Two variant cohesin complexes containing SMC1, SMC3, RAD21 and either STAG/SA1 or SA2 are present in all cell types. We report here their genomic distribution and their specific contributions to genome organization in human cells. While both variants are found at CTCF sites, a fraction of cohesin-SA2 localizes to enhancers lacking CTCF, is linked to tissue-specific transcription and cannot be replaced by cohesin-SA1 when SA2 is absent, a condition observed in several tumours. Downregulation of either variant has different consequences for gene expression and genome architecture. Our results suggest that cohesin-SA1 preferentially contributes to the stabilization of TAD boundaries together with CTCF, while cohesin-SA2 promotes cell type-specific contacts between enhancers and promoters independently of CTCF. Loss of SA2 rewires local chromatin contacts and alters gene expression. These findings provide insights on how cohesin mediates chromosome folding and establish a novel framework to address the consequences of cohesin mutations in cancer.

Highlights d Cohesin variants have distinct effects on mESC chromatin architecture and transcription d Cohesin-SA1 preferentially contributes to TAD boundary strength d Cohesin-SA2 facilitates Polycomb domain compaction through PRC1 recruitment d Cohesin-SA1 impedes and cohesin-SA2 promotes aggregation of distal Polycomb domains

In addition to mediating sister chromatid cohesion, cohesin plays a central role in DNA looping and segmentation of the genome into contact domains (TADs). Two variant cohesin complexes that contain either STAG/SA1 or SA2 are present in all cell types. Here we addressed their specific contribution to genome architecture in non-transformed human cells. We found that cohesin-SA1 drives stacking of cohesin rings at CTCF-bound sites and thereby contributes to the stabilization and preservation of TAD boundaries. In contrast, a more dynamic cohesin-SA2 promotes cell type-specific contacts between enhancers and promoters within TADs independently of CTCF. SA2 loss, a condition frequently observed in cancer cells, results in increased intra-TAD interactions, likely altering the expression of key cell identity genes.

Highlights d Cohesin variants have distinct effects on mESC chromatin architecture and transcription d Cohesin-SA1 preferentially contributes to TAD boundary strength d Cohesin-SA2 facilitates Polycomb domain compaction through PRC1 recruitment d Cohesin-SA1 impedes and cohesin-SA2 promotes aggregation of distal Polycomb domains

Background The cohesin complex organizes the genome-forming dynamic chromatin loops that impact on all DNA-mediated processes. There are two different cohesin complexes in vertebrate somatic cells, carrying the STAG1 or STAG2 subunit, and two versions of the regulatory subunit PDS5, PDS5A and PDS5B. Mice deficient for any of the variant subunits are embryonic lethal, which indicates that they are not functionally redundant. However, their specific behavior at the molecular level is not fully understood. Results The genome-wide distribution of cohesin provides important information with functional consequences. Here, we have characterized the distribution of cohesin subunits and regulators in mouse embryo fibroblasts (MEFs) either wild type or deficient for cohesin subunits and regulators by chromatin immunoprecipitation and deep sequencing. We identify non-CTCF cohesin-binding sites in addition to the commonly detected CTCF cohesin sites and show that cohesin-STAG2 is the preferred variant at these positions. Moreover, this complex has a more dynamic association with chromatin as judged by fluorescence recovery after photobleaching (FRAP), associates preferentially with WAPL and is more easily extracted from chromatin with salt than cohesin-STAG1. We observe that both PDS5A and PDS5B are exclusively located at cohesin-CTCF positions and that ablation of a single paralog has no noticeable consequences for cohesin distribution while double knocked out cells show decreased accumulation of cohesin at all its binding sites. With the exception of a fraction of cohesin positions in which we find binding of all regulators, including CTCF and WAPL, the presence of NIPBL and PDS5 is mutually exclusive, consistent with our immunoprecipitation analyses in mammalian cell extracts and previous results in yeast. Conclusion Our findings support the idea that non-CTCF cohesin-binding sites represent sites of cohesin loading or pausing and are preferentially occupied by the more dynamic cohesin-STAG2. PDS5 proteins redundantly contribute to arrest cohesin at CTCF sites, possibly by preventing binding of NIPBL, but are not essential for this arrest. These results add important insights towards understanding how cohesin regulates genome folding and the specific contributions of the different variants that coexist in the cell.