Background

Controlling brown adipose tissue (BAT) plasticity offers potential for novel obesity therapies. DNA methylation is closely linked to thermogenic and metabolic pathways and thereby influences BAT function. How metabolic state and cold exposure interact to shape methylation-dependent BAT gene regulation was investigated.

Methods

Five-week-old mice were fed either chow for 11 weeks (lean) or high-fat diet for 22 weeks to induce obesity (DIO), after which cold exposure was applied for seven days. BAT transcriptomes (RNAseq) and methylomes (RRBS) were generated, and differentially methylated and expressed genes (DMEGs) showing metabolic state–dependent cold responses were identified. Pathway enrichment, epigenetic regulator screening, and transcription factor (TF) motif analyses were performed. DNA methylation was experimentally modulated in vitro to validate selected gene expression responses.

Results

A total of 1,364 differentially expressed genes (DEGs) were uniquely affected by the interaction of metabolic state and cold, with most downregulated in DIO mice. Sixty-five DMEGs (4 % of DEGs) showed metabolic state–specific responses to cold. In DIO mice, DMEGs were enriched in pathways associated with mitochondrial dysfunction, altered lipid metabolism, neuroendocrine signaling, and stress responses. Several epigenetic regulators, including Tet2, Dnmt3a, and Apobec1, exhibited metabolic state- and cold-dependent expression, and TF-motif analyses highlighted roles for AhrArnt and Foxn1. In vitro assays confirmed that DNA methylation influences expression of thermogenic genes.

Conclusion

These findings provide the first evidence that the epigenetic cold response of BAT differs by metabolic condition. BAT remodeling is shaped by coordinated transcriptional and epigenetic mechanisms integrating environmental and metabolic cues.

Graphical abstract

Mice were housed under cold exposure (8 °C) or thermoneutrality (30 °C) and fed either chow (lean) or high-fat diet (HFD; diet-induced obese (DIO)). RNA sequencing (RNAseq) and reduced representation bisulfite sequencing (RRBS) was performed on brown adipose tissue (BAT) which provide the fundament for the identification of differentially methylated positions (DMP) and regions (DMR) as well as differentially expressed genes (DEG) in three models: COLD_lean, comparing 8° versus 30 °C mice on chow diet; COLD_DIO, comparing 8° versus 30 °C mice on HFD diet and ΔCOLD, comparing COLD_lean versus COLD_DIO. Differentially methylated and expressed genes (DMEGs) were identified for all comparisons, based on DMPs within DMRs that significantly correlated with differentially expressed genes (DEGs; the heatmap shows a representative example). DMEGs were further characterized. First, using a published single cell dataset (Shamsi et al., 2023, https://doi.org/10.1038/s42003-023-05140-2), the distribution of DMEGs across cell types was analyzed. For further characterization, DMEGS were categorized into similarly (among COLD_lean and COLD_DIO) and uniquely (COLD_lean, COLD_DIO and ΔCOLD) regulated. Pathway enrichment analyses were performed and to identify differentially regulated epigenetic regulators among the DMEGs, the EpiFactor Database (https://epifactors.autosome.org) was used. A TF-binding site motive enrichment analyses (https://jaspar.elixir.no)was applied to add information of most enriched Transcription Factors (TF). Finally, to gain functional insights of DNA methylation changes on selected DMEG candidates, cell cultures of immortalized brown adipocytes and primary brown adipocytes were treated with 5′aza-2′-deoxycytidine (demethylation) or S-adenosylmethionine (upregulation of methylation), following gene expression analyses. The figure was created with BioRender.