Regulation of microtubule dynamics is required for many biological processes. Here we uncovered a novel mechanism to regulate microtubule dynamics: local tubulin condensation. We show that two microtubule tip- binding proteins, CLIP-170 and EB3, undergo phase separation and form a liquid-like network at growing microtubule tips. In vitro and in cells this liquid network can condense tubulin. This process of tubulin co-condensation is spatially initiated at the microtubule tip and temporally regulated to occur only when microtubules grow. Tubulin condensation at the growing microtubule tip increases growth speeds up to four-fold and strongly reduces depolymerization events. Since tubulin concentration determines microtubule growth, our discovery shows a regulatory mechanism for enriching tubulin availability at strategically important locations: the growing microtubule tips.

AI can act as a lever of human ingenuity, supporting the identification of new biological mechanisms involved in disease, therapies and medicines. While artificial intelligence and machine learning are often portrayed as prediction systems, their capacity to combine unrelated concepts in a sensible way makes them also excellent partners to inspire innovation. The talk will present the mechanisms of AI creativity, and practical use cases for research and industry.

Phenotypic screens are powerful to identify small molecules that act on a biological process of interest, but the elucidation of the cellular target and/or mechanism of action of the hit compounds presents a major challenge. Consequently, hit compounds often do not reach their full potential as pharmacological leads or chemical biology tool compounds.

Exemplary of this, Hedgehog Pathway Inhibitor 1 (HPI-1) was found as a hit in a phenotypic screen for the Hedgehog (Hh) signaling pathway – a major developmental signaling cascade that establishes the embryonic body plan, and dysregulation of which underlies various cancers. HPI-1 robustly inhibits the Hh pathway in a variety of cell lines, downstream of the activator Smoothened, yet its cellular target has remained elusive for many years.

Here, we present the target elucidation of HPI-1 through the design, synthesis, and evaluation of corresponding proteolysis targeting chimeras (Hedgehog Pathway PROTACs, HPPs) coupled with label-free quantitative proteomics. We show that HPP-9 robustly reports on HPI-1 action on various BET bromodomain proteins, epigenetic modulators known to be important for Hedgehog signal transduction, through their degradation. Moreover, HPP-9 is the first example of a PROTAC targeting the Hedgehog pathway, enabling novel pharmacological strategies to combat Hh pathway-driven disease.

AI can act as a lever of human ingenuity, supporting the identification of new biological mechanisms involved in disease, therapies and medicines. While artificial intelligence and machine learning are often portrayed as prediction systems, their capacity to combine unrelated concepts in a sensible way makes them also excellent partners to inspire innovation. The talk will present the mechanisms of AI creativity, and practical use cases for research and industry.