Roman Barth

Huge thanks to everyone involved, especially my co-first authors, @ceesdekker.bsky.social , and our great collaborators at Jan-Michael Peters' lab @impvienna.bsky.social

CTCF doesn't just have one stop signal—it has two!The YDF & KTYQR motifs stall the cohesin loop extruder in different ways: one is a roadblock, the other forces one-way traffic.

🚨New paper out! In our @MolCellNews paper, we solve a piece of the puzzle of how our genome folds. We found the two specific molecular "brakes" in the CTCF protein that stop DNA loop extrusion. One of the final pieces of my PhD!
tinyurl.com/29uuc5by

Telomeres stall DNA loop extrusion by condensin Analikwu et al. show that dense linear arrays of the telomeric protein Rap1 on DNA block DNA loop extrusion by condensin. Rap1-induced stalling depends on array length and protein density and leads to a local chromatin decompaction during anaphase, revealing how protein arrays can regulate SMC-driven genome organization.

Telomeres stall DNA loop extrusion by condensin

Here's a new very exciting paper from our #CDlab: www.cell.com/cell-reports...

In an exemplary in vitro/in vivo study, 1st authors ‪@btanalikwu.bsky.social‬ & Alice Deshayes showed that dense linear protein arrays (e.g. telomeres) do stop loop-extruding SMCs!

Great collaboration with Marcand lab.

Latest #CDlab paper online now:
Cohesin supercoils DNA during loop extrusion
www.cell.com/cell-reports...

This extensive study was led by Jan Michael Peters at Vienna; our lab contributed mostly modeling of plasmid supercoiling (@romanbarth.bsky.social‬) and single-molecule data (Richard Janissen).

Congratulations to the 2025 Schmidt Science Fellows - exciting that you'll be joining the community!!

We have 3 openings for ambitious postdocs or PhD students in our #CDlab - for exciting single-molecule biophysics research on nuclear pores, peroxisomes, or archaeal divisomes.

Check it out and apply: ceesdekkerlab.nl/come-join-us/

RT=nice!

Fred Hutch announces 10 recipients of the 2025 Harold M. Weintraub Graduate Student Award Fred Hutch announces 10 recipients of the prestigious 2025 Harold M. Weintraub Graduate Student Award

www.fredhutch.org/en/news/rele...

Huge congrats to our #CDlab PhD student @romanbarth.bsky.social for winning the prestigious 2025 Weintraub Award for his PhD thesis work on SMC motor proteins!

www.fredhutch.org/en/news/rele...

Great honor and well deserved Roman!

Science is a collaborative adventure, and this moment really feels like a celebration of the collective effort, curiosity, and passion that drives research.

I'm excited to share that I've been selected as one of the 10 recipients of the prestigious Weintraub Award from @fredhutch.bsky.social this year! 🏆
I owe a massive debt of gratitude to the remarkable colleagues and mentors who have shaped my path, especially @ceesdekker.bsky.social !!

Check out this cool article by @newscientistnl.bsky.social on our latest work about direction-switching molecular motors racing along DNA: www.newscientist.nl/nieuws/eiwit...

in a PDS5 knockdown? STAG2 seems to be a liiiiittle enriched there (Fig. 4c) and AF is more confident in STAG2 than in STAG1. Yet we didn't test STAG2.
Nevertheless, in the (local) absence of PDS5, CTCF might use the KTYQYR motif to interfere with cohesin as our data suggests..

seems to happen also experimentally. We didn't explore PDS5 binding to KTYQR computationally or experimentally. It might be that KTYQR-PDS5 binding is stronger than KTYQR-STAG binding and this is thus the dominant mode seen in vivo. Are you aware of Hi-C and/or binding (as in your Fig. 4) data..

Hi @elphegenoralab.bsky.social and @eecutts.bsky.social . True, the KTYQR motif has been identified by both of your cited articles as PDS5 binding regions. We didn't test PDS5 in our assays, only cohesin-NIPBL. Seeing that AF suggested KTYQR-STAG binding was surprising to us and more so that that..

Latest #CDlab paper now at @biorxivpreprint.bsky.social: www.biorxiv.org/content/10.1...

In this paper, we identify two amino acid motifs on CTCF that impede cohesin-mediated DNA loop extrusion - important as CTCF and cohesin are crucial to the structure and regulation of our chromosomes.

We started

Thank you to all that contributed!! Some on bluesky: @jacotorre.bsky.social @ceesdekker.bsky.social, especially to the fantastic team at the IMP Vienna Gabriele Litos, Iain Davidson, and Jan-Michael Peters!!

We hope that by disentangling the multiple contributions of the CTCF N-terminus to the stalling of cohesin at CTCF sites, we are another step closer to understanding how cohesin and CTCF control and shape genomes across the tree of life 🌳

To summarize:

KTYQR is distinct from YxF: it reduces cohesin's LE activity overall, but doesn't impact cohesin's directionality.

Together, the two motifs account pretty much for the impact of the full NTR!! We hypothesize that connecting the two on one protein chain and some adjacent sequences modulate further.

A computational AlphaFold screen suggested the KTYQR motif, located N-terminally of the YxF, that appears to also bind STAG1.

Even though YxF has a large impact on cohesin's LE activity, it doesn't fully account for the impact of the full N-terminal region of CTCF. So we searched more 🔎

And indeed, FCS experiments showed that STAG1's DNA affinity increases upon addition of the YxF motif.

Curiously, we observed that cohesin also changes direction much less frequently in the presence of YxF! Taking inspiration from Shaltiel et al, Science, 2022, we hypothesized that this is due to a higher DNA affinity of some cohesin subunit in the presence of YxF. Our main candidate: STAG1-kleisin

We started with the known YxF motif. In brief, the YxF motif:
- reduces cohesin's ATPase rate
- reduces LE initiation
- reduces the fraction of complete LE steps (in Magnetic Tweezers)
So these few amino acids alone already tinker quite a lot with cohesin!

Observing LE by Magnetic Tweezers provides a much more detailed view, however. We can see individual steps made by cohesin. Previously, we already observed a plethora of phenomena: cohesin often makes full LE steps, as seen by a height decrease of the magnetic bead. But these are also often reversed

SMC motor proteins extrude DNA asymmetrically and can switch directions All eukaryotic SMC motor proteins extrude DNA asymmetrically but may undergo direction switches, suggesting a common loop extrusion mechanism among all eukaryotic SMC proteins while permitting bidirec...

Single-molecule fluorescence visualization lets us assess quantitative metrics such as the loop extrusion (LE) initiation rate, LE rate, directionality changes (see www.cell.com/cell/fulltex..., etc.

We took the CTCF N-terminal region that is already well-known to interact with cohesin and cut it up into smaller pieces. Then, we exposed cohesin to these various fragments in buffer (no DNA binding domain here!) and observed how cohesin handles that.