BIOCHEMICAL INSIGHTS INTO THE PROTEIN NUMA AND ITS BINDING PARTNERS BETWEEN MITOTIC SPINDLE AND NUCLEAR COMPARTMENTS

During my PhD program I worked on two Nuclear Mitotic Apparatus (NuMA)-related projects. The first one centred on the biochemical and structural characterization of the NuMA-dynein mitotic interaction and was published last year on Structure. The second project focused on the study of the largely unknown role of NuMA in the nucleus during interphase. Regarding this part, I got interesting details on the NuMA-53BP1 (p53-binding protein 1) interaction in the context of liquid-liquid phase separation (LLPS). In multicellular organisms, the proper organization of the mitotic spindle is essential for accurate cell division, tissue development and homeostasis. In vertebrate cells, the protein NuMA is a master regulator of mitotic spindle functions, implicated in spindle assembly and orientation, working together with the high molecular weight dynein-dynactin microtubule-motor complex. The domain structure of NuMA consists of an N-terminal globular domain, a central extended coiled-coil, and an unstructured C-terminal cargo-binding region. Whether NuMA is a dynein-dynactin activating adaptor is still not known. On these premises, the first part of my PhD project focused on the characterization of the NuMA-dynein binding interface, which I performed in collaboration with other members of the group. The crystal structure of the N-terminal head of NuMA (NuMA_1-153) revealed that it folds into a hook domain, a conserved feature of the Hook-family dynein-dynactin adaptors interacting directly with the Light Intermediate Chain (LIC) subunit of dynein. Pulldown assays performed with purified proteins indicated a direct interaction between NuMA_1-705 and LIC and identified four conserved residues in the NuMA hook domain that are crucial for LIC binding. Interestingly, sequence alignment between NuMA and known CC1-box containing dynein-dynactin adaptors revealed the existence of a CC1-box-like motif in the NuMA N-terminal coiled-coil domain (NuMA_365-376) that we demonstrated to be also implicated in contacting LIC. Thus, our studies identified two sites on NuMA’ N-terminus required for the interaction with a conserved hydrophobic helix in LIC1 C-terminus. Spindle positioning assays in human HeLa cells showed that these newly identified dynein-binding interfaces of NuMA are essential for correct mitotic progression. Collectively, these results support the notion that NuMA acts as a mitotic dynein-dynactin adaptor, forming complexes with similar topology to what observed for other known hook and CC1-box containing adaptors. In vertebrate cells, NuMA accumulates in the nucleus during interphase and contributes to the DNA damage response (DDR), negatively regulating the 53BP1 double strand break (DSB) repair function. The second part of my PhD project focused on the characterization of the NuMA-53BP1 binding interface. By co-immunoprecipitation (co-IP) experiments in human HEK293T nuclear extracts with anti-NuMA antibodies, I confirmed that endogenous NuMA interacts with 53BP1, and that this interaction is decreased upon DNA damage induction. Interestingly, analytical size-exclusion chromatography (SEC) experiments with purified fragments revealed that the C-terminus of 53BP1 (53BP1_1484-1972) interacts directly with the C-terminus of NuMA (NuMA_1821-2115). These are two intrinsically disordered domains, common to proteins that undergo LLPS, a mechanism conferring spatial and temporal regulation to biological processes. Since 53BP1 forms DNA damage foci, which are LLPS condensates promoted by its C-terminal disordered region, I tested whether also NuMA is involved in this mechanism. Interestingly, I found that NuMA_1821-2115 forms liquid droplets in vitro at 20 uM and physiological salt concentrations, promoted by electrostatic and polar interactions. By co-IP experiments in HEK293T nuclear extracts, I also detected an interaction of NuMA with the MT nucleator TPX2. Since TPX2 counteracts the 53BP1 DSB repair function during replication stress and undergoes LLPS, I hypothesized that NuMA could work with TPX2 in regulating the DDR by forming dynamic LLPS condensates. Surprisingly, by co-IP experiments, an interaction between NuMA and 53BP1 was also scored during mitosis, where 53BP1 is known to be part of the centrosome surveillance pathway, another condensate-associated regulatory process. Further studies are required to uncover the molecular basis and the functional role of the NuMA interaction with 53BP1 both in the DDR and in the centrosome surveillance pathway.