CORDIS - Forschungsergebnisse der EU

Ubiquitin-dependent regulation of the microtubule cytoskeleton

Periodic Reporting for period 1 - UBIMAPS (Ubiquitin-dependent regulation of the microtubule cytoskeleton)

Berichtszeitraum: 2017-11-01 bis 2019-10-31

Ubiquitylation of proteins is catalyzed by ubiquitin E3 ligases and frequently leads to subsequent degradation of target proteins via the proteasome.
Using comprehensive substrate screening on protein microarrays, we identified several microtubule-associated proteins as substrates of the multi-subunit ubiquitin E3 ligase SCFFbxw5 in vitro implicating its crucial role in the regulation of microtubules. After initial confirmation of these targets, our work focused on Kif2c/MCAK (mitotic centromere-associated kinesin), a microtubule depolymerase that plays important roles in mitosis.

Medical relevance
In agreement with its pivotal function in cell division, overexpression of Kif2c is frequently observed in different cancer types and is often associated with poor patient prognosis. It has been proposed that via its microtubule depolymerase activity, Kif2c overexpression may counteract the activity of the broadly applied anti-cancer drug paclitaxel (taxol) – a microtubule stabilizing agent. Hence, by elucidating a potential ubiquitylation and degradation of Kif2c by SCFFbxw5, our work may eventually contribute to the development of potential new therapy approaches for tumors overexpressing Kif2c.

We first aimed at validating proteins identified in the screen as substrates of SCFFbxw5 (WP1) in order to focus on one specific target protein afterwards. Using different in vitro and in cellulo approaches we tried to identify binding sites on substrate for Fbxw5 and sought to elucidate the physiological role of the ubiquitylation reaction (WP2). In the last part, we wanted to address temporal and spatial characteristics of the SCFFbxw5-dependent regulation of the substrate and the molecular mechanism behind such a potential confinement (WP 3).

By combining different in vitro and in cellulo assays, we could establish Kif2c as a bona fide substrate of SCFFbxw5 in cells. In addition, we could demonstrate that the SCFFbxw5 targets Kif2c in the G2 stage of the cell cycle and that this process is required for subsequent ciliogenesis later in G0/G1. Importantly, knock down of Fbxw5 displayed also a severe synthetic lethality with Kif2c overexpression indicating a new therapeutic potential of Kif2c overexpressing cancer cells.

In line with being a potential substrate, Kif2c bound strongly to Fbxw5 in an in vitro pull down experiment and was very efficiently and specifically ubiquitylated by SCFFbxw5, forming predominantly K48 chains (Figure 1).


We tried to identify binding sites within Kif2c for Fbxw5 by chemical crosslinking and mutational mapping. However, chemical crosslinking with recombinant proteins yielded no specific cross-links for Fbxw5 and Kif2c and truncations of Kif2c behaved inconclusive compared to full-length protein (Figure 2). We thus decided to change strategy by employing more sophisticated approaches in the future, such as Cryo-EM and H/D exchange.
In line with proteasomal degradation of Kif2c after ubiquitylation via SCFFbxw5, siRNA-mediated knockdown of Fbxw5 led to an increase of Kif2c levels in cells arrested in mitosis or in G1/G0 (Figure 3). In the latter, it accumulated at the basal body and impaired formation of primary cilia in a Kif2c-dependent manner, indicating that the regulation process is important for ciliogenesis. Furthermore, knockdown of Fbxw5 showed severe synthetic lethality with Kif2c overexpression, which is characterized by a prometaphase arrest. This finding suggests that cells expressing elevated Kif2c levels depend on a functional Fbxw5-dependent degradation pathway in order to prevent mitotic errors.


Using a combination of cycloheximide chase experiments and live-cell imaging, we were able to place the regulation of Kif2c by SCFFbxw5 to the G2 phase of the cell cycle (Figure 4). Attempts to identify the intracellular location of interaction and ubiquitiylation via split-FAST or proximity ligation assays were so far not successful, but will be continued in the future.
A potential mechanism behind the temporal confinement could be phosphorylation of Kif2c or Fbxw5 by kinases that are active only during G2. Previous studies implicated that the stability of Kif2c during mitosis is promoted by its phosphorylation via the kinase Plk1. However, adding recombinant Plk1 had no effect on Kif2c ubiquitylation. Next steps will be the systematic analysis of G2/M specific phosphorylations of both interaction partners, mutational analyses and the identification of responsible kinases.


During this project we could establish the following key points:
- Kif2c is a bona fide substrate of SCFFbxw5
o SCFFbxw5 ubiquitylates Kif2c very efficiently in vitro
o Fbxw5 and Kif2c interact strongly with each other
o Stability of Kif2c depends on Fbxw5
- Fbxw5-dependent regulation of Kif2c occurs during G2
o Kif2c becomes instable after G2 arrest, which is dependent on Fbxw5
o Kif2c levels at mitosic exit are significantly higher upon Fbxw5 knockdown
o The half-life of Kif2c at the basal body in G1/G0 does not differ between wt and knockdown cells
- The SCFFbxw5-dependent regulation of Kif2c is essential in cells overexpressing Kif2c and is required for proper ciliogenesis in unperturbed cells
o Knockdown of Fbxw5 is synthetic lethal with Kif2c overexpression
o Knockdown of Fbxw5 decreases amount of ciliated cells upon serum starvation in a Kif2c-dependent manner


This work has been presented during a poster session of the EMBO course on Methods in Cell Biology in Heidelberg September 2019, during an oral talk at a meeting of the DFG-funded Collaborative Research Center (SFB1036) in December 2019, and in the monthly seminar series of my host institution. Moreover, I showed the results of this work during a poster session at the ASCB meeting in Washington (USA) in December 2019, which has more than 3000 participants. Most importantly, I have started to write a manuscript for a peer-reviewed journal and after successful publication we will work on a review covering ubiquitin-dependent control of cytoskeletal elements.
In terms of public engagement, I introduced a group of pupils to our general laboratory work in 2018, participated in the “Nacht der Forschung” in Heidelberg 2019 and supervised 4 students during lab rotations on the UBIMAPS project.
In this project, we have established Kif2c as a bona fide substrate of SCFFbxw5, pinned down the ubiquitylation process to the G2 stage of the cell cycle and could further demonstrate its importance for cilia formation in the G0/G1 phase of the subsequent cell cycle.
In the near future, we would like to elaborate on two key points:
First, we will identify the mode by which Kif2c is bound by Fbxw5 using cryo-EM and H/D exchange. Second, we will work further on the medical implications of our findings. Given that Kif2c is often overexpressed in a variety of different tumors accompanied with poor patient prognosis, our finding that loss of Fbxw5 generates synthetic lethality with Kif2c overexpression in RPE-1 cells indicates a promising therapeutic potential. Thus, we would like to explore this observation by testing the effect of Fbxw5 knockdown and MCAK overexpression on the sensitivity of different cell types to cancer drugs that affect microtubule stability. Thereby, we may be able to generate a solid foundation for future research on combined and personalized treatment of cancer cells that overexpress Kif2c.
Validation of putative targets of SCFFbxw5 by interaction and in vitro ubiquitylation
Identification of binding sites on Kif2c for Fbxw5
Physiological relevance of Kif2c regulation by SCFFbxw5
SCFFbxw5 targets Kif2c during G2