Lennart Hilbert

Sharing a recent, deep insight:

The concept of a membrane-less organelle is an interesting one, like calling something a house without walls

Happy 200th bday KIT 🥳

Sadly, yes, and applies also past junior level! Sorry to everyone affected by my unreal email and task overload in the form of non-responsiveness 😣

This might explain a lot of emails I am getting. Using Mac Mail…

Thought the same recently, glad we have an open and international discussion platform again! Still, engagement seems lower than on Twitter during the good days, despite the higher follower count here

I was about to ask :-)

Thanks, no need to defend, your points are clear & justified! We have data where gene and enhancer don’t come closer than ~200-300 nm, but do come into that distance systematically. I’m trying to understand whether that distance is sufficient to mainly rule out it’s driven by your/a cohesion effect.

Warming up for October start of KIT teaching by guest lecturing for @unistra.fr Cell Physics Master “Active Droplets”. First of three lectures today. 3:00 h sharp, start to finish 💪

Some travel impressions, including fancy and very white “Le Studium” on Strasbourg Central Campus 🤍

And, Gyoza 🥟

Right, that makes sense, also and especially considering the question of your study. I was wondering how close a gene would have to cone to an enhancer to be “caught” by a cohesion. It would help distinguish what type of event we see in our own data. Any idea?

One question, after reading your short overview:

How close, in terms of nm distance, would one locus have to come to a second locus for them to be caught into this type of cohesin-based long-lived engagement? Something below 100 nm? I hope my question makes sense…

The goal of a PhD is not to learn some facts or read a few papers or learn a bunch of techniques. The goal of a PhD is to learn independence, problem solving, how to finish things you start, resilience, & gain the ability to adapt & think creatively. Learning these things is hard.

Very much on the case, also motivated as the project came back today during the defense in Erlangen 👌

☺️

It’s moving for you! ☺️

It was a pleasure and an honor to walk this part of the path together, and amazing to look back at the achievements today. Also, two very smart oral examiners, Alexandra Chambony & Helmut Brandstätter, quite insightful and sharp questions! They’re here, right after the grade was announced.

What a great doctoral defense by Tim Klingberg @fau.de

Based in the group of our collaborator @vasilyzaburdaev.bsky.social Tim decisively contributed to at least 3 of our own lab’s foundational projects, which he showed today at “Max Planck Zentrum Physik und Medizin”

Some impressions

Welcome on board, happy to see you joined!

Congratulations to the Fritz Lipmann Lecturer Michael Rosen (Dallas) by the GBM president Harald Kolmar.

GBM Compact Focus on Condensate Biology is opened by Edward Lemke. Looking forward to a great meeting!

If you are at GBM compact "Focus on Condensate Biology", we will be happy to talk to you at our lab's posters:

Mona Wellhäuser, MC15 @mofrawe.bsky.social
Yuzhi Bao, CA03 @yuzhibao.bsky.social
Lennart Hilbert, MC06

Conference abstracts:

gbm-compact.org/abstracts.html

*sketch added as click bait

The label maker is the single most gratifying machine one can buy for a lab!

Um das Verhalten der DNS nachvollziehbar und vorhersagbar zu machen, kombinieren die Forschenden Laborexperimente mit Computersimulationen.

Molekulare #Bionik: Der Zellkern als Vorbild für DNS-basierte #Computerchips. Forschende des #KITKarlsruhe (u. a. @lennarthilbert.bsky.social) nutzen zelluläre Informationsverarbeitung als Inspiration für die #Biotechnologie. www.kit.edu/kit/pi_2025_...

Preferential associations between co-regulated genes reveal a transcriptional interactome in erythroid cells - Nature Genetics Peter Fraser and colleagues report a genome-wide analysis of transcription interactions involving the globin genes in mouse erythroid cells. They demonstrate that the transcription factor Klf1 mediate...

If you want actual evidence for cis and trans interactions, this might be one classic paper that has it:

www.nature.com/articles/ng....

We also have a full nucleus model (unpublished). Surprise: it seems to do both 🤷🏼‍♂️

GitHub - lhilbert/EnhancerShoeBox: Molecular Dynamics Simulations of Enhancer-Promoter Interactions in a Model Shoe Box Molecular Dynamics Simulations of Enhancer-Promoter Interactions in a Model Shoe Box - lhilbert/EnhancerShoeBox

Our postdoc at the time aptly called this the Enhancer Shoe Box model. So, all that I can say for this model: the two chromosome segments are each attached to the opposing short sides of the shoe box ;-)

Simulation code is here, if anyone needs it:

github.com/lhilbert/Enh...

Simulation of the formation of a chromatin-attached condensate, which is visited by a gene undergoing transcriptional activation.

Read more about it here:

nyaspubs.onlinelibrary.wiley.com/doi/10.1111/...

The paper provides a broader discussion of the challenges of DNA computing and designing a chip architecture, and how methods from accelerated/virtual materials design can drive progress.

Thanks to all for contributions and input along the way! Co-author @mofrawe.bsky.social is also on BlueSky

Simulation of surface condensation in the presence of a model polymer chain. (A) Basic interaction rules in a system of diffusing particles result in liquid–liquid phase separation and surface condensation. 2D-projections of simulations show that for a low number of condensate-forming particles, droplets only form on interactive polymer sections. Only at higher numbers of condensate-forming particles can droplets form independently. (B) The association of particles into droplets is detected at lower particle concentrations in the presence of a condensation surface. (C) When increasing the bulk concentration of condensate-forming particles above the pre-wetting concentration, , a dense droplet phase appears in the presence of a condensation surface (top). Without a surface, a dense droplet phase forms only above the higher saturation concentration . Droplet phase concentration obtained using the convex hull around DBSCAN detected clusters. (D) Simulating the limited resolution of a microscope and implementing a quantification of condensate intensities similar to the analysis of microscopy images leads to a progressive increase of intensity above (in the presence of surface, top) or above (without surface, bottom). All values are mean SEM, independent simulations per condition.

A coarse-grained simulation reproduces, again, the behaviors seen in our synthetic DNA experiments, and explains some effects of finite microscope resolution, as well as complex interactions of condensates and surface DNA strands.

Use of synthetic DNA nanostructures to reconstruct the surface condensation process underlying transcription factory formation. (A) The condensed material of a transcription factory is mimicked using X-shaped nanomotifs made of short DNA oligomers. The nanomotifs have sequence-encoded binding sites for other nanomotifs (magenta) and for target areas located on DNA strands (green). These target areas enable the local formation of surface condensates, similar to the condensate formation at genomic control elements. (B) Two-color microscopy images showing nanomotifs (magenta), to which surfaces (green) were added or not added. Images are maximum intensity z-projections, intensity look-up tables were adjusted for each image individually to facilitate visual interpretation. For separate channel images, see Figure S1. (C) Number of condensates per volumetric stack, the nanomotif channel intensity within these condensates, and the nanomotif channel background intensity for increasing nanomotif concentration, in the presence or absence of added condensation surfaces. All values are mean with 95% confidence intervals, 500 bootstrap resamples, volumetric stacks per condition for condensate number and background intensity; for condensate intensity, is the total number of condensates pooled from 20 stacks per condition. For the middle and bottom intensity plots, the circle area is proportional to the number of detected condensates.

We took a short-cut, designed DNA-nanomotifs behaving similarly to the in vivo liquid phase, condensing in targeted fashion on long, repetitive DNA strands produced bz rolling circle amplification (RCA). Crucial behaviors of the surface condensate regime are reproduced! @xeniatschurikow.bsky.social

Delineation of the surface condensation regime in contrast to canonical liquid–liquid phase separation (LLPS). Reliable operation of a computing system that uses properties and functions afforded by surface condensation can be expected within the inner regions of the surface condensation regime, above the pre-wetting concentration () and below the saturation concentration (). Blue and red lines in the bottom diagram represent stable concentrations that can be obtained by demixing into condensates via surface condensation or LLPS, respectively. Dashed lines are theoretical expectations and solid lines are expected observations considering a finite microscope resolution. Sketch based on simulations in our previous work.

The surface condensation regime holds very specific implications for the behavior of liquid-phase condensates: they are growth-limited by amount of available surface, and localize to these surfaces. The regime occurs at sub-saturated (C<C_sat) concentrations, also relevant for the in vivo regime.