Integrative Genomics Identifies the Corepressor SMRT as a Gatekeeper of Adipogenesis through the Transcription Factors C/EBPβ and KAISO

Abstract

The molecular role of corepressors is poorly understood. Here, we studied the transcriptional function of the corepressor SMRT during terminal adipogenesis. Genome-wide DNA-binding profiling revealed that this corepressor is predominantly located in active chromatin regions and that most distal SMRT binding events are lost after differentiation induction. Promoter-proximal tethering of SMRT in preadipocytes is primarily mediated by KAISO through the conserved TCTCGCGAGA motif. Further characterization revealed that KAISO, similar to SMRT, accelerates the cell cycle and increases fat accumulation upon knockdown, identifying KAISO as an adipogenic repressor that likely modulates the mitotic clonal expansion phase of this process. SMRT-bound promoter-distal sites tend to overlap with C/EBPβ-bound regions, which become occupied by proadipogenic transcription factors after SMRT clearance. This reveals a role for SMRT in masking enhancers from proadipogenic factors in preadipocytes. Finally, we identified SMRT as an adipogenic gatekeeper as it directly fine-tunes transcription of pro- and antiadipogenic genes.

Absolute quantification of transcription factors during cellular differentiation using multiplexed targeted proteomics

“…However, our approach is in principle not limited to ten target proteins, provided npg online methods 3T3-L1 cell culture, differentiation and protein extraction. 3T3-L1 cells were cultured and differentiated into adipocytes as detailed in Raghav et al 31 . Briefly, 3T3-L1 cells were collected at six different differentiation time points (0 h, 2 h, day 1, day 2, day 4 and day 6).…”

“…3b). Although it is difficult to precisely estimate the latter threshold, a survey of our own and published ChIP data revealed that the signal associated with DNA binding is typically considered as positive from around 1% (as compared to input) 30,31 . Notably, when we used 1.35% as the detection threshold, our model predicted binding-event numbers that closely mimic the experimentally derived data (Fig.…”

“…Mouse TF ORFs were cloned in Gateway format essentially as described earlier 31,34 . To make the wheat germ (WG) in vitro transcription-translation expression system (Promega) compatible with the mouse TF ORF clones and to allow the purification of TFs, we modified the SP6 pF3A WG (BYDV) Flexi vector (Promega) to accommodate the Gateway reading frame A cassette (Invitrogen).…”

“…We therefore had to devise a strategy that did not rely on the use of publicly available information to discover proteotypic peptides of TFs. To alleviate concerns regarding the detectability of TFs, we implemented three different bidimensional peptide fractionation pipelines on 3T3-L1 total nuclear protein extracts at day 4 of differentiation (a time point at which the adipogenic master regulators PPARγ and RXRα are known to be highly expressed 31,41 ) with the precise aim of discovering TF-specific peptide candidates that could be used in downstream quantitative analyses. However, only two peptides for each TF were retrieved in spite of the higher resolving power provided by the extensive fractionation and the consecutive higher sensitivity expected in mass spectrometry (Supplementary Fig.…”

Mapping genome-wide transcription-factor binding sites using DAP-seq

“…Alternative methods are needed for capturing genome-wide binding data for many organisms. In vitro TFBS identification methods such as Protein Binding Microarrays (PBM) and High Throughput Systematic Evolution of Ligands by Exponential Enrichment (HT-SELEX) have achieved the highest throughput for deducing TF binding specificities in vitro [9][10][11] , but these methods use short synthetic oligonucleotides lacking secondary DNA modifications and genomic context, both important determinants of selective TF binding in vivo [12][13][14][15][16] . The DAP-seq technique 15 described here employs an in vitro-expressed affinitytagged TF in combination with high-throughput sequencing of a genomic DNA library, allowing for the generation of genome-wide binding site maps reflective of both local sequence context and DNA methylation status.…”