrieval was performed using 0.01-M citrate buffer and 2 min of boiling.
Hydrogen peroxide was applied to block endogenous peroxidase activity, and then sections were incubated with normal goat serum to block nonspecific protein binding. Sections stained with primary antibody for Tim-3 and galectin-9 were incubated overnight at 4 °C. Sections were stained in parallel with PBS as a negative control. Tim-3 and galectin-9 expression were then detected using DAB, and slides were counterstained with hematoxylin. Slides were view at 400× magnification. The antibodies are listed in table 2.

Xenograft mouse model
BALB/c nude mice (4-week-old) used in this study and were maintained in a specific-pathogen-free (SPF) condition facility. Mouse injected subcutaneously with 1×10 7 SiHa-SUV39H1/HeLa-SUV39H1 cells were randomly divided into four groups when tumor volumes were around 100mm 3 : (1/2) SiHamock/HeLa-mock control groups; (3/4) SiHa-SUV39H1/HeLa-SUV39H1 groups. Two diameters of the individual tumor were measured by electronic slide caliper every two days. Tumor volume was calculated using the following formula: tumor volume (mm 3 ) =0.5×length×width 2 . Mice were monitored for 21 days, at which time mice were euthanized and tumors and organs were extracted.

Statistical analysis
Statistical analyses were performed using GraphPad Prism 7 software (GraphPad Software, USA).
Paired t test and one-way ANOVA analysis were carried out on samples within groups. The p value of <0.05 was considered statistically significant. The p values are represented as **p<0.01, *p<0.05.
The data are presented as mean ± standard error of the mean (SEM). All experiments were independently repeated at least thrice, with consistent results.

Tim-3 and galectin-9 expression were increased due to genes methylation level decreased in cervical cancer
Using the Oncomine databases (https://www.oncomine.org/), we compared the mRNA expression of HAVCR2 and LGALS9 between cervical cancer and normal cervical samples. The results indicated that the expression levels of HAVCR2 and LGALS9 were all higher in cancer than in normal cervical samples (Fig. 1a, b). Furthermore, HAVCR2 was positively corrected with LGALS9 (R = 0.26, p <0.05) based on GEPIA (Gene Expression Profiling Interactive Analysis) dataset (http://gepia.cancer-pku.cn/) ( Fig. 1c). Hence HAVCR2 has a positive correlation with LGALS9 in cervical cancer. The expression of Tim-3 and galectin-9 protein in cancer tissues were higher than in normal cervix tissues (Fig. 1d, e).
The detail data of patients' clinicopathological is shown in table 3.
The online software "MethPrimer" (http://www.urogene.org/methprimer/) profiled CpG island in the region that was located from -2000 to -200 bp upstream from ATG, the transcription starts site (TSS) in the HAVCR2 and LGALS9 promoters respectively (Fig. 1f). One pair of primers was designed to amplify the genes promoter regions respectively. HAVCR2 and LGALS9 promoters in cervical cancer tissues displayed hypermethylation status in normal cervical tissues (Fig. 1g, h), possibly leading to the inhibition of its gene expression in normal cervical tissues. Immunohistochemistry results revealed that Tim-3 and galectin-9 expressed in tumor cells of cervical cancer tissues (Fig. 1i).

Tim-3 and galectin-9 expression were reversed by alter the methylation status in the promoter regions of HAVCR2 and LGALS9 in cervical cancer cells
It showed that HAVCR2 and LGALS9 promoter regions from ATG in SiHa, HeLa and C33A cells were all partially methylated (Fig. 2a). The Tim-3 and galectin-9 expressed in SiHa, HeLa and C33A cells (Fig.   2b). The mRNA expression level of HAVCR2 and LGALS9 in SiHa, HeLa and C33A cell lines after treatment with DNA demethylation reagent 5-aza-2'-deoxycytidine (5-Aza-CdR) to identify whether the methylation status in the promoter regions regulate the expression of these genes at the transcription level. The results suggested that the expression of HAVCR2 and LGALS9 mRNA in SiHa, HeLa and C33A cells increased in dose-dependent manner after cellular DNA demethylation (Fig. 2c,   d). It illustrated the Tim-3 and galectin-9 expression were reversed by 5-Aza-CdR, which promoted the expression of these genes at the transcriptional level. What's more, immunofluorescence assay showed that the SiHa and HeLa cells all staining cytoplasm and nuclear Tim-3 and galectin-9 (Fig. 2e).
As shown in Fig. 4b, Overexpressed SUV39H1 in SiHa and HeLa cells (Fig. 4a) significantly increased H3K9me3 and DNMT3A expression. Fig. 4c showed that knocking-down SUV39H1 expression in SiHa and HeLa cells displayed dramatically down-regulation in H3K9me3 and DNMT3A protein expression (Fig. 4d). Collectively, these results indicated that SUV39H1 participate in regulation of DNMT3A through changing H3K9me3 expression in cervical cancer cells.
In a screening for epigenetic mechanisms that regulating DNMT3A expression, ChIP analysis revealed that the SiHa and HeLa cells exhibit the highest H3K9me3 level in a region upstream of the transcription initiation region (Fig. 4f). H3K9me3 regulated the expression of DNMT3A by acting on the -1000 to +1 region of the promoter region of DNMT3A (Fig. 4g, h). Taken together, above results revealed that SUV39H1 regulated the expression of DNMT3A through elevating H3K9me3 level on the DNMT3A promoter in cervical cancer cells.

The methylation status of HAVCR2 and LGALS9 affected by SUV39H1 in cervical cancer cells
We have determined the baseline levels of DNA methylation on HAVCR2 and LGALS9 promoters among cervical cancer cell lines and then evaluated whether SUV39H1 mediated DNA methylation through DNMT3A is required for HAVCR2 and LGALS9 transcription. The results showed that SUV39H1 overexpression increased methylation levels at the HAVCR2 and LGALS9 promoters (Fig. 5a, b), these changed methylation level contributed to the decrease of Tim-3 and galectin-9 expression among overexpressed SUV39H1 cell lines (Fig. 5e). SUV39H1-knockdown cells showed the opposite results ( Fig. 5c, d, f).
These results indicating that changes in histone modification precede the changes in DNA methylation level of HAVCR2 and LGALS9. Consistent with SUV39H1 affecting the expression of DNMT3A.

SUV39H1 mediated Tim-3 and galectin-9 expression through DNA methylation in vivo
For the purpose of investigating SUV39H1 mediated the costimulatory factors Tim-3 and galectin-9 expression through DNA methylation in vivo, we generated SiHa-SUV39H1 and HeLa-SUV39H1 tumor xenografts in nude mice. As shown in Fig. 6a, tumors formed from the SiHa-mock and HeLa-mock cells (Fig. 6b, c) grew faster than those formed from the SiHa-SUV39H1 and HeLa-SUV39H1 cells, respectively. All these data indicated that up-regulating SUV39H1 may inhibited the tumor growth in SiHa and HeLa cells.
To determine SUV39H1 mediated Tim-3 and galectin-9 expression through DNA methylation in vivo, the levels of H3K9me3, DNMT3A and Tim-3 and galectin-9 in the tumor xenograft tissues were examined by western blotting. As shown in Fig. 6d-f, the expression of DNMT3A increased significantly when SUV39H1 overexpressed, followed by the down-regulation of Tim-3 and galectin-9 in SiHa-SUV39H1 and HeLa-SUV39H1 cells derived tumors. SUV39H1 overexpression significantly upregulated the methylation level of HAVCR2 and LGALS9 in tumor tissues (Fig. 6g). ChIP analysis revealed that H3K9me3 regulated the expression of DNMT3A by acting on the -1000 to +1 region of the promoter region of DNMT3A (Fig. 6h, i) in tumor tissues. H3K9me3 directly regulate the expression of DNMT3A in vivo, these results indicating that SUV39H1 precede the changes in DNA methylation. All the results in xenograft tissues were consistent with those in vitro, indicating that similar SUV39H1 mediated Tim-3 and galectin-9 expression through DNA methylation in vivo.

H3K9me3 expression was independent from HR-HPV oncogenes
As persistent HR-HPV infection contributes to almost all cervical cancer cases [19], we attempted to explore whether HR-HPV oncogenes E6 and E7 participated in SUV39H1 mediated DNA methylation in SiHa and HeLa cell lines. We found that there was no difference in the expression level of H3K9me3 in cervical cancer cells SiHa, HeLa and C33A (Fig. 4e). Expression of HPV16/18 E6 and E7 oncogenes were detected by western blotting in overexpression or knockdown SUV39H1 SiHa and HeLa cells respectively. As show in Fig. 7i and j, the expression level of HR-HPV oncogenes E6 and E7 were no difference between overexpressed or knockdown SUV39H1 SiHa and HeLa cells with control. In the following studies we transiently overexpressed or knocked-down HPV16/18 E6 and E7 in cervical cancer cells for further illustration (Fig. 7a-d). The results showed that, the level of H3K9me3 was not changed in over-expressed or knocked-down HPV16/18 E6 and E7 cells compared with control ( Fig.   7e-h).
Taken together, our data suggested that H3K9me3 expression was independent from HR-HPV oncogene E6 and E7 in cervical cancer.

Discussion
Aberrant DNA methylation is recognized as one of the most important events in cervical cancer carcinogenesis, which causes silencing of certain genes [20,21]. Epigenetic modification plays an important role in regulating immune cell differentiation [22,23]. The methylation status of immune genes influences the tumor immune response in the tumor microenvironment (TME) [24,25]. The study showed that decreased activity of DNMTs in CD4 + Tregs was accompanied by demethylation of the forkhead box P3 (FOXP3) gene promoter and downregulation of immune responses in the TME [26]. Hypermethylation associated SMAD family member 3 (SMAD3) silencing in CAFs, which was associated with aberrant response to exogenous TGF-β1 [27]. Here, we identified a novel function of SUV39H1 regulates DNMT3A expression through elevating H3K9me3 level in DNMT3A promoter, which could mediate Tim-3 and galectin-9 expression through DNA methylation in cervical cancer.
Tim-3 and galectin-9 are over-expressed in cervical cancer tissues, this biological effect is mediated through the aberrant epigenetic of Tim-3 and galectin-9, which is facilitated by the recruitment of DNMT3A to their promoter regions. Meanwhile, SUV39H1 contributed to Tim-3 and galectin-9 regulation by up-regulation H3K9me3 level in DNMT3A promoter which directly binding to the promoter of DNMT3A.
We found that Tim-3 and galectin-9 were over-expressed in cervical cancer tissues related to promoter regions of HAVCR2 and LGALS9 were hypo-methylated, and they were partial methylation in

Conclusion
In summary, our present study highlights the role of SUV39H1 and H3K9me3 in the DNA methylation regulation of Tim-3 and galectin-9 in cervical cancer microenvironment (Fig. 8). We provide a potential direction in exploring the relationship between SUV39H1 and DNMT3A. These findings add diverse roles and mechanistic insight into our understanding of crosstalk of SUV39H1 with DNMT3A.
Declarations wrote and edited the manuscript. All authors read and approved the final manuscript.

Data availability statement
The dataset analyzed during the current study are publicly available from the online database: GEPIA database (http://gepia.cancer-pku.cn/) and Oncomine database (www.oncomine.org).

Consent for publication
Not applicable.

Conflicts of interest
The authors declare no potential conflicts of interest.   Tables   Table 1 Primer sequences Figure 1 The expression of Tim-3 and galectin-9 in cervical cancer tissues and the genes methylation HR-HPV E6/E7 wasn't participate in H3K9me3 mediated DNA methylation in cervical cancer.