RNA factories shape trans chromatin topology
|
Our guiding hypothesis is that RNA biogenesis and 3D genome organization are mutually instructive processes. Specifically, we propose that transcription and RNA processing occur within dynamic inter-chromosomal domains—spatial clusters where genes located on different chromosomes come together through shared trans-acting nucleic acid–binding proteins acting as "genome architects". These concentrate at core loci to coordinate transcription and splicing at accessory loci. In this model, the same molecular forces that shape RNA biogenesis also sculpt the global architecture of the genome.
This idea builds on our discovery that the splicing regulator RBM20, mutated in familial cardiomyopathy, nucleates such an RNA factory around its target gene TTN in cardiomyocytes. Within this compartment, co-regulated loci interact across chromosomes, forming what we term trans-interacting chromatin domains (TIDs). We hypothesize that these domains enhance the efficiency and fidelity of gene-expression programs—precisely the properties required to achieve full cellular maturation. |
The inter-chromosomal RNA factories hypothesis and examples thereof, adapted from Bertero A., Frontiers in Genetics 2021
|
Mapping Inter-chromosomal networks with trans-C
To systematically explore this architecture, we developed trans-C, a computational framework that identifies statistically significant inter-chromosomal gene cliques from Hi-C data. trans-C performs probabilistic random walks across networks extracted from genome-wide contact maps to reveal loci that interact in trans. Applying this method to multiple species and cell types, we found that genes bound by the same transcription or splicing factors frequently co-localize in three dimensions, forming higher-order TIDs. These are particularly enriched for proteins with intrinsically disordered domains—a feature linked to phase separation—and are often positioned near nuclear speckles, sites of intense RNA processing. These findings support our central hypothesis: inter-molecular proximity between co-regulated nucleic acids is a pervasive, functional mechanism in biology, linking RNA metabolism to nuclear organization.
Overview of the trans-C algorithm and its goal, adapted from Hristov A. et al., Genome Research 2024
Dissecting single-locus microenvironments with o-MAP
To validate and characterize these interactions biochemically, we co-developed single-locus o-MAP (oligonucleotide-mediated proximity-interactome mapping) with Dave Shechner (University of Washington). o-MAP is an RNA-centric proximity ligation method to profile the proteins, RNAs, and chromatin fragments surrounding specific RNA molecules. When applied to pre-mRNAs, this allows mapping their nuclear neighbourhood. Demonstrating this method, targeting o-MAP to introns of TTN pre-mRNA revealed a muscle-specific RNA factory containing QKI, SAFB, and other architectural factors, surrounded by a compartment of transcriptionally silenced chromatin. Disruption of RBM20 remodeled nearly every facet of this structure, demonstrating that individual genes can function as organizers of nuclear architecture.
By combining trans-C computational mapping with o-MAP biochemical profiling, we can now bridge the structural and functional dimensions of genome organization. Together, these complementary approaches allow us to interrogate the chromatin interaction networks of candidate genes emerging from our functional studies, directly linking sequence-level perturbations to higher-order nuclear architecture.
By combining trans-C computational mapping with o-MAP biochemical profiling, we can now bridge the structural and functional dimensions of genome organization. Together, these complementary approaches allow us to interrogate the chromatin interaction networks of candidate genes emerging from our functional studies, directly linking sequence-level perturbations to higher-order nuclear architecture.
Schematic principle for single-locus o-MAP, from Kania E. et al., bioRxiv 2025