In this study, we obtained high-purity of human CD4+
Treg cells and aTconv cells by combining in vitro
expansion, MACS- and FACS- sorting methods (). As reported by other researchers employing this technique, these cells maintained all phenotypic, functional and epigenetic Treg cell characteristics, even after extensive in vitro
. We utilized these cells for ChIP-Seq analysis to generate high-resolution maps of the genome-wide distribution of H3K4me1 and H3K4me3 in both cell subtypes. Ultimately, we identified a number of cell type-specific H3K4me1 regions and H3K4me3 marked proximal promoters in Treg cells. The majority of the differential H3K4me1 regions were found to be located in promoter-distal sites, and we selected some for verification of their enhancer activity by using reporter gene assays.
CD4-positivity and CD25-positivity have long been considered as the cell-specific indicators of Treg cells. However, CD4+CD25- T cells were demonstrated to be able to up-regulate their CD25 expression upon activation by antigen, indicating that CD4 and CD25 double-positive T cells actually represent a heterogeneous cell population and these surface markers are not sufficient identifiers of Treg cells. Thus, we used CD4+CD25+FOXP3+ triple expression to define Treg cells since FOXP3 gene expression is essential for Treg cell function. We carried out comparative analysis of the genome wide epigenetic methylation status for H3K4 in CD4+CD25+FOXP3+ (Treg) and CD4+CD25+FOXP3- T cells (aTconv).
The low frequency of Treg cells in normal human peripheral blood has thus far limited the detailed characterization and potential clinical application of human Treg cells. In many previous studies, Treg expansion has been carried out to obtain enough cell material to perform analysis. Unfortunately, CD4 and CD25 were usually used to identify the Treg cells. Here, we found that although CD4+CD25+ T cells were expanded up to 1000-fold, most of the expanded cells were FOXP3-negative. Thus, we performed FACS-sorting to obtain high-purity Treg cells with CD4, CD25 and FOXP3 expressions immediately prior to our ChIP-Seq assay.
Previously, Heintzman determined the chromatin modification states at high resolution along 30 Mb of the human genome, and found that active promoters were marked by H3K4me3 and enhancers by H3K4me1 
. We also found that most proximal promoters enriched by H3K4me3 were common between the Treg and aTconv cells, suggesting that the related genes of the proximal promoters were co-expressed in the two lineages. Although some genes are widely used as markers for Treg cells, such as IL2RA, CTLA4 and TNFRSF18, accumulating evidence has unfortunately suggested that these markers are not strictly Treg-specific. Upon activation, all T cells express CD25, the alpha-chain of the IL-2 receptor 
, and its combination with IL-2 is essential for T cell clonal expansion. CTLA-4, which is the receptor for APC-B7, negatively regulates the IL-2 production of the newly activated T cell and inhibits further T cell proliferation upon binding of B7 and is up-regulated on all CD4+
T cells, two to three days following activation 
. Similarly, the expression of TNFRSF18, which is a possible target molecule in cell contact-dependent suppression, is induced in T cells upon activation 
. This could explain why we observed H3K4me3 in the proximal promoters of these genes in aTconv cells. STAT family TFs are critical for T cell differentiation; however, their expression is not sufficient to drive lineage commitment. Consistent with the ubiquitous expression patterns of STAT family TFs, we found that most STATs were marked in their promoter regions by H3K4me3, in both Treg and aTconv cells. Based on these results, we predicted that the common 19927 genes between Treg and aTconv cells () may be expressed in both lineages.
We found that, apart from the common H3K4me3 promoters, there were also some Treg cell-type specific proximal promoters marked by H3K4me3, such as FOXP3 (). It may be these types of specific proximal promoters, especially the FOXP3, that are responsible for the differences between Treg cell and aTconv cells. The proximal promoter of FOXP3, which is believed to serve as a master regulator of Treg cells, was found to be enriched by H3K4me3 in Treg cells. Moreover, a 50-fold higher mRNA expression level was observed in Treg cells, as compared to aTconv cells. This nearly exclusive expression of FOXP3 in Treg cells was in accordance with the current concept that FOXP3 represents the critical TF of Treg cells. In addition, we also found that the CCR7 gene was marked by H3K4me3 in its proximal promoter only in Treg cells, and exhibited a nearly 20-fold increase of mRNA expression in Treg, as compared to aTconv cells.
H3K4me3 is usually associated with promoters, and its occurrence at enhancers remains a topic of debate. Whereas Heintzman, et al
. found little or no H3K4me3 at p300-associated enhancers 
, Barski, et al
. identified all three methylation states at the related functional enhancers 
. It is, therefore, unclear whether the promoter-distal H3K4me3 sites identified in this study are associated with uncharacterized functional transcription units, or whether they are able to act as enhancer regions themselves. For example, we found there was a region located about 6 kb downstream of the FOXP3 promoter (ChrX:49001620–49002192), which showed enhancer activity in transient transfection assays; the existence of this region suggests that there may be some non-promoter H3K4me3 regions associated with enhancers. Another region located about 4 kb downstream of the FOXP3 promoter was specifically enriched by H3K4me3 in Treg cells (ChrX:49004128–49005080), and also exhibited enhancer activity; interestingly, previous studies have shown that this region was enriched for STAT5 consensus sites. Treg cell survival critically depends on interaction with IL-2. The TF STAT5 is activated through the IL-2 receptor 
, has an essential role in Treg cell homeostasis 
, and is known to regulate the lineage-specific TF FOXP3 through an intronic, methylation-sensitive enhancer 
. Together, all the data indicate that certain promoter-distal H3K4me3 modified regions may have enhancer activity. Moreover, it is likely that some of the 27000 Treg cell-type specific H3K4me3 non-promoter regions that were identified in this study might be important for Treg cell-type specific patterns of gene expression.
Although the promoter region represents a primary element of gene expression, it is controlled by distal regulatory elements like enhancers and silencers. Previous studies have shown that H3K4me1 at promoter-distal sites was often associated with enhancer function 
. Our results indicated that most of the H3K4me1 islands were cell-type specific, suggesting that enhancers are the most variable class of transcriptional regulatory element between Treg and aTconv cells, and are probably of primary importance in driving Treg cell-type specific patterns of gene expression. Our study identified a number of putative regulatory elements for genes that are highly important for Treg cell functions. For instance, we found that there was a region located in intron 1 of the FOXP3 gene (ChrX:49001620–49002192) enriched by both H3K4me1 and H3K4me3 in Treg cells and which showed enhancer activity in transfected Jurkat cells. A region located upstream of IL2RA (Chr10:6148000–6148784) also showed enhancer activity. Since cultured and expanded conventional T cells express high levels of CD25 as a consequence of TCR activation, it is possible that this region may contribute to regulating constitutive (rather than activation-induced) CD25 expression in Treg cells.
In addition, we found that most H3K4me1 enriched regions were not enriched by H3K4me3, suggesting that most potential regulatory elements were only enriched by H3K4me1 but lacked H3K4me3 in the whole genome of human Treg and aTconv cells. This finding is consistent with the observations of p300-associated enhancers that were found to have little or no H3K4me3 
. However, there were also some regions simultaneously enriched by H3K4me1 and H3K4me3, such as the region located in intron 1 of the FOXP3 gene, which did show enhancer activity. Whether or not the regions enriched by the two types of histone methylations may harbor more potential to act as enhancers remains unknown.
In conclusion, we identified genome-wide H3K4me1 and H3K4me3 modification regions in Treg and aTconv cells. The H3K4me3 modifications located in proximal-promoter regions were nearly identical in both Treg and aTconv cells, with the exception of a few promoters of genes, such as FOXP3 and CCR7, which are expressed uniquely in Treg cells. In contrast to the H3K4me3 modification, H3K4me1 exhibited cell-type specific locations, indicating that enhancers are the most variable class of transcriptional regulatory elements between Treg and aTconv cells. Furthermore, enhancers are likely to be of primary importance in driving Treg cell-type specific patterns of gene expression. The Treg- and aTconv-specific H3K4me1 and H3K43 patterns may function as significant mediators of differentiation events, lineage commitment and cell type-specific gene expression. It is likely that this basic principle is not confined to these two closely related T cell populations, but may apply generally to somatic cell lineages in adult organisms.