Trichloroethylene metabolite modulates DNA methylation-dependent gene expression in Th1-polarized CD4+ T cells from autoimmune-prone mice

Abstract

Trichloroethylene (TCE) is an industrial solvent and widespread environmental contaminant associated with CD4+ T-cell activation and autoimmune disease. Prior studies showed that exposure to TCE in the drinking water of autoimmune-prone mice expanded effector/memory CD4⁺ T cells with an interferon-γ (IFN-γ)-secreting Th1-like phenotype. However, very little is known how TCE exposure skews CD4⁺ T cells towards this pro-inflammatory Th1 subset. As observed previously, TCE exposure was associated with hypermethylation of regions of the genome related to transcriptional repression in purified effector/memory CD4 T cells. We hypothesized that TCE modulates transcriptional and/or epigenetic programming of CD4⁺ T cells as they differentiate from a naive to effector phenotype. In the current study, purified naive CD4 T cells from both male and female autoimmune-prone MRL/MpJ mice were activated ex vivo and polarized towards a Th1 subset for 4 days in the presence or absence of the oxidative metabolite of TCE, trichloroacetaldehyde hydrate (TCAH) in vitro. An RNA-seq assessment and reduced representation bisulfite sequencing for DNA methylation were conducted on Th1 cells or activated, non-polarized cells. The results demonstrated TCAH’s ability to regulate key genes involved in the immune response and autoimmunity, including Ifng, by altering the level of DNA methylation at the gene promoter. Intriguing sex differences were observed and for the most part, the effects were more robust in females compared to males. In conclusion, TCE via TCAH epigenetically regulates gene expression in CD4⁺ T cells. These results may have implications for mechanistic understanding or future therapeutics for autoimmunity.

Trichloroethylene (TCE) is a halocarbon used as an industrial solvent and metal degreasing agent. This pervasive pollutant has contaminated many water systems in the United States and is an occupational hazard. There are well-documented epidemiological associations between TCE exposure and autoimmunity and other immune-mediated inflammatory diseases linked to occupational and environmental exposures in human populations (Huang et al., 2014; Nakajima et al., 2022; Purdue et al., 2022). However, the mechanism of how TCE promotes autoimmune responses is not known. Experimental studies have shown that TCE is immunotoxic to CD4⁺ T cells, and exposure to TCE in the drinking water of autoimmune-prone mice expanded effector/memory (CD44^hi/CD62L^lo) CD4⁺ T cells with an interferon-γ (IFN-γ)-secreting pro-inflammatory Th1-like phenotype (Banerjee et al., 2020; Griffin et al., 2000; Li et al., 2018). Because autoimmune responses are mostly CD4⁺ T-cell driven, we hypothesized that TCE directly modulates transcriptional and/or epigenetic programming of CD4⁺ T cells as they differentiate from a naive to effector phenotype.

Under normal circumstances, CD4⁺ T cells differentiate into subsets to deal with pathogenic threats or resolution of inflammation. Type 1 T helper (Th1) cells are among the CD4⁺ T-cell effector subsets that provide protection from pathogens and are characterized by the production of high levels of IFN-γ among several other cytokines and factors, and can be distinguished from other Th subsets by the transcription factor T-bet (Fang et al., 2022). Uncontrolled Th1 expansion is involved in several autoimmune disorders, such as type 1 diabetes mellitus, multiple sclerosis, rheumatoid arthritis, and Crohn's disease (Akbar et al., 2023; Ayass et al., 2023; Wang et al., 2023a). It has yet to be fully explored whether TCE-induced CD4⁺ T-cell changes are regulated by transcription and/or DNA methylation, and whether TCE specifically promotes Th1 cell polarization or more generalized non-specific activation.

Previously, a genome-wide DNA methylation assessment in our lab using reduced representation bisulfite sequencing (RRBS) in purified effector/memory CD4⁺ cells isolated from female autoimmune-prone MRL/MpJ mice exposed to TCE in the drinking water revealed that TCE-altered DNA methylation. The regions that were most affected were those involved in transcriptional repression (Byrum et al., 2019; Gilbert et al., 2016, 2017). Because the CD4⁺ T cells analyzed in these studies were purified effector cells (ie, CD44^hi/CD62L^lo/IFN-γ⁺) that had already differentiated in vivo, hypermethylation of these regions related to transcriptional repression as a mechanism of downregulating gene expression would be predicted. Unfortunately, these experiments did not have complimentary gene expression data. Since both DNA methylation and gene expression changes could explain the dynamic process of T-cell differentiation, assessing whether TCE alters differentiation towards the Th1 subset might be more revealing in providing insight into mechanism.

Precise CD4⁺ T-cell differentiation events are time-dependent and difficult to elucidate in vivo. Using an in vitro system would circumvent several issues pertaining to capturing the often biphasic/non-linear response of cytokine production and any genomic patterns by effector/memory CD4⁺ T cells as they differentiate in vivo. Furthermore, TCE is a highly volatile chemical with poor water solubility and difficult to work with or maintain a consistent dose in vitro. TCE’s metabolism involves the generation of several metabolites. Upon exposure in vivo, lipophilic TCE is quickly absorbed regardless of exposure route (eg, oral, inhalation, dermal), and at low concentrations removed from the blood after a single pass through the liver. Pharmacokinetic studies have shown that TCE levels in the liver peak at less than 20 min before it undergoes oxidative metabolism mediated by Cytochrome P450 Family 2 Subfamily E Member 1 (CYP2E1) (Cichocki et al., 2016). While TCE can be metabolized by a glutathione-dependent pathway in the kidney (Lash et al., 2006), the majority of TCE absorbed into the circulation is metabolized by the oxidative pathway. CYP2E1 converts TCE to trichloroacetaldehyde or chloral whichin solution is in equilibrium with trichloroacetaldehyde hydrate (TCAH). For the sake of simplicity, we will call this mixture, TCAH. A simplified figure depicting oxidative metabolism of TCAH is shown in Figure 1. TCAH is further metabolized by either of two pathways: oxidation to trichloroacetic acid (TCA) catalyzed by aldehyde dehydrogenase or reduction mediated by alcohol dehydrogenase to trichloroethanol which is excreted in the urine as the alcohol glucuronide. This latter pathway is considered to be reversible back to TCAH. Thus, the level of TCAH depends on the activity of several enzymes (Ramdhan et al., 2008).

Figure 1.

Metabolism of trichloroethylene to TCAH and subsequent disposition into excreted metabolites, TCA and TCOH. Abbreviations: TCA, trichloroacetic acid; TCAH, trichloroacetaldehyde hydrate; TCOH, trichloroethanol.

Because of TCE’s complicated metabolism and the fact that CD4⁺ T cells do not have intrinsic CYP2E1 activity we used one of TCE’s water-soluble oxidative metabolites, TCAH as a surrogate for TCE in vitro. Previous studies have underscored the importance of oxidative metabolism in TCE-mediated CD4⁺ T-cell immunotoxicity. When CYP2E1 was inhibited by diallyl sulfide in MRL/MpJ mice exposed to TCE, CD4⁺ T-cell responses including proliferation and cytokine levels were suppressed (Griffin et al., 2000). Since TCAH is a known intermediate in the pathway of TCE metabolism that generates the metabolite, trichloroacetic acid or TCA, we studied whether TCAH or TCA when administered to the drinking water of MRL/MpJ mice for 4 weeks imparted CD4 T-cell affects similar to TCE (Blossom et al., 2004). While both TCA and TCAH inhibited apoptosis and enhanced cytokine levels from CD4⁺ T cells, it was not clear whether the effects of TCAH on CD4⁺ cells were due to TCAH itself or due to the TCA that may have resulted from further metabolism in vivo. In the same report, TCAH increased levels of serum autoantibodies whereas TCA did not. Thus, we shifted our focus on using TCAH as a mechanistic surrogate for TCE. In a follow-up study, exposure of MRL/MpJ mice to TCAH in the drinking water for 40 weeks accelerated the appearance of autoimmune tissue pathology commensurate with CD4⁺ alterations similar to that observed with TCE exposure (Blossom and Gilbert, 2006).

The approach in the current study was to determine the direct effects of TCAH on gene expression and DNA methylation patterns in polarized Th1 CD4⁺ T cells in vitro. Th1 cells were compared with anti-CD3/anti-CD28-activated CD4⁺ cells (activated, but non-polarized) to test whether TCAH altered gene expression or methylation patterns similar to those elicited in Th1 cells. While in general, females are more likely to develop autoimmune disease than males, and most of our prior work has been conducted in females, the incidence for many types of autoimmune diseases in males in the United States appears to be increasing (Dinse et al., 2020). Thus, male MRL/MpJ mice were included in this study for comparison purposes.

Materials and methods

 

Mice

This study was conducted at the University of New Mexico under and approved by the Institutional Animal Care and Use Committee and the University of New Mexico (protocol # 201094; approval date October 03, 2021). Eight-week-old male and female MRL//MpJ mice were purchased from Jackson Laboratories, Bar Harbor, ME, USA.

Reagents

TCAH (purity >99%) was obtained from Sigma, resuspended in deionized water as a stock solution, and diluted with 1× PBS. A fresh stock solution preparation was made before each experiment consisting of three experimental replicates in triplicate. All cells were cultured in Complete RPMI (C-RPMI) consisting of standard RPMI medium supplemented with 2 mM l-glutamine, 1 mM, nonessential amino acids, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 µg/ml streptomycin, 5 × 10⁵-M 2-ME, and 5% fetal calf serum (FCS). TCAH-treated cultures were used at a concentration of 0.5 mM where this dose was previously shown to alter T-cell function, and culturing primary CD4⁺ T cells with a range of TCAH up to 1.0 mM for 5 days did not significantly affect cell viability (Blossom and Gilbert, 2006).

CD4+ T-cell isolation and polarization experiments

Naive CD4⁺ T cells were isolated from spleen cell preparations using mouse-naive CD4⁺ T-cell isolation kits purchased from Miltinyi Biotec and purified according to the manufacturer’s instructions. All resulting naive CD4⁺ cells were resuspended in complete media, seeded into six-well plates at a density of 1 × 10⁶ cells/ml, and activated with immobilized anti-CD3 antibody (Biolegend; clone 145-2C11) and soluble anti-CD28 antibody (BioLegend; clone 145-2C11). Subsets of purified CD4⁺ cells differentiated toward Th1. Procedures for Th1 cell polarization were performed according to the manufacturer’s specifications from kits purchased from BioLegend. Anti-mouse IL-4, clone 11B11 (10 µg/ml), recombinant mouse IL-2 (5 ng/ml), and recombinant mouse IL-12 (10 ng/ml) were added to the cultures. On day 4, cells were harvested, and RNA and DNA were isolated for genome-wide transcriptomics using RNA-sequencing (RNA-seq) and genome-wide DNA methylation assessment using RRBS analysis, respectively. Cells from the cultures were assessed for cell viability and absolute cell counts using ViaStain AOPI Staining Solution and imaged with a Nexcelom Cellometer Auto 2000. To obtain an n = 3 per group, we conducted three independent experiments. For each experiment, purified CD4 cells were pooled from 2 age- and sex-matched mice in each group and cultured in triplicate.

Targeted gene expression analysis

To determine gene expression changes in activated, non-polarized CD4⁺ T cells or Th1 CD4⁺ T cells, total RNA was isolated from cells as described and reverse-transcribed as described (Blossom et al., 2022). RNA concentrations were determined using a NanoDrop instrument (Thermo Scientific). cDNAs were prepared High Capacity cDNA Reverse Transcriptase kit (Applied Biosystems). Genes were quantified by qRT-PCR using predesignated TaqMan Gene Expression Assays (Thermo-Fisher). Samples were run in duplicate and averaged to obtain mean fold-change expression of Th1-related genes, Ifng (IFN-γ), Txb21 (T-bet), and Il10 (IL-10) and compared with the housekeeping control gene (β actin) using 2 delta delta Ct method.

IFN-γ ELISA procedure

Supernatants for cell culture were collected on either day 1 or day 4 post-isolation and plating. Supernatants were kept frozen at −80°C until used for assays. IFN-γ levels in the cell supernatants were determined using the R&D Systems Mouse IFN-gamma Quantikine ELISA kit (# MIF00). Cell supernatants were diluted 1:100 prior to the ELISA procedure. The plate was imaged on a Molecular Devices SPECTRAmax 340PC plate reader and 450 nm. IFN-γ levels (ng/ml) were quantified based on a standard curve.

Extraction of genomic DNA and RNA

Genomic DNA (gDNA) and RNA samples were extracted using the AllPrep DNA/RNA mini kit (Qiagen, Hilden, Germany). The extracted gDNA and RNA concentration were determined with Qubit fluorometers (Thermo Scientific, CA, USA) under a broad-range assay setup. The integrity of extracted genomic materials was confirmed with TapeStation (Agilent Technologies, CA, USA).

Statistical analysis

Statistical analyses were performed using GraphPad Prism 10.0 (GraphPad Software, La Jolla, CA). Data are presented as mean (SD). Data were evaluated by one-way ANOVA followed by a Tukey’s post hoc test. T tests were used to compare between two groups. p < .05 was considered statistically significant.

Transcriptomic analysis

RNA-seq was performed to determine differential gene expression among the different treatment groups. RNA samples obtained from purified CD4 cells isolated from male or female MRL//MpJ mice that were either activated (anti-CD3/anti-CD28) or Th1 polarized for 4 days were sequenced with Novagene (Sacramento, CA) using Illumina Novaseq platform. Differential expression of log2 fold-change (FC) > 1 between treatment and control groups was considered significant (p < .05). Data are provided in supplementary tables that depict the top 20 differentially expressed genes sorted by FC for each group (p < .05 and p < .1 adjusted p-value cut off). A complete data file of differentially expressed genes is available upon request.

DNA methylation analysis by RRBS

Bioinformatics analysis

RRBS libraries were prepared by following our previously published protocol (Choudhury and Walker, 2018). Briefly, 500 ng of gDNA were digested overnight with MspI, followed by end-repair, A-tailing, adapter-ligation, size-selection, bisulfite conversion, and enrichment with index primers. RRBS libraries were multiplexed and sequenced using 75-bp single-end reads at the UAMS Genomics core facility as described (Byrum et al., 2019). The nf-core nextflow pipeline methylseq was used to pre-process raw data from FastQ inputs, align the reads, and perform quality control (Ewels et al., 2020). Sequencing reads were trimmed to remove adapters using TrimGalore!, reads were aligned using bwa-meth to the mouse genome (GRCm39), duplicates were marked with Picard MarkDuplicates, methylation calls were extracted with MethylDackel, and alignment quality control evaluated with Qualimap. The coverage files were then analyzed further using edgeR Bioconductor package as previously described (Robinson et al., 2010) and following guidance from Chen et al. (2017). Linear models are then used to fit the total read count (methylated plus unmethylated or M + U) at each genomic locus and methylated reads are modeled indirectly as an over-dispersed binomial distribution. Differentially methylated sites and regions (DMRs) were assessed by generalized linear models with likelihood ratio tests using edgeR generalized linear model likelihood ratio test (glmLRT). The p-values were corrected using the false discovery rate.

Results

 

TCAH treatment of naive CD4⁺ T cells in vitro enhances expression of core response genes during Th1 polarization

Naive mouse CD4s skewed toward Th1 effector phenotypes in vitro upregulate many early core response genes (eg, Ifng and its master transcriptional regulator, T-bet) detectable as early as day 1 after activation. These genes persist and overlap with later-occurring turnover genes (eg, Il10) that occur at around day 4 (van den Ham et al., 2013). We sought to determine whether TCAH altered the expression of core response/turnover genes using qRT-PCR. As expected, in Figures 2A and 2B, Th1 core response genes were upregulated early (day 1) and persisted (day 4). Tbx21, the gene that encodes for T-bet, sustained a robust expression over time (Figure 2A) in TCAH-treated cultures. However, in cells not treated with TCAH, this expression was decreased by day 4 compared to day 1 controls (p = .003) and TCAH-treated cells at day 4 (p = .001). A similar pattern was observed with Ifng expression. Ifng was maintained over time, especially in TCAH-treated Th1 cells, compared to non-treated Th1 cells (p < .0001).

Figure 2.

TCAH enhanced gene expression in Th1 cells. CD4 T cells were purified from spleens of female MRL/MpJ mice using magnetic bead separation and incubated with reagents designed to polarize the cells toward Th1 as described in Materials and Methods. On days 1 and 4, cells were harvested and processed for RNA and gene expression by qRT-PCR. Numbers in the bar graphs represent mean (SD) fold-change values relative to unstimulated CD4 cells and normalized by b-actin housekeeping gene. Data were analyzed by Student’s t test with Welch correction as described in methods. Shown in the graphs are p values indicating statistical significance (**p < .05 and ***p < .001) after Welch correction comparing the means in each group.

In addition to gene expression, supernatants were collected to measure the secreted IFN-γ in the supernatants on days 1 and 4. Results confirmed that as suspected, IFN-γ secretion substantially increased by 69% (mean of 9.9 vs 14.3 ng/ml) in control cultures and by 79% (mean of 14.4 vs 18.1 ng/ml) in cultures treated with TCAH confirming that a majority of the cultured cells (eg, >50%) incubated with Th1 polarizing agents were polyclonally activated and skewed towards Th1 as described (Aso et al., 2023). TCAH significantly increased these levels at the protein level similar to what was observed for gene expression (Supplementary Figure 1). Although TCAH did not significantly alter IL-10 expression compared to controls, Figure 2C confirmed that IL-10 served as a late-occurring turnover gene in Th1 polarized cells. Based on these experiments, day 4 was selected as the optimal time point for capturing overlapping expression of early and late-occurring genes involved in Th1 cell differentiation without a significant change in cell viability, absolute number, or proliferation among the groups (Supplementary Figure 2). This result also confirmed the utility of using TCAH as a TCE surrogate for modulating Th cell gene expression and differentiation in vitro.

TCAH alters gene expression of polarized Th1 cells

To more comprehensively assess the transcriptomic profile, we conducted RNA-seq on day four polarized Th1 cells from male and female MRL/MpJ mice ± TCAH. Volcano plots revealed that females showed an even more robust response to TCAH treatment compared to males (Figs. 3A and 3C). In female Th1 cells treated with TCAH, there were 1124 significantly (p < .05) upregulated and differentially expressed genes. Several notable immune-related genes significantly increased by TCAH are annotated in Figure 3A. The statistically significant genes that were increased by TCAH were ranked according to Log2 fold-change (Supplementary Table 1) and included several chemokines, Ccl1 (6.25-fold), Xcl1 (4.00-fold), Ccl20 (3.62-fold), and Cxcl9 (2.62-fold). The pro-proliferative/anti-apoptotic functional cytokine, Il9 was also increased (2.72-fold). The cytokine Ifng that is expressed most predominantly by Th1 cells over other subsets (see Figure 2) was also significantly upregulated by 2.43-fold in female mice.

Figure 3.

A, Volcano plot representing FC for the differentially expressed genes in Th1 cell population of female MRL/MpJ mice treated with TCAH at 0.5 mM, compared to the age-matched nontreated mice (control). Genes involved in immune responses have been highlighted among the significantly (p < .05) upregulated (FC >1, log2; red dots) or downregulated (FC <1) clusters. B, Bubble map depicting the KEGG-annotated pathways enriched with (rich factor >0.5) differentially upregulated in the female MRL/MpJ mice. No pathways were predicted based on the differentially downregulated genes in the Th1 cells population of the female MRL mice. C, Volcano plot representing FC for the differentially expressed genes in Th1 cell population of male MRL/MpJ mice treated with TCAH at 0.5 mM, compared to the age-matched non-treated mice (control). Genes involved in immune responses have been highlighted among the significantly (p < .05) upregulated (FC >1, log2) or downregulated (FC <1) clusters. However, the differentially upregulated or downregulated genes were not enriched for any predicted KEGG-annotated pathways. D, The number of significantly differentially expressed genes from TCAH exposure within each sex are shown in males vs. females. Abbreviations: FC, fold change; TCAH, trichloroacetaldehyde hydrate.

Next, enrichment of the differentially expressed genes was predicted using KEGG-annotated pathways (Figure 3B). Genes upregulated in the Th1 cells from female mice (Figure 3B) were highly enriched for metabolic pathways and RNA transport. Using a Venn intersection analysis (Figure 3D), 21 upregulated genes and one downregulated gene were identified as being mutually expressed in both sexes (Supplementary Table 2). Ccl20 was the most highly expressed gene in male mice (3.64-fold). However, we could not detect enriched signaling Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways with differentially expressed and mutually inclusive genes in both sexes. Similarly, despite the number of significantly downregulated genes in females (n = 1030) (Supplementary Table 3), KEGG-annotated analysis did not reveal significant differences in pathways. In males, there were very few differentially upregulated or downregulated genes in response to TCAH (75 vs 31, respectively), and none of those genes were enriched for predicted KEGG-annotated pathways. Several important Th1 cytokines that were upregulated in females were not observed in males including the Th1 cytokine, Ifng (0.358-fold; p = .261). The top 20 differentially expressed genes in male mice are listed in Supplementary Table 4.

TCAH alters gene expression in non-polarized activated CD4s

In the absence of polarizing cytokines based on the pattern of gene expression observed in Th1 cells, based on our in vivo data with the parent compound, TCE, we predicted that TCAH would enhance the expression of Th1-related genes in naive CD4s that were treated sub-optimally with anti-CD3/CD28 for 4 days. As shown in Figures 4A, 4D, and 4F, TCAH treatment resulted in 855 significantly upregulated genes in 4-day activated CD4⁺ cells from female mice compared to only 195 genes in males. Several genes essential in the immune response and apoptosis are annotated. Interestingly, when the top 20 significantly upregulated genes were ranked by fold-change (Supplementary Table 5), 9 of those genes were also differentially upregulated in Th1 cells (Supplementary Table 1) including, chemokine ligand and receptors, Ccl1, Xcl2, Pf4, Ppbp, and Xcl1, cytokines, Ifng and Il1a, as well as Thrombospondin 1 (Thbs1), and Grp (Gastrin-releasing peptide).

Figure 4.

Differentially expressed genes in CD4+ T-cell population of female MRL/MpJ mice treated with TCAH at 0.5 mM, compared to the age-matched non-treated controls. Genes involved in immune responses have been highlighted among the significantly (p < .05) upregulated (FC >1, log2; red dots) or downregulated (FC <1; blue dots) clusters. Bubble map depicting the KEGG-annotated pathways enriched with (rich factor >0.5) differentially upregulated (B) or downregulated (C) genes in the female MRL mice. D, Volcano plot representing FC for the differentially expressed genes in CD4+ T-cell population of male MRL/MpJ mice treated with 0.5 mM TCAH, compared to the age-matched non-treated mice (control). Genes involved in immune responses have been highlighted among the significantly (p < .05) upregulated (FC >1, log2) or downregulated (FC <1) clusters. E, Differentially downregulated genes in male MRL/MpJ mice were enriched (rich factor >0.5) for several KEGG-annotated pathways. F, Using a Venn analysis, we identified 68 upregulated and 39 downregulated genes mutually inclusive to both sexes. Enrichment bar plots were prepared to predict the commonly impacted pathways containing differentially upregulated (G) or downregulated (H) genes in TCE treatment in the activated, non-polarized CD4+ T-cell population of both sexes. Abbreviations: FC, fold change; TCAH, trichloroacetaldehyde hydrate.

KEGG pathway analysis of the differentially expressed genes (Figs. 4B, 4C, and 4E) revealed that several relevant pathways were enriched in female mice corresponding with TCAH-mediated up- or downregulation of genes. Similar to Th1 cells, the most significantly upregulated genes by TCAH in female mice in the unpolarized subset were also enriched for metabolic pathways (Figure 4B). Some annotated pathways downregulated in TCAH-treated CD4 cells from female mice included T-cell signaling. The genes that were the most significantly downregulated in female CD4s relative to TCAH treatment are presented in Supplementary Table 6. Among the top 20 genes, only 5 were downregulated in Th1s and in activated/unpolarized CD4⁺ T cells including Tgm1, Gpld1, Plcd1, Adamtsl4, and the anti-inflammatory factor, tgfb3.

In male mice, there were a total of 195 upregulated and 144 downregulated genes that were statistically significant (Supplementary Tables 7 and 8). A subsequent KEGG analysis showed some statistically significant enriched pathways for differentially downregulated genes in male mice (Figure 4E). Several pathways were linked to neurological disorders including Parkinson disease. In contrast, enriched pathways predicted for the differentially upregulated genes in males were not statistically significant.

Among the significant differentially expressed genes, 68 upregulated and 39 downregulated genes were common in both sexes (Figure 4F; Supplementary Table 9). The chemokine, Ccl1 was highly differentially expressed in both males and females (4.81- and 4.64-fold, respectively). Some shared downregulated genes include the nuclear transcription factor, Pparg, the Th2 cytokine, Il4, and several histone-related genes (eg, H3c10, H2bc9, and H3c1). Data are shown in the enrichment bar plots for shared TCAH-mediated upregulated and downregulated genes. Interestingly, several gene families that were significantly upregulated in both sexes are involved in the regulation of the immune system and inflammatory response (Figure 4G). Notable pathways related to TCAH downregulated genes in both sexes (Figure 4H) were related to the responses associated with chromatin assembly and organization responses.

TCAH-induced differential DNA methylation assessment in Th1 cells and activated CD4⁺ T cells from female mice

We next investigated whether TCAH instigates epigenetic modifications that could regulate differential expression of the immunomodulatory genes. We observed a relatively lower fraction (2%, n = 17 304) of significant (p < .01) DMRs in polarized Th1 cell population from female mice compared to the non-TCAH-treated Th1 controls. We enumerated the significant DMRs belonging to the gene-promoters and CpG islands and observed that 67% (n = 276) DMRs were hypermethylated and 33% (n = 133) DMRs were hypomethylated at the promoters (Figure 5B). In contrast, we observed 87% (n = 10 232) of hypermethylated DMRs and 13% (n = 1604) hypomethylated DMRs at the CpG islands (Figure 5C). Next, we combined the median DNA-methylation at the DMRs in gene promoters and integrated it with gene expression data. We observed 18 hypomethylated genes upregulated with TCAH treatment, including selective immunomodulatory genes such as Ifng, Il9, and Ccl1 (Figure 4D and Supplementary Table 10). We observed 39 genes to be hypermethylated at the promoter and downregulated. This cluster includes Il33, which encodes for IL-33, a notable immunomodulatory cytokine involved in suppression of autoimmune responses (Janyga et al., 2023; Song et al., 2023; Wang et al., 2023b). We also observed genes that were hypomethylated at the promoter and downregulated (n = 14) or hypermethylated at the promoter and upregulated (n = 15).

Figure 5.

A, Manhattan plot displaying adjusted p values of the association between the differentially methylated regions (DMRs) and the effect of TCE on Th1 cells in female MRL mice. The line in the plot represents the FDR value (p < .01), and the dots above the line represent the signficant DMRs (beta value >10% between treated and control samples). B, Pie chart representing the 67% hypermethylated (n = 276) and 33% hypomethylated (n = 133) DMRs at the gene promoters. C, The DMRs that were hypermethylated (87%; n = 10 232) and hypomethylated (13%; n = 1604) were noted in the CpG island regions. D, Quadrant plot representing the correlcation between differential (log2) DNA-methylation at promoters and gene expression. There were clusters containing hypomethylated and upregulated genes (n = 18) and differentially hypermethylated and downregulated (n = 39) genes. Four critical immunomodulatory genes, Ccl1, Il9, adn Ifng were found to be among the hypomethylated and upregulated genes, while Il33 was observed among the hypermethylated and downregulated genes.

Next, we similarly investigated the association between TCAH treatment and DMRs in the activated but non-polarized CD4 cells from female mice. We identified 1.59% (n = 13 570) statistically significant (p < .01) DMRs (Figure 6A). Of these DMRs, we observed 37% (n = 112) hypermethylated DMRs and 63% (n = 202) hypomethylated DMRs to be at the gene-promoters (Figure 6B). In contrast, we noticed 13% (n = 1593) hypermethylated DMRs and 87% (n = 10 124) hypomethylated DMRs at the CpG-island (Figure 6C). Upon combining the methylation and expression data for the unpolarized activated CD4⁺ T cells, we observed 36 genes to be hypomethylated and upregulated, 43 genes to be hypomethylated and downregulated, 33 genes to be hypermethylated and upregulated, and 19 genes to be hypermethylated and downregulated (Figure 6D). We detected very few albeit important shared genes (n = 4) including Ifng and Ccl1 among the hypomethylated and upregulated genes in TCAH-treated CD4 cells.

Figure 6.

A, Manhattan plot displaying adjusted p values of the association between the differentially methylated regions (DMRs) and the effect of TCE on the CD4 T cells (activated and unpolarized) from female MRL/MpJK mice. The line in the plot represents the FDR values (p < .01), and the dots above the line represent the significant DMRs (beta-value >10% between control and treated samples). B, Pie chart representing the 37% hypermethylated (n = 122) and 63% hypomethylated (n = 202) DMRs at the gene promoters. C, Thirteen percent (n = 1593) DMRs were hypermethylated and 87% (n = 10 124) DMRs were hypomethylated at CpG islands. D, Quadrant plot representing the correlation between differential (log2) DNA-methylation at promoters and gene expression. Clusters containinig hypomethylated and upregulated (n = 36) and differentially hypermathylated and downregulated (n = 19) genes were noted. Two critical immunomodulatory genes, Ifng and Ccl1 were found to be among the hypomethylated and upregulated gene cluster.

Discussion

Despite a number of reports examining effects of in vivo exposure to TCE on the immune system, there is very little understanding of the mechanistic underpinnings that drive Th1 responses to promote autoimmunity. Several studies point to epigenetic factors including DNA methylation as potential drivers of these responses (Huang et al., 2021; Liu et al., 2022; Yue et al., 2022). As far as TCE exposure, we previously reported that TCE promotes epigenetic effects (genome-wide demethylation) in cervical (using HeLa cell line) and hepatocellular (using HepG2 cell line) carcinoma cells (Cui et al., 2016). Using single-molecule fluorescence microscopy and fluorescence correlation spectroscopy, we demonstrated that TCE-induced demethylation is correlated to the dissociation of DNA methyltransferase 3A oligomers from the heterochromatic foci. Considering the critical role of TCE in inducing autoimmune responses, it is important to evaluate the dependency between alterations in DNA methylation with the expression of genes related to autoimmunity. The relevance of exploring epigenetic mechanisms in the context of TCE exposure is underscored by recent findings that human occupational exposure to trichloroethylene was linked to alterations in DNA methylation in peripheral blood (Phillips et al., 2019). Intriguingly, these documented exposures were linked to DMRs relevant to autoimmunity and cancer and a significant hypomethylation in the TRIM68, a negative regulator of IFN-γ. Although purified immune cell populations were not assessed, this study provides human evidence that DNA methylation alterations play a role in TCE-mediated immunotoxicology.

One limitation of our previous work was that DNA methylation was studied either in the aforementioned transformed cell lines or in autoimmune-prone MRL/MpJ mice following in vivo exposure where the isolated cells were already terminally differentiated and functionally Th1-like (Gilbert et al., 2012, 2016). Our hypothesis was that a more precise depiction of CD4⁺ T-cell function in response to toxicant exposure could be obtained by assessment of gene expression/methylation patterns generated during dynamic events during differentiation in vitro. In the current study, we evaluated anti-CD3/CD28 activated, non-polarized CD4s in addition to terminally differentiated Th1 polarized cells for comparison purposes. Thus, the activated, non-polarized cells were not committed towards any subset. Based on our previous data in vivo showing TCE exposure enhanced IFN-γ expression, we predicted that TCAH would alter gene expression and/or DNA Methylation patterns in a similar manner. The results confirmed that direct exposure of CD4⁺ T cells to TCAH indeed regulated Th1 responses at the level of the transcriptome and epigenome in vitro and agreed with our prior work documenting TCAH’s immunomodulatory role in vitro and in vivo (Blossom et al., 2007; Blossom and Gilbert, 2006; Gilbert et al., 2004). TCAH may also be clinically relevant since this metabolite has been implicated as a key trigger of TCE hypersensitivity syndrome in workers exposed occupationally to TCE (Huang et al., 2014).

We demonstrated that TCAH induced specific patterns in DNA methylation between different T-cell lineages in the female mice. For instance, the promoter and CpG islands of the polarized Th1 cells were predisposed to hypermethylation, while the regions were predisposed to hypomethylation in CD4⁺ T cells. Similar patterns have been found in previous in vivo studies where hypermethylation of CpG regions in effector/memory CD4s was observed with TCE exposure in vivo (Byrum et al., 2019). From our results, TCAH appeared to strategically alter the methylation pattern to favor the expression of certain genes related to immunoregulation and autoimmunity. This could be exemplified by consistent hypomethylation and upregulation of certain immunomodulatory genes that code for interferon-ɣ (Ifng) or chemokine ligand 1 (Ccl1).

The methylation data presented here also represents some ambiguity such that we observed group of genes at the promoter regions in both the Th1 and activated/non-polarized CD4 cell population, which does not satisfy the canonical methylation expression relationship. For instance, we have noticed genes that are hypermethylated at the promoters yet upregulated, or hypomethylated at the promoters yet downregulated. These atypical epigenetic events could be explained with the finding that DNA methylation coordinates with chromatin modifications in an orchestrated fashion to regulate dynamic changes in gene expression under environmental stimuli. The data in the current study implicated the possible involvement of chromatin enhancer regulation of gene expression in TCAH-treated cells as evidenced by the downregulation of gene pathways involved in chromatin assembly/organization and nucleosome assembly in TCAH-treated CD4 cells in both sexes. These findings are further supported by recent reports in the literature in human populations where occupational TCE exposure was associated with DNA methylation alterations in blood peripheral blood mononuclear cells including markers associated with leukocyte telomere length and epigenetic aging known to significantly impact chromatin organization (van der Laan et al., 2022). One future goal will be aimed at integrating TCAH-induced differentially methylated loci with the key chromatin modifications.

One particularly intriguing finding in the current study was the striking sex-dependent effects of direct TCAH exposure on CD4⁺ T cells. At the pathway level in female mice, but not males, TCAH increased expression of genes involved in inflammatory pathways, cytokine–cytokine interactions, and metabolic pathways. Checkpoint inhibitor genes, PDL1 and PD-1 were significantly downregulated by TCAH in females. These molecules are involved in the progression of autoimmune disease since autoreactive CD4⁺ T cells are controlled by these signals (Álvarez-Sierra et al., 2023; Bukhari et al., 2023; Guan et al., 2022).

As for males, there were fewer pathways that were significantly up- or downregulated by TCAH compared to females. However, there were several significantly altered pathways by TCAH in neurologic disorders including Parkinson’s disease. TCE is increasingly being recognized as a risk factor for the development of Parkinson’s disease in both animal models and human populations (Adamson et al., 2023; Goldman et al., 2023; Wadman, 2023). The mechanism of how TCE promotes Parkinson’s disease is currently being investigated. Rats exposed to TCE by inhalation had increased LRRK2 kinase activity in the brain (De Miranda et al., 2021). In the current study, TCAH significantly enhanced the expression of LRRK2 in CD4⁺ T cells from both male and female mice. Although mutations and subsequent deficits in LRRK2 activity are associated with Parkinson’s disease, they are also found in immune-related disorders and are important in immune cell function and autophagy (Wallings and Tansey, 2019). Therefore, alterations in this gene by TCE/TCAH may have far-reaching implications for different disease susceptibility in TCE-exposed populations.

Not all of the expression of pro-inflammatory genes was sex dependent. Our data also showed that TCAH robustly significantly upregulated Ccl1 gene expression robustly by ∼4-fold in both male and female mice. Ccl1 is expressed in several immune cell subtypes and regulates Th1 differentiation (Iellem et al., 2000). As such, autoimmune mice in which the gene for IFN-γ was deleted have exacerbated pathology linked with increased expression of Ccl1 (Su et al., 2007). Although the significance of this finding is not clear, our data demonstrated the ability of TCAH to impart clear sex differences in some genes and pathways, but not others. Males and females have different immune capabilities and display different degrees of susceptibility to various diseases including autoimmune disorders. Many factors appear to predispose women to autoimmunity. However, it has been suggested that past studies, including our own, that exclusively used female mice, did not consider males are at least just as likely as females to develop toxicant-induced autoimmunity (Pollard, 2012). Although less studied, male MRL/MpJ mice eventually develop a milder form of lupus-like disease compared to female MRL/MpJ mice and die on average at 93 weeks of age, which is similar to lifespan of standard mouse strains (Storer, 1966). In the current study, the inclusion of both sexes was designed to enhance understanding of sex disparity at the gene/methylation level in CD4⁺ T cells. Similar investigations in human peripheral blood mononuclear cells are being carried out in human populations to better understand why females are more predisposed to autoimmunity and immune-mediated diseases in general (Mamrut et al., 2015). These investigations should carefully consider environmental factors that enhance disease susceptibility in individuals who may be genetically predisposed towards developing autoimmune disease.

One of the more intriguing findings of the current study was the effect of TCAH on Ifng gene expression. TCAH enhanced IFN-γ gene expression commensurate with reduction in DNA methylation in the activated, non-polarized population as well as the Th1 cells from female, but not male, mice. It is tempting to speculate that TCE skews cells towards a promotes a Th1 phenotype in vitro similar to what we observe in vivo with TCE exposure. Disease development in our model is preceded by CD4⁺ T-cell production of IFN-γ. In female MRL/MpJ mice, exposure to TCE in the drinking water increased IFN-γ production by effector/memory CD4⁺ T cells (Gilbert et al., 2016). A follow-up time course study revealed a biphasic response pattern IFN-γ up- or downregulation by TCE (Gilbert et al., 2016). This non-linear response was consistent with several studies documenting time-dependent/compensatory inflammatory mediator fluctuations, including IFN-γ, in autoimmune mouse models (Kuerten et al., 2010). This pattern has also been demonstrated in humans with autoimmune disease and correlates with an early disease phase followed by a temporary recovery and clinical relapse (Ryden et al., 2009).

Collectively, our results suggest that TCE via TCAH regulates Ifng by altering the level of DNA methylation at the gene promoter during activation and differentiation, which may have far-reaching implications for future therapeutics and/or insight into mechanism. However, despite the number of significant and shared genes between Th1s and non-polarized cells (Supplementary Tables 1 and 5), it is impossible to definitively conclude that TCAH skews an otherwise uncommitted Th cell type towards a Th1 subtype in an in vitro scenario. Even for Th1 polarized cells, there is a fair amount of plasticity involved in lineage-committed Th cell subtypes not only generated in vitro, but those differentiating in vivo (Zhu, 2018). From our own data, the expression of Tbx21, the master regulator of Th1 cell subset lineage was downregulated in TCAH-treated activated, non-polarized cells from both male and female mice, and this effect was not statistically significant (data not shown). In vivo experiments involving adoptive transfer of TCAH-skewed Th subsets are planned to address the functional and autoimmune disease-promoting effects of TCAH.

One other limitation of the current study is that while TCAH appears to behave similarly to the parent compound, TCE, it is still not known how TCAH activates CD4⁺ T cells. A clue to TCAH effector function can be determined from its chemical structure. TCAH has been shown to form a transient chemical bond known as a Schiff base with molecules on the surface of CD4s (Gilbert et al., 2004). Reports of immunostimulatory effects of Schiff-based forming compounds have been described for the drug, Tucaresol, that can activate CD4⁺ Th1 responses via bystander co-stimulation (Rhodes et al., 1995). It is plausible to hypothesize that TCAH may alter T cells in a similar fashion upon metabolism of TCE in vivo to favor stimulatory yet more generalized non-specific effects on CD4 T cells. Future experiments will address how TCAH interacts with T cells to break tolerance and promote autoimmunity.

Supplementary data

Supplementary data are available at Toxicological Sciences online.

Declaration of conflicting interests

The authors declare no conflicts of interest.

Author contributions

S.R.C.: conceptualization, data curation, methodology, formal analysis, validation, writing—original draft. S.D.B.: data curation, methodology, formal analysis, validation, writing—original draft. S.J.B.: conceptualization, data curation, funding acquisition, methodology, investigation, supervision, writing—original draft.

Funding

National Institutes of Health Institute of Environmental Health Sciences (NIH/NIEHS) grant R01ES030323; P30ES032755; the University of New Mexico College of Pharmacy; the Arkansas Children's Research Institute and the Center for Translational Pediatric Research funded under the National Institutes of Health National Institute of General Medical Sciences (NIH/NIGMS) grant P20GM121293.

Data availability

Supplementary data are available at Toxicological Sciences online. RNA-seq and DNA methylation metadata files are available on Dryad (doi:10.5061/dryad.msbcc2g5k).