Plasticity of Naturally Occurring Regulatory T Cells in Allergic Airway Disease Is Modulated by the Transcriptional Activity of Il-6

The impact of naturally occurring regulatory T cells (nTregs) on the suppression or induction of lung allergic responses in mice depends on the nuclear environment and the production of the pro-inflammatory cytokine interleukin 6 (IL-6). These activities were shown to be different in nTregs derived from wild-type (WT) and CD8-deficient mice (CD8−/−), with increased IL-6 levels in nTregs from CD8−/− mice in comparison to WT nTregs. Thus, identification of the molecular mechanisms regulating IL-6 production is critical to understanding the phenotypic plasticity of nTregs. Electrophoretic mobility shift assays (EMSA) were performed to determine transcription factor binding to four Il-6 promoter loci using nuclear extracts from nTregs of WT and CD8−/− mice. Increased transcription factor binding for each of the Il-6 loci was identified in CD8−/− compared to WT nTregs. The impact of transcription factor binding and a novel short tandem repeat (STR) on Il-6 promoter activity was analyzed by luciferase reporter assays. The Il-6 promoter regions closer to the transcription start site (TSS) were more relevant to the regulation of Il-6 depending on NF-κB, c-Fos, and SP and USF family members. Two Il-6 promoter loci were most critical for the inducibility by lipopolysaccharide (LPS) and tumor necrosis factor α (TNFα). A novel STR of variable length in the Il-6 promoter was identified with diverging prevalence in nTregs from WT or CD8−/− mice. The predominant GT repeat in CD8−/− nTregs revealed the highest luciferase activity. These novel regulatory mechanisms controlling the transcriptional regulation of the Il-6 promoter are proposed to contribute to nTregs plasticity and may be central to disease pathogenesis.


Introduction
IL-6 acts as a pro-inflammatory cytokine with a critical role in immune responses [1]. Impaired regulation of IL-6 can lead to immune-related diseases including rheumatoid arthritis, where increased IL-6 levels were observed, and has thus been used as a therapeutic target [2]. While IL-6 knockout mice have also been shown to present with heightened inflammatory responses with regional tissue damage [1,3], an excess of IL-6 is linked to Castleman's disease [4], indicating the importance of the appropriate regulation of IL-6 levels. IL-6 has also emerged as a pivotal factor in the pathogenesis of asthma [5][6][7], a heterogeneous syndrome involving complex pathophysiological pathways. In addition to IL-6, soluble IL-6 receptor (sIL-6R) levels have consistently been elevated in children and adults with asthma [5][6][7][8][9][10]. Levels were positively correlated with disease severity and inversely correlated with forced expiratory volume in 1 s (FEV1). T regulatory cells (Tregs), including naturally occurring T regulatory cells (nTregs, CD4 + CD28 high Foxp3 + )

Transcription Factor
Binding to the Il-6 Promoter Is Increased in nTregs from CD8 −/− Mice 2.1.1. Human and mouse IL-6 promoter share high homology Using the Visualization Tool for Alignment (VISTA) Genome Browser database [15,16], the human IL-6 and mouse Il-6 promoter revealed up to 95% homology (Online Supplemental Figure S1A). The region of the Il-6 promoter directly upstream of the TSS was more conserved between human and mice in comparison to its corresponding exons. nTregs from CD8 −/− mice produce significantly higher amounts of IL-6 than WT nTregs [14]. To identify cis-regulatory elements potentially contributing to the differential expression of IL-6 levels in nTregs from WT and CD8 −/− mice and their distinct suppressive capacities [13,14], in silico transcription factor binding analyses of approximately 500 bp upstream of the TSS of the mouse Il-6 promoter were performed. The Il-6 promoter region in mice revealed a sequence homology to human IL-6 of 71% (Online Supplemental Figure S1B). Using the MatInspector software tool [17], a total of 126 potential transcription factor binding sites were identified (Online Supplemental Table S1). Consistent with previous reports [18][19][20][21][22], this included predicted binding sites for NF-κB, AP-1, and C/EBP, which were highly homologous in humans (Online Supplemental Figure S1A,B).

Elevated
Binding of NF-κB, SP1, and SP3 in nTregs from CD8 −/− Compared to WT Mice To delineate formed DNA/protein interactions specific to nTregs, these three loci in the mouse Il-6 promoter upstream of the TSS homologous to human IL-6 were analyzed. They were suggested to influence its regulatory function (Online Supplemental Figures S1B and S2) when examined by EMSA with nuclear extracts isolated from nTregs of WT and CD8 −/− mice. A previously uncharacterized region within the mouse Il-6 promoter, most distinct from human IL-6 (Online Supplemental Figure S1B), was investigated (Il-6_4). Each 3 of 14 potential binding site under study was numbered in accordance with the relative distance to the TSS, respectively (Il-6_1, Il-6_2, Il-6_3, and Il-6_4, see Online Supplemental Figure S2).
For the region closest to the TSS of Il-6 (Il-6_1), a number of complexes were identified with an overall higher binding in unstimulated and stimulated (phorbol 12-myristate 13acetate (PMA)/ionomycin) nuclear extracts from CD8 −/− nTregs compared to WT nTregs ( Figure 1A, lanes 3 and 4 vs. 1 vs. 2). To verify the predicted binding of NF-κB, competition experiments with a high affinity NF-κB consensus site (100-fold molar excess) [23,24] were performed, leading to the abrogation of all complexes (Online Supplemental Figure S3A, lane 3). A DNA/protein complex was confirmed to selectively contain the NF-κB subunit p65 (Online Supplemental Figure S3A, lane 7), while no binding of the p50 subunit was observed (lane 6). A novel protein/DNA interaction of transcription factors of the specificity protein (SP) family including SP1 and SP3 in the nuclear environment of nTregs was identified. Their binding to Il-6_1 was confirmed by competition experiments with a SP consensus site [25,26] and using specific antibodies (Online Supplemental Figure S3A, lanes 4 and 8-11). Overall, the locus closest to the TSS of Il-6 showed greater binding of p65 and SP family members in CD8 −/− nTregs compared to WT nTregs before and after stimulation.

Mice
To delineate formed DNA/protein interactions specific to nTregs, these three loci in the mouse Il-6 promoter upstream of the TSS homologous to human IL-6 were analyzed. They were suggested to influence its regulatory function (Online Supplemental Figures S1B and 2) when examined by EMSA with nuclear extracts isolated from nTregs of WT and CD8 −/− mice. A previously uncharacterized region within the mouse Il-6 promoter, most distinct from human IL-6 (Online Supplemental Figures S1B), was investigated (Il-6_4). Each potential binding site under study was numbered in accordance with the relative distance to the TSS, respectively (Il-6_1, Il-6_2, Il-6_3, and Il-6_4, see Online Supplemental Figure S2).
For the region closest to the TSS of Il-6 (Il-6_1), a number of complexes were identified with an overall higher binding in unstimulated and stimulated (phorbol 12-myristate 13acetate (PMA)/ionomycin) nuclear extracts from CD8 −/− nTregs compared to WT nTregs ( Figure 1A, lanes 3 and 4 vs. 1 vs. 2). To verify the predicted binding of NF-κB, competition experiments with a high affinity NF-κB consensus site (100-fold molar excess) [23,24] were performed, leading to the abrogation of all complexes (Online Supplemental Figure S3A, lane 3). A DNA/protein complex was confirmed to selectively contain the NF-κB subunit p65 (Online Supplemental Figure S3A, lane 7), while no binding of the p50 subunit was observed (lane 6). A novel protein/DNA interaction of transcription factors of the specificity protein (SP) family including SP1 and SP3 in the nuclear environment of nTregs was identified. Their binding to Il-6_1 was confirmed by competition experiments with a SP consensus site [25,26] and using specific antibodies (Online Supplemental Figure S3A, lanes 4 and 8-11). Overall, the locus closest to the TSS of Il-6 showed greater binding of p65 and SP family members in CD8 −/− nTregs compared to WT nTregs before and after stimulation. Transcription factor binding to promoter regions within the Il-6 gene is increased in nTregs from CD8 −/− mice compared to wild-type (WT) nTregs. (A-D). EMSA analyses of four different Il-6 promoter loci upstream of the transcription start site of the Il-6 gene were performed, including (A) Il-6_1, (B) Il-6_2, (C) Il-6_3, and (D) Il-6_4 using nuclear extract isolated from nTregs from WT or CD8 −/− (KO) mice (5 µg) cultured for 3 h either left in medium (−) or stimulated (+) with PMA/ionomycin (50 ng/mL/1 µM).

Three Loci in the Il-6 Promoter Revealed Increased Transcription Factor Binding
Signaling for c-Fos, USF1, and USF2 in nTregs from CD8 −/− Mice Stronger transcription factor binding patterns to different sequence regions within the Il-6 promoter were present for three additional potential regulatory elements in unstimulated and stimulated CD8 −/− nTregs compared to WT nTregs ( Figure 1B-D). In contrast to the in silico prediction of C/EBP binding to Il-6_2 (Online Supplemental Figure S2,

Three Loci in the Il-6 Promoter Revealed Increased Transcription Factor Binding
Signaling for c-Fos, USF1, and USF2 in nTregs from CD8 −/− Mice Stronger transcription factor binding patterns to different sequence regions within the Il-6 promoter were present for three additional potential regulatory elements in unstimulated and stimulated CD8 −/− nTregs compared to WT nTregs ( Figure 1B-D). In contrast to the in silico prediction of C/EBP binding to Il-6_2 (Online Supplemental Figure S2, Online Supplemental Table S1), a supershift for c-Fos, USF1, and USF2 using specific antibodies (Online Supplemental Figure S3B, lanes 6, 8, and 9) was seen with a stronger complex formation signal using nuclear extracts from unstimulated nTregs from CD8 −/− mice ( Figure 1B, lanes 1 vs. 3). The specific binding of these transcription factors was confirmed by competition experiments with an AP1 [26] and USF [27] consensus site (Online Supplemental Figure S3B, lanes 3 and 4).
Increased protein/DNA binding in nTregs from CD8 −/− mice in contrast to WT nTregs before and after stimulation was also seen for Il-6_3 ( Figure 1C, lanes 3 and 4 vs. 1 and 2). As suggested by the prediction model (Online Supplemental Figure S2, Online Supplemental Table S1), binding of AP1 to Il-6_3 was verified but only for the c-Fos subunit (Online Supplemental Figure S3C, lanes 3 and 6), independently of c-Jun (Online Supplemental Figure S3C, lane 7). No specific complex formation of USF1 and USF2 to this Il-6 promoter region was detected (Online Supplemental Figure S3C, lanes 8 and 9).
Similarly, the transcription factors involved as potential regulatory elements most distant to the TSS of Il-6 (Il-6_4) included c-Fos, USF1, and USF2 (Online Supplemental Figure S3D, lanes 6, 8, and 9). The DNA/protein complex formation intensity significantly increased using nuclear extracts from CD8 −/− compared to WT nTregs ( Figure 1D, lanes 3 and 4 vs. 1 and 2). Taken together, using nuclear extracts from nTregs of CD8 −/− mice revealed an elevated transcription factor binding signal for all four Il-6 DNA sequence regions under study ( Figure 1A-D). This may contribute to the previously observed higher IL-6 levels in nTregs from these mice [14].

Defined Regions Are Vital for LPS-and TNFα-Mediated Il-6 Promoter Activation
To determine whether these transcription factors acted as regulatory elements of Il-6, plasmids containing the firefly luciferase gene under the transcriptional control of the Il-6 promoter were generated (Online Supplemental Figure S1B). We first confirmed that the Il-6 promoter activity was highly inducible in response to LPS or TNFα stimulation in a mouse fibroblast cell line (NIH3T3 cells, Figure 2A). Each of the binding sites analyzed by EMSA was then individually depleted by inserting a point-mutation. At baseline, the Il-6 promoter activity significantly increased after disruption of the binding sites for p65 and members of the SP family (Il-6_1, Online Supplemental Figure S4A). This introduced the full loss of the induction of the Il-6 promoter in response to LPS and TNFα stimulation (p < 0.01, Figure 2A, Online Supplemental Figure S4A).
The mutation of the transcription factor binding sites for c-Fos and USF1/2 within Il-6_2 led to significantly lower baseline promoter activity compared to the Il-6 wild-type luciferase construct (p < 0.01, Online Supplemental Figure S4A). When treated with LPS or TNFα, the Il-6 promoter activity was still inducible, albeit to a lower degree ( Figure 2A).
The depletion of the c-Fos/USF binding site in Il-6_3 did not influence baseline Il-6 activity (Online Supplemental Figure S4A), but resulted in a loss of induction after stimulation with LPS or TNFα ( Figure 2A).
In contrast, the independent depletions of the transcription factor binding sites including c-Fos and USF1/2 most upstream of the Il-6 TSS (Il-6_4) had no discernable effect on the overall luciferase signal (Online Supplemental Figure S4A, Figure 2A), suggesting that this region may not be relevant for the transcriptional regulation of the Il-6 promoter at baseline. However, the region furthest from to the TSS had the capacity to react to extrinsic stimuli including LPS and TNFα, mainly mediated through USF1/2 as the depletion led to a significantly lower inducibility of the Il-6 promoter (Online Supplemental Figure S4A). Overall, each of the loci within the Il-6 promoter acted as a transcriptional regulator with distinct binding signals of specific transcription factors including NF-κB, c-Fos, and members of the SP and USF family, thus potentially promoting the previously observed elevated IL-6 levels in nTregs from CD8 −/− mice in comparison to WT nTregs [14].
the Il-6 promoter activity was highly inducible in response to LPS or TNFα stimulation in a mouse fibroblast cell line (NIH3T3 cells, Figure 2A). Each of the binding sites analyzed by EMSA was then individually depleted by inserting a point-mutation. At baseline, the Il-6 promoter activity significantly increased after disruption of the binding sites for p65 and members of the SP family (Il-6_1, Online Supplemental Figure S4A). This introduced the full loss of the induction of the Il-6 promoter in response to LPS and TNFα stimulation (p < 0.01, Figure 2A, Online Supplemental Figure S4A).

Figure 2.
Transcription factor binding sites of NF-κB, c-Fos, and members of the SP and USF family as well as a novel GT repeat affect Il-6 promoter activity. (A) NIH3T3 cells were transiently transfected with the empty pGL4 luciferase vector (pGL4 empty) or with different reporter constructs of the Il-6 promoter (1138 bp). In the construct Il-6_1 (NF-κB)depl., the NF-κB binding site as identified by EMSA within Il-6_1 was depleted by site-directed mutagenesis. Correspondingly, the c-Fos binding site within Il-6_2 (Il-6_2(c-Fos)depl.), the c-Fos binding site within Il-6_3 (Il-6_3(c-Fos)depl.), the USF binding site within Il-6_4 (Il-6_4(USF)depl.), and the c-Fos binding site within Il-6_4 (Il-6_4(c-Fos)depl.) were mutated (4-5 independent experiments, 2 technical replicates). Cells were left in medium or stimulated with 10 ng/mL LPS or 0.5 ng/mL TNFα. (B) Il-6 luciferase promoter constructs were generated using DNA isolated from WT or CD8 −/− nTregs. To determine the frequency of a GT short tandem repeat in nTregs of both strains, a total of 104 clones (WT = 51, CD8 −/− = 53) were sequenced. The percentage of each of the detected GT STRs in WT or CD8 −/− nTregs is depicted (C) NIH3T3 cells were transiently transfected with the empty pGL4 luciferase vector or with reporter constructs of the Il-6 promoter region containing 22 or 23 GT repeats. Cells were left in the medium or stimulated with 10 ng/mL LPS or 1 ng/mL TNFα. For all luciferase experiments, luciferase activity was normalized for transfection efficiency using the control plasmid pRL-TK. The relative luciferase activity is presented in relative light units (RLU). Paired, two tailed students t-test was performed. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001; p-value comparing different stimulation conditions: ### p ≤ 0.001.

A Novel Short Tandem Repeat (STR) Influences Mouse Il-6 Promoter Activity
Within 500 bp upstream of the Il-6 TSS (Online Supplemental Figure S2), we identified a novel variable STR (microsatellite) in nTregs ranging from 20 to 24 GT repeats in length. Following sequencing of 104 Il-6 luciferase clones originating from nTregs from WT (n = 51 clones) or CD8 −/− (n = 53 clones) mice, the STR with 23 GT repeats was most common in both strains, with 66% and 43%, respectively ( Figure 2B). This repeat was found to occur with a 3.5-fold increase in CD8 −/− nTregs compared to the 22 GT repeat while in nTregs from WT mice, only a marginal difference between these two STRs was detectable (Online Supplemental Figure S4B). The luciferase construct containing the 23 GT STR resulted in a significantly (p < 0.01) increased signal under the control of the Il-6 promoter in unstimulated cells in comparison to the plasmid with 22 GT repeats ( Figure 2C). Furthermore, the Il-6 promoter signal was significantly higher in response to LPS and TNFα stimulation in the presence of 23 compared to 22 GT repeats ( Figure 2C). These data suggest that the GT STR was associated with the transcriptional regulation of the Il-6 promoter, providing a novel transcriptional mechanism explaining the observed increases in Il-6 expression and the consequent failure of nTregs from CD8 −/− mice to suppress lung allergic responses [14].

Discussion
In models of experimental asthma, the plasticity and loss of suppressive activities of nTregs from CD8 −/− mice compared to WT mice was largely dependent on levels of IL-6 production [14,28]. The underlying molecular mechanisms controlling Il-6 in nTregs have not been characterized to date. Here, we show that Il-6 in nTregs from WT and CD8 −/− mice is regulated through alterations of the binding of a number of transcription factors which was related to the response of Il-6 induction to external stimuli. Moreover, a novel STR was identified in the Il-6 promoter, significantly influencing its activity. These studies emphasized that differential expression of IL-6 is complex and is mediated by several cis-acting regulatory elements, thus contributing to the distinct role of nTreg function.
IL-6 participates in a broad range of biological events and has been shown to engage in key processes to promote inflammatory diseases [1,3]. In asthmatics, IL-6 levels were significantly elevated in plasma [7], bronchoalveolar lavage fluid [8], and sputum [5, 6,9] and correlated with disease severity [5][6][7]. IL-6 production was also elevated by viral infections and obesity [29][30][31], two important comorbid factors resulting in asthma exacerbations and severity. In patients with adult-onset asthma, elevated IL-6 was associated with high-dose inhaled corticosteroid use, systemic inflammation, and was linked to poor asthma control [32]. Outside of its role during disease, infection or stress, IL-6 is present at low expression levels [33]. Due to its many pleiotropic effects, IL-6 production is therefore tightly controlled. This complex regulation is cell type-specific and can vary in the same cell type depending on the stimuli [34]. As a result, it is essential that the transcriptional activities of IL-6 are studied in the distinct nuclear environment. In order to better understand the transcriptional regulation of Il-6, we performed transcription factor binding analyses in nTregs from WT and CD8 −/− mice as well as promoter mapping. As summarized in Figure 3, a total of four sites were studied, some of which have previously been shown to contribute to IL-6 regulation in different human and mouse cell lines [1,[18][19][20][21][22]. To our knowledge, this is the first study investigating regulatory mechanisms of Il-6 in primary nTregs comparing different mouse strains. Of note, across all four loci within the Il-6 promoter, increased occupation of transcription factor binding in nTregs from CD8 −/− compared to WT mice was seen at baseline and was enhanced in response to PMA and ionomycin, stimuli leading to the production of a variety of cytokines and a strong T cell activation [35]. The elevated binding signal using nuclear extracts from nTregs from CD8 −/− could be the result of overall increased binding affinity or a higher abundance of these transcription factors in the nucleus of CD8 −/− . Il-6 expression and the consequent failure of nTregs from CD8 −/− mice to suppress lung allergic responses [14].

Discussion
In models of experimental asthma, the plasticity and loss of suppressive activities of nTregs from CD8 −/− mice compared to WT mice was largely dependent on levels of IL-6 production [14,28]. The underlying molecular mechanisms controlling Il-6 in nTregs have not been characterized to date. Here, we show that Il-6 in nTregs from WT and CD8 −/− mice is regulated through alterations of the binding of a number of transcription factors which was related to the response of Il-6 induction to external stimuli. Moreover, a novel STR was identified in the Il-6 promoter, significantly influencing its activity. These studies emphasized that differential expression of IL-6 is complex and is mediated by several cisacting regulatory elements, thus contributing to the distinct role of nTreg function.
IL-6 participates in a broad range of biological events and has been shown to engage in key processes to promote inflammatory diseases [1,3]. In asthmatics, IL-6 levels were significantly elevated in plasma [7], bronchoalveolar lavage fluid [8], and sputum [5, 6,9] and correlated with disease severity [5][6][7]. IL-6 production was also elevated by viral infections and obesity [29][30][31], two important comorbid factors resulting in asthma exacerbations and severity. In patients with adult-onset asthma, elevated IL-6 was associated with high-dose inhaled corticosteroid use, systemic inflammation, and was linked to poor asthma control [32]. Outside of its role during disease, infection or stress, IL-6 is present at low expression levels [33]. Due to its many pleiotropic effects, IL-6 production is therefore tightly controlled. This complex regulation is cell type-specific and can vary in the same cell type depending on the stimuli [34]. As a result, it is essential that the transcriptional activities of IL-6 are studied in the distinct nuclear environment. In order to better understand the transcriptional regulation of Il-6, we performed transcription factor binding analyses in nTregs from WT and CD8 −/− mice as well as promoter mapping. As summarized in Figure 3, a total of four sites were studied, some of which have previously been shown to contribute to IL-6 regulation in different human and mouse cell lines [1,[18][19][20][21][22]. To our knowledge, this is the first study investigating regulatory mechanisms of Il-6 in primary nTregs comparing different mouse strains. Of note, across all four loci within the Il-6 promoter, increased occupation of transcription factor binding in nTregs from CD8 −/− compared to WT mice was seen at baseline and was enhanced in response to PMA and ionomycin, stimuli leading to the production of a variety of cytokines and a strong T cell activation [35]. The elevated binding signal using nuclear extracts from nTregs from CD8 −/− could be the result of overall increased binding affinity or a higher abundance of these transcription factors in the nucleus of CD8 −/− . Intact NF-κB binding in primary nTregs to a region closest to the TSS of Il-6 was essential for its transcriptional regulation, as previously seen in different cell lines [19][20][21]36]. Concomitantly, we identified a novel DNA/protein interaction of members of the SP family including SP1 and SP3 to the same binding site. This region was required for Intact NF-κB binding in primary nTregs to a region closest to the TSS of Il-6 was essential for its transcriptional regulation, as previously seen in different cell lines [19][20][21]36]. Concomitantly, we identified a novel DNA/protein interaction of members of the SP family including SP1 and SP3 to the same binding site. This region was required for significant induction of the Il-6 promoter as one potential mechanism contributing to the observed increase in IL-6 levels in nTregs from CD8 −/− compared to WT mice [14]. Given the limitation of the unfavorable ex vivo expansion properties of nTregs, these promoter expression studies were performed in NIH3T3 cells, a cell line of choice known to express IL-6 with a functional IL-6 signal transduction pathway [37][38][39]. In nTregs, SP1 not only influences Il-6 levels but also acts as a central element controlling the function of nTregs in mice due to its binding upstream of a Foxp3 enhancer element [40]. Further upstream within the Il-6 promoter, SP1 contributed to the transcriptional regulation in a mouse fibroblast B cell line, thus serving as a potential bridge to interact with NF-κB [36]. SP1 has been shown to bind with high affinity to specific NF-κB sites across the genome which were suggested to be mutually exclusive [41], while others observed a direct interaction between NF-κB and SP1 in a lung epithelial cell line [42]. Our data in nTregs support the latter. As NF-κB is fundamental in the regulation of IL-6 in a number of different cell types, it is postulated that NF-κB is crucial for the constitutive expression of Il-6, while the interaction with other transcription factors including AP1 [41] or SP1/SP3 is part of a program to control precise patterns of Il-6 in a distinct cell type. Although the binding of C/EBP was relevant for the transcriptional regulation of IL-6 in cell lines [18,20,21,[43][44][45], for example in concert with the NF-κB binding site [21], the ability to bind C/EBP, at least in nTregs from mice, was not required.
In addition, c-Fos may play a vital role to promote Il-6 as c-Fos binding was observed to three cis-regulatory elements within the Il-6 promoter, one of which has previously not been investigated. C-Fos is a subunit of the AP-1 complex forming homodimers as well as heterodimers with c-Jun [46]. The phosphorylation of c-Jun is mediated through the activation of the c-Jun N-terminal kinase (JNK) which in turn enables the interaction of c-Fos JunB, JunD, and ATF leading to the AP-1 complex formation [47]. Earlier studies in mouse cell lines demonstrated an involvement in the transcriptional regulation of Il-6 for both AP-1 subunits, c-Fos and c-Jun [18,20]. In nTregs, the protein complex binding to the mouse Il-6 promoter consisted only of the c-Fos subunit. This was surprising as JNK2 was identified to be essential in the enhancement of lung allergic responses mediated through nTregs, as demonstrated by the complete absence of exacerbating asthma-like immunopathology in sensitized and challenged WT mice following transfer of nTregs from JNK2-deficient mice [28]. It is therefore suggested that in nTregs, other regions within the Il-6 promoter are controlled by c-Jun. The depletion of the c-Fos binding in Il-6_2 significantly decreased the promoter activity of Il-6 in unstimulated cells, which was not observed for Il-6_3 and Il-6_4, suggesting that the transcriptional regulation of Il-6 mediated through c-Fos binding closest to the transcription starts was more relevant for the baseline Il-6 expression. Concomitant binding of c-Fos with members of the USF family (USF1/USF2) was detected for two regulatory elements within Il-6. USF is ubiquitously expressed and is involved in the transcription of a wide variety of genes, including the asthma susceptibility genes ORMDL sphingolipid biosynthesis regulator 3 (ORMDL3) [48] and plasminogen activator inhibitor 1 [49]. USF family members are part of helix-loop-helix proteins capable of interacting with enhancer box elements regulating target promoters [50,51]. For the binding site Il-6_4, the in silico model predicted an enhancer box binding region for which a protein/DNA complex formation of USF1/USF2 together with c-Fos was detected.
Based on our observations, the transcription factors NF-κB, c-Fos, SP1/SP3 and USF1/USF2 were found in a highly activated state, specifically in nTregs from CD8 −/− mice. Disrupting the binding sites within each of the four regulatory regions under study significantly decreased the Il-6 promoter activity in response to LPS and TNFα. LPS stimulates cells through Toll like receptor 4 signaling [52] mimicking an inflammatory response necessary for the activation of Il-6 in nTregs [53]. TNFα is a pro-inflammatory cytokine that can downregulate the suppressive capacity of nTregs [54], which in turn leads to increased IL-6 levels. In contrast to USF1/USF2, mutating the cFos binding site within Il-6_4 showed little effect on the ability to activate the Il-6 promoter, suggesting that the presence of this c-Fos regulatory site is not necessarily required. Overall, we believe that a complex interaction of these transcription factors is sufficient at basal level, but most importantly to the induction of IL-6, specifically in nTregs from CD8 −/− mice [14].
We discovered a novel STR at variable length in nTregs from WT and CD8 −/− mice upstream of the TSS of Il-6. Due to their association with human disease including asthma and atopic disease [55][56][57], the biological contribution of STRs has been studied intensively, leading to new paradigms to improve our current understanding of their effect on genome structure and function [58]. GT microsatellites are the second most common simple STR in the human genome within 5 kb of the TSS [55]. The prevalence of STRs including GT repeats specifically in regulatory regions, as seen in different organisms, suggests that repeat variation might be a common, evolutionarily conserved, mechanism regulating gene expression, thus supporting our findings. In contrast to the function of STRs in the predisposition to human disease, their role in disease-related mouse models including experimental asthma have to date been less well-characterized. Although the IL-6 promoter region between human and mouse is homologous and has some overlapping regulatory elements in both species, this microsatellite was not present in the human IL-6 promoter. The degree of conservation significantly declines directly upstream of this repeat suggesting the potential for evolutionary adaptation in humans as previously shown for other tandem repeats [59]. Noteworthy, within the same region of the human IL-6 promoter, a base composition of a polyA (n = 8) followed by a polyT (n = 13) was present. It remains to be determined if this is a repetitive element of variable length with a mechanistic effect on the regulation of IL-6 gene expression, thus contributing to human disease.
Amongst the detected number of repeats, 22 and 23 GT repeats were most frequent in nTregs from WT and CD8 −/− mice. Of note, the variation in length of this GT repeat significantly affected the promoter activity of Il-6. A similar effect has been described for the STAT6 gene, critical to the Th2 cytokine signaling cascade [56]. A GT repeat influencing the regulation of the promoter activity of STAT6 was associated with susceptibility to atopic asthma and total serum IgE levels. Given that the STR with 23 GT repeats led to the highest induction of the Il-6 luciferase signal without or with stimulation and was most commonly found in nTregs from CD8 −/− in comparison to nTregs from WT mice, we believe that this microsatellite may be a novel mechanism contributing to the increased IL-6 production observed in CD8 −/− nTregs. One potential model to additionally explain our findings is through alterations of the chromatin architecture of Il-6, as seen for other disease-related STRs in close proximity to boundaries of CpG islands [58]. As this novel GT repeat is located only 10 bp upstream of a potential CpG island, this hypothesis opens up future studies aimed at elucidating the cause-and-effect relationship between STR, epigenetic changes, and the cis-regulation on the Il-6 promoter in nTregs from CD8 −/− compared to WT mice.
There is compelling evidence of a phenotypic and functional instability of nTregs [60][61][62] which, amongst other factors [28,[63][64][65], is related to shifted production of IL-6 specifically in nTregs from WT and CD8 −/− mice [14,28]. Suppressive nTregs were shown to be capable of converting in vivo into pathogenic cells, enhancing the full spectrum of lung allergic responses. To this end, we described underlying molecular mechanisms in Il-6 controlling this conversion. Transcription factors close to the TSS directly interacted with the Il-6 promoter in nTregs, influencing its activity, and responded to inflammatory stimuli. The binding signal of these transcription factors was distinct in nTregs from WT and CD8 −/− mice. The expansion of a newly identified STR in nTregs from CD8 −/− mice increased the transcriptional activity of Il-6. Taken together, these findings exhibit alternative regulation of Il-6 in nTregs with distinct functions contributing to their plasticity and ability to undergo conversion, thereby influencing lung allergic responses and may be other IL-6-related diseases impacted by the nTregs.

In Silico Transcription Factor Binding Analyses
The MATinspector software tool (Genomatrix Suite, Precigen Bioinformation, Munich, Germany, www.genomatrix.de, accessed on 23 March 2021) was used to predict putative binding sites for transcription factors within 500 bp upstream the translation start site of Il-6 by comparing the respective sequence with similarities of binding matrices from the database. Briefly, using this software tool, an index was calculated, representing the essential binding regions for the respective binding site. A matrix was defined as a selective description of DNA patterns to bind transcription factors attributed to a database. The matrix similarity evaluated the matches of all necessary bases of the binding sequence of a transcription factor to the region of interest and was considered relevant for values ≥0.8. Each matrix contained a core sequence which was defined as the highest conserved positions of the matrix (usually 4 bp) reflected by the core similarity with a maximum score of 1.0.

IL-6 Reporter Constructs
DNA was isolated from nTregs from WT or CD8 −/− mice. To study the promoter activity of Il-6, 1138 bp upstream of the translation start site of Il-6 (Online Supplemental Figure S2) was cloned into the pGL4.10 basic vector (Promega, Madison, WI) using the Gibson Assembly Cloning kit according to manufacturer's protocol (Thermo Fisher Scientific, Waltham, MA, USA). A total of 51 colonies for WT and 53 for the CD8 −/− strain were screened by Sanger sequencing to delineate the distribution of a GT STR in the Il-6 promoter. Following the manufacturer's protocol, site-directed mutagenesis (Agilent Technologies, Santa Clara, CA, USA) of the pGL4 Il-6 construct was performed to deplete potential transcription factor binding sites in the Il-6 promoter region. The sequence was confirmed by re-sequencing of all Il-6 promoter constructs. Primers for amplification of the Il-6 promoter region, for Sanger sequencing, and for the site-directed mutagenesis are depicted in the Online Supplemental Table S3.

Luciferase Assay
NIH3T3 mouse fibroblast cells were seeded in 96-well plates at a density of 7 × 10 4 per well. The next day, cells were transfected with 750 ng pGL4.10 plasmids (Promega, Madison, WI, USA) expressing the luciferase gene under the control of the Il-6 promoter containing the respective wild-type, point mutation at one of the four analyzed transcription factor binding sites, or different lengths of the GT repeat clones together with 10 ng of the pRL-TK Renilla reporter plasmid (Promega, Madison, WI, USA) for normalization of transfection efficiency and cell viability. Lipofectamine 2000 was used as the transfection reagent according to the manufacturer's protocol (Thermo Fisher Scientific, Waltham, MA, USA). Directly after transfection, the medium was exchanged with a medium containing pure medium (unstimulated) or 10 ng/mL LPS, 0.5 ng/mL TNFα, 1 ng/mL TNFα. Fourteen hours after stimulation, cells were washed in PBS and lysed in 1x passive lysis buffer (Promega, Madison, WI, USA). Dual luciferase assay was performed according to manufacturer's protocol (Promega, Madison, WI, USA). Luciferase and renilla signals were measured using a Synergy HT luminometer (BioTek, Winooski, VT, USA) and quantified as relative light units (RLU). Experiments were conducted independently four to five times with two technical replicates for each construct. A paired, two-tailed Student's t-test was performed to identify significant differences between groups (p ≤ 0.05).

Conclusions
Changes in transcription factor binding and a novel variable dinucleotide repeat of variable length impact the regulation of Il-6 contributing to nTreg plasticity important for IL-6-related disease with impaired nTreg function. Data Availability Statement: The data supporting the findings of this study are available from the corresponding author upon request.

Acknowledgments:
The assistance of Diana Nabighian in preparation of this manuscript is gratefully acknowledged.

Conflicts of Interest:
The authors declare no conflict of interest.