STAT5 and TET2 Cooperate to Regulate FOXP3-TSDR Demethylation in CD4+ T Cells of Patients with Colorectal Cancer

The tumor-infiltrating Tregs are linked to colorectal cancer progression and outcome. FOXP3 is regarded as a critical developmental and functional factor for Tregs. FOXP3-TSDR demethylation is required for stable expression of FOXP3 and maintenance of Treg function. In our study, we found specific DNA hypomethylation of FOXP3-TSDR in CD4+ T cells from colon tumor tissues as compared with normal colonic tissues. Moreover, we also found that the expression of STAT5 and TET2 was increased in CD4+ T cells from colon tumor tissues, and the superfluous STAT5 and TET2 binding to FOXP3-TSDR resulted in DNA hypomethylation. In conclusion, we have demonstrated that excessive amounts of STAT5 may bind more TET2 to the FOXP3-TSDR and upregulate FOXP3 expression via DNA demethylation. Our study improved the mechanism of FOXP3-TSDR hypomethylation in tumor-infiltrating CD4+ T cells of CRC patients.


Introduction
Colorectal cancer (CRC) is the development of cancer in the colon or rectum. It is one of the most common fatal malignancies in the world; globally, more than 1 million people get colorectal cancer every year [1]. The incidence of CRC in China is lower than that in the west countries but has increased in recent years and has become a substantial cancer burden in China, particularly in the more developed areas [2].
Tregs are considered to be a major cell population involved in tumor immune tolerance [3,4]. The transcription factor forkhead box P3 (FOXP3) is a subfamily member of forkhead transcription factor, which is specifically expressed in CD4 + CD25 + Tregs, and it is regarded as a critical developmental and functional factor for Tregs [5]. Treg-specific demethylated region (TSDR) is a CpG dinucleotide-rich and highly conserved region within the conserved noncoding sequences 2 (CNS2) at the FOXP3 locus [6]. TSDR demethylation is necessary for stable FOXP3 expression and maintenance of the suppressive phenotype for nTregs [7].
Recently, investigators have evaluated that the tumorinfiltrating FOXP3 + Tregs are linked to colorectal cancer progression and outcome [8][9][10]. Zhuo et al. found significantly higher FOXP3-TSDR demethylation rates in tumor sites versus normal sites in patients with CRC, as well as significantly more FOXP3 mRNA expression and higher protein synthesis in tumor tissues [11]. However, the molecular mechanism of FOXP3-TSDR excessive demethylation in patients with CRC has not been fully understood.
Ten-eleven translocation proteins (TET1, TET2, and TET3) catalyze 5-methylcytosine (5mC) conversion to 5-hydroxymethylcytosine (5hmC) to mediate DNA demethylation [12]. It has been found that TET1 and TET2 bind to the FOXP3 CNS2 region in Tregs, suggesting a role for the TET proteins in the maintenance of Tregs identity [13]. Moreover, IL2 can maintain the high level of TET2 during the thymic Treg development, and downregulation of TET2 expression prevents FOXP3-TSDR demethylation [14]. Nevertheless, the molecular mechanism of TET2 targeting regulation of FOXP3-TSDR demethylation is still unclear. STAT5 is an important transcription factor for regulating FOXP3 expression. In the FOXP3 locus, there are multiple STAT5 binding sites, which are located in the 5 ′ -flanking and intron regions including CNS2 [15][16][17]. However, whether STAT5 participates in the FOXP3-TSDR demethylation process is still unknown.
In this study, we aimed to investigate the correlation between STAT5 and TET2 in demethylation process of FOXP3-TSDR in CD4 + T cells of patients with CRC. Our data indicated that STAT5 and TET2 were significantly upregulated in tumor-infiltrating CD4 + T cells of CRC. We also provided evidences indicating that overexpressed STAT5 combines with TET2 to form complexes that bind to the FOXP3-TSDR, resulting in excessive demethylation in tumor-infiltrating CD4 + T cells. Taking together, our findings reveal that the increased STAT5 and TET2 play an important role in augment of FOXP3 and the pathogenesis of CRC. Only patients who underwent colectomy for colon cancer without chemotherapy or radiotherapy before surgery were selected, no matter their TNM stage. The characteristics of these patients are shown in Table 1.

Materials and Methods
Fresh colon tumor tissues and corresponding normal colonic tissues were obtained immediately after surgery. The mononuclear cells were separated from these tissues in the following protocol. In short, the tissue samples were minced and incubated at 37°C for 20 min in Hanks' buffered saline solution (HBSS) containing 2.5% FBS and 1 mM dithiothreitol, and then incubated in 0.75 mM EDTA at 37°C in three rounds of 15 min each, mucus and epithelial cells were removed, respectively. Further, samples were transferred in gentleMACS C tubes (Miltenyi) containing calcium and magnesium supplemented HBSS with 0.5 mg/mL Collagenase IV (Sigma), 50 ng/mL DNAse I (Thermo Fisher), 2% FBS, and 10% BSA. Tissue dissociation was made on a gentle-MACS Octo Dissociator with Heaters (Miltenyi). Single cell suspensions were washed with cold HBSS. Then, cells were pelleted through a 40% isotonic Percoll solution and finally centrifuged over a Lympholyte density gradient. CD4 + T cells were further isolated from mononuclear cells using human CD4 beads (Miltenyi) according to protocols provided by the manufacturer.
2.2. RNA Isolation and Real-Time PCR. Total RNA extraction in CD4 + T cells was performed with TRIZOL (Invitrogen). cDNA was synthesized according to the instructions in the SuperScript II reverse transcriptase (Invitrogen). mRNA expression was assessed by real-time PCR using the 7500 Real-Time PCR Systems (Applied Biosystems). The mRNA levels of various genes were calculated after normalizing with GAPDH, and the 2 −ΔΔCt method was used for analyzing target gene expression level. Primers are listed in Table 2. 2.3. Western Blot Analysis. CD4 + T cells were lysed in 1 ml of lysis buffer containing protease inhibitor (Thermo Scientific) at 4°C. Lysates were centrifuged for 15 min at 14,000g at 4°C, and protein concentration was determined by Bradford protein assay (Thermo Scientific). Soluble lysates were loaded in each lane and separated by sodium dodecyl sulfatepolyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Millipore). Membranes were blocked with 5% nonfat dry milk in Tris-buffered saline containing 0.1% Tween-20 buffer (TBST) for 1 hr, washed with TBST, and incubated with primary antibodies. Antibodies include anti-STAT5 (ab126832), anti-TET2 (ab94580), anti-FOXP3 (ab10901), and anti-GAPDH (ab181602); all antibodies were obtained from Abcam Company. Blots were rinsed with TBST and subsequently incubated with horseradish peroxidase-conjugated secondary antibodies for 1 hr at room temperature. The intensity of the identified bands was quantified using Quantity One software (Bio-Rad).

DNA Methylation of FOXP3-TSDR in Colon Tumor
Tissues and Normal Colonic Tissues. To investigate whether the FOXP3-TSDR was aberrantly DNA methylated in CD4 + T cells from colon tumor tissues, we did bisulfate sequencing analysis for TSDR. The FOXP3-TSDR was found to be significantly hypomethylated in CD4 + T cells from colon tumor tissues as compared with normal colonic tissues (Figure 1(a)). Further, we measured the expression of FOXP3 in colon tumor tissues and normal colonic tissues. The results showed that FOXP3 mRNA (Figure 1(b)) and protein levels ( Figure 1(d)) were sharply increased in CD4 + T cells from colon tumor tissues. Further, we found that the FOXP3-TSDR DNA methylation was negatively correlated with its mRNA level through the correlation analysis (Figure 2(c)). Together, our results suggest that the abnormally reduced Foxp3-TSDR methylation may be an important factor to increase the expression of FOXP3 in CD4 + T cells of CRC patients.

The
Interrelationship between STAT5 and TET2 in CRC CD4 + T Cells. We measured the mRNA and protein expression levels of STAT5 and TET2 in CD4 + T cells from colon tumor tissues coupled with corresponding normal colonic tissues. The results from real-time PCR showed that STAT5 and TET2 expression was significantly upregulated in tumor tissues compared with normal tissues (Figures 2(a) and 2(b)). Western blot further confirmed the elevation of STAT5 and TET2 (Figure 2(c)).
In order to validate the interaction between STAT5 and TET2, we performed coimmunoprecipitation using anti-STAT5 in CD4 + T cells from colon tumor tissues. The results showed that STAT5 could combine with TET2 to form a complex (Figure 2(d)), and more complexes were found in tumor tissues compared with normal tissues (Figure 2(e)).

TET2 Binds to the FOXP3-TSDR and Regulates Its DNA Methylation via STAT5.
To confirm that the abnormally increased STAT5 and TET2 bind more strongly to FOXP3-TSDR in CD4 + T cells from colon tumor tissues compared with which from normal tissues, we performed ChIP-qPCR analysis using anti-STAT5 and TET2 antibody. As shown in Figures 3(a) and 3(b), CRC patients showed stronger binding of STAT5 and TET2 to FOXP3-TSDR in tumor tissues compared with normal tissues. Further, we suppressed the expression of STAT5 by transfection of siRNA-STAT5 and detected the levels of TET2 binding to FOXP3-TDSR in CD4 + T cells from tumor tissues. The result showed that the expression of STAT5 and FOXP3 (Figure 3(c)) and the binding level of TET2 in FOXP3-TSDR (Figure 3(d)) were decreased obviously, while the FOXP3-TSDR methylation (Figure 3(e)) was increased significantly after interference with STAT5 expression. In addition, we upregulated the expression of TET2 with or without STAT5 inhibition in CD4 + T cells from normal human donors. We found that overexpression of TET2 resulted in a significant increase in the combination of FOXP3-TSDR (Figure 3(g)), obviously reducing the DNA methylation of TSDR (Figure 3(h)), and upregulating the expression of FOXP3 (Figure 3(f)). Interference with STAT5 expression inhibits the binding of TET2 to FOXP3-TSDR (Figure 3(g)) and partly reverses the reduced DNA methylation of TSDR (Figure 3(h)) and the increased FOXP3 expression caused by TET2 overexpression (Figure 3(f)). Overall, these data suggest that STAT5 may be a key factor in the process of TET2mediated demethylation of FOXP3-TSDR; excessive amounts of STAT5 may bind more TET2 to the FOXP3-TSDR and upregulate FOXP3 expression via DNA demethylation.

Discussion
Numerous studies have verified an increase of Tregs in the circulation, draining lymph nodes, and at the tumor sites of patients with malignancy, including breast, lung, pancreas, and skin cancers [8,[20][21][22][23][24]. Colorectal cancer is one of the most common cancers worldwide and a major cause of cancer-related death [25]. Therefore, it is meaningful to explore the Tregs and their possible role in CRC, as well as the potential significance in the targeted therapeutics. For CRC patients, increased numbers of Tregs had been shown in peripheral blood, tumor-draining lymph node, and tumor site, and these Tregs could actively migrate to the site of immune activity [8,26,27]. Moreover, Salama et al. found that the high FOXP3 + Tregs density in normal colonic mucosa of CRC patients was associated with worse prognosis [9]. Liu   expression of FOXP3 may be a key factor in CRC pathological process; however, the molecular mechanism of FOXP3 overexpression remains unclear. STAT5 is necessary and sufficient for Treg development. Specifically, T cell specific deletion of STAT5 prevents Treg development. TGF-β signaling pathway can activate STAT5, mediate it binding to the CNS2 region of FOXP3, and promote transcription of FOXP3, but the mechanism is unknown [17]. Ichiyama et al. found that TET2 could promote DNA demethylation and activation of cytokine gene expression in T cells [28]. Nair and Oh through the studies on experimental animals found that downregulation of TET2 prevents FOXP3-TSDR demethylation in IL2 deficient regulatory T cells [14]. In combination with their research, we speculate that TET2 may participate in the process of FOXP3-TSDR demethylation via STAT5 guidance. Our study found significant hypomethylation of FOXP3-TSDR in tumor-infiltrating CD4 + T cells of CRC patients, which is consistent with the results of other investigators [11]. Furthermore, we also found that the expression of STAT5 and TET2 were increased in CD4 + T cells from tumor tissues, and the massive STAT5 binds more TET2 to the FOXP3-TSDR and upregulates FOXP3 expression. This provides an explanation for the molecular mechanism of STAT5 and TET2 upregulation of FOXP3 expression.
In conclusion, we demonstrated that the excessive expression of FOXP3 in tumor-infiltrating CD4 + T cells of CRC patients is caused, at least in part, by the upregulation of STAT5 and TET2. The unbalanced TET2 may bind to the FOXP3-TSDR overly via superfluous STAT5, thereby contributing to DNA demethylation and mRNA transcription. Thus, our study improved the mechanism of FOXP3-TSDR hypomethylation in CD4 + T cells of CRC patients and provided a novel therapeutic target of CRC.

Conflicts of Interest
The authors declare that they do not have any conflicts of interest.