Beautiful 😎

This work was led by Marion Mouginot, who notably established a locus-specific proteomics approach that identified Lola-I as a meta-loop-associated factor,
together with Sahar Hani (supported by Arianna Ravera, UNIL’s image analysis expert!), who uncovered hierarchical meta-loop formation. 👏

Lola-I was, however, only required to form a single meta-domain. This highlights a dichotomy between factors like Lola-I, which may control neuronal fate largely independently of inter-TAD interactions, and structural proteins like Cp190, which may enable ultra-long range gene regulation.

We developed an unbiased proteomics approach (inspired by Baldi..Becker 2018) to identify proteins bound to meta-loop anchors. This identified a specific isoform of the axon guidance protein Longitudinals lacking (Lola-I) as enriched at several meta-loop anchors, in our pull-downs and in vivo.

Many meta-loop anchors coincide with TAD boundaries. We found that a major determinant of TAD boundary formation in flies (called Cp190) is required to form a large fraction of meta-loops, including CTCF-dependent meta-loops.

How do meta-loops form?
Most meta-domains contain 2 or more meta-loops, with some of these forming earlier during neurogenesis than the other(s) in the same meta-domain.
We found that in many meta-domains, specific meta-loops (often the earlier forming ones) nucleate others.

1 meta-domain we studied contains a meta-loop connecting promoters of Glutamate receptors 1A and 1B. Disrupting the meta-domain uncoupled their co-transcription (similar to Levo..Levine 2022 for intra-TAD loops connecting paralogous gene promoters).

We precisely disrupted 3 independent meta-domains and found transcriptional defects in all 3 cases.
2 meta-domains we studied contain a meta-loop connecting a neuronal gene promoter to a distant intergenic element. Disrupting the meta-domain decreased neuronal gene transcription by 2-to-3-fold.

We recently showed that in fly neurons, specific pairs of TADs megabases apart specifically come together to form meta-domains (Mohana et al. 2023).
Within meta-domains, meta-loops connect accessible chromatin peaks present at promoters of specific neuronal genes or in non-coding intergenic DNA.

and thank you to the reviewers whose feedback greatly improved the review 😎

Fun to do this with talented Sanyami Zunjarrao in our lab!
Thank you @andersshansen.bsky.social and Marcelo Nollmann for including our review in the 2025 Genome Architecture & Expression edition of COGEDE 😀

yes indeed, BEAF-32 mat- zyg- mutants are not happy...

Thank you to the outstanding efforts of notably Anastasiia Tonelli and Pascal Cousin and another great collaboration with Aleksander Jankowski @[email protected] 💪

For more info:
Open access paper: www.sciencedirect.com/science/arti...
Step-by-step protocol: www.sciencedirect.com/science/arti...

Exciting findings using insulator-seq are:
Not all insulator-binding protein sites act as enhancer-blockers.
Not all TAD boundaries act as enhancer-blockers.
Both a CTCF motif and its surrounding sequence context are essential for enhancer-blocking.

(2) The insulator activity of CTCF binding sites depends on an intact CTCF motif.

Here’s some proof that insulator-seq works:
(1) It identifies biologically meaningful insulators in genomic DNA (both known and novel).

That’s right, insulators works on transiently transfected plasmids! This allows us to test the enhancer-blocking activities of up to 1 million DNA test fragments at a time.

Anastasiia Tonelli in the lab established insulator-seq - a massively parallel reporter assay for enhancer-blockers in Drosophila cells