Work

Microtubule plus-end tracking protein EB1 orchestrates mitotic progression via liquid–liquid phase separation

Year: 2024

Type: article

Abstract: Dear Editor, Microtubule cytoskeleton is consisted of α/β-tubulin dimers, which assemble end-to-end for protofilaments to form the polar and hollow tube that is 25 nm in diameter (Gudimchuk and McIntosh, 2021). Microtubule dynamics normally present at the plus end, driven by β-tubulin-bound guanosine triphosphate (GTP) hydrolysis and the end-associated proteins, such as end-binding protein 1 (EB1). In eukaryotes, EB1 tracks the growing microtubule plus end, which is dependent on the presence of the β-tubulin-bound GTP nucleotide, to promote the microtubule growth and catastrophe frequency (Zhang et al., 2015). EB1 consists of a calponin homology (CH) domain for microtubule binding, a linker region (intrinsic disorder region), and an end-binding homolog (EBH) domain for dimerization. The dimerized EBH domains form a hydrophobic cavity to recruit the SxIP-containing microtubule plus-end tracking proteins (+TIPs), and the flexible C-terminal tail of EB1 contains an evolutionarily conserved glutamic acid–glutamic acid–tyrosine motif that is responsible for interactions with the CAP-GLY domain-containing binding partners (Gudimchuk and McIntosh, 2021). Several recent studies revealed that EB1 forms dynamic condensates, also called EB1 bodies, to mediate the proper kinetochore–microtubule attachment, microtubule shrinking, and cargo transportation in cells (Maan et al., 2023; Meier et al., 2023; Song et al., 2023). The evolutionarily conserved lysine and arginine residues in the intrinsic disorder region constitute molecular grammar for EB1 body architecture (Xia et al., 2012; Yang et al., 2024), and the number and position of these positively charged amino acids are essential for EB1 condensation. To explore the functional relevance of these amino acids in the regulation of EB1 phase separation by post-translational modifications, we combined published datasets with proteomic validation. As shown in Supplementary Table S1, the survey of the listed cancer-related databases revealed the presence of lysine modifications, such as lactylation, crotonylation, and acetylation. Among these potential modifications, the conserved K174 showed the greatest tendency of context-dependent acylation in tumor samples, especially of lactylation and crotonylation (Figure 1A). To evaluate the potential effects of the context-dependent acylation of K174 on EB1 phase separation, we introduced the K174Q substitution to mimic the alteration of the electric charge of Lys174 and used the K174R mutant as the control. EB1 K174Q attenuates its phase separation to modulate mitosis progression. (A) Schematic diagram of EB1 domain structure and K174 site and multiple alignment of EB1 protein sequences from different species (K174 labelled by blue star). (B) Characterization of recombinant purified proteins used in the in vitro phase separation assay. Recombinant EB1 proteins were subjected to electrophoresis followed by Coomassie Brilliant Blue staining. (C) Summary of in vitro phase-separation behavior of wild-type EB1 (EB1-WT) and K174Q mutant (EB1-K174Q). The red circles indicate phase separation (above saturation concentration), and the blue circles indicate no phase separation (below saturation concentration). (D) Representative in vitro LLPS images of EB1 proteins as visualized by fluorescence microscopy in 100 mM KCl buffer. Scale bar, 5 μm. (E) Quantitative analyses of the EB1 condensates area in D. For 2 μM EB1 protein, WT (n = 177), K174Q (n = 86), and K174R (n = 144). For 1 μM EB1 protein, WT (n = 156), K174Q (n = 85), and K174R (n = 100). (F) GST-EB1 pull-down assay. GST-EB1 was purified and absorbed on GST agarose as affinity matrix to isolate EB1-WT, EB1-K174Q, or EB1-K174R. The intermolecular interaction was assessed by sodium dodecyl sulphate–polyacrylamide gel electrophoresis and Coomassie Brilliant Blue staining. (G) HeLa cells expressing FLAG-EB1 were synchronized and released at G1/S and mitosis for immunoprecipitation. The EB1 lactylation levels were analyzed by western blotting with anti-pan-LacK antibody. (H) Schematic illustration of experimental design and expression of exogenous wild-type, K174Q mutant, or K174R mutant EB1-GFP in endogenous EB1-deficient HeLa cells (by EB1 shRNA) and representative real-time imaging of chromosome movements in the cells expressing EB1-WT-GFP, EB1-K174Q-GFP, or EB1-K174R-GFP. HeLa cells were synchronized and released for time series imaging as illustrated. Arrowheads indicate spindle poles; arrowheads indicate lagging chromosomes. Scale bar, 10 μm. (I and J) Statistical analysis of mitotic phenotypes in H. WT, n = 33; K174Q, n = 45; K174R, n = 21. Data in E, I, and J represent mean ± SEM from three independent experiments. Statistical significance was assessed by multiple t-tests. NS, P > 0.05; *P < 0.05; **P < 0.01. To test the effect of K174Q on EB1 phase separation, we purified GPF-tagged wild-type EB1 and the K174Q and K174R mutant proteins from bacteria (Figure 1B) and established phase diagrams by performing condensate formation assays with various concentrations of the protein and salt. Wild-type EB1-GFP underwent phase separation above 2 μM in the buffer containing 150 mM KCl, which is consistent with the previous study (Figure 1C; Supplementary Figure S1A; Song et al., 2023). Wild-type EB1 and the K174Q mutant exhibited different phase-separation abilities at 100 mM NaCl, with the critical protein concentration at 0.5 μM for wild-type EB1 but elevated to 1 μM for the K174Q mutant (Figure 1C and D). In addition, condensate size of the K174Q mutant was dramatically reduced when compared to that of wild-type EB1 or the K174R mutant (Figure 1E), supporting the importance of K174 in EB1 liquid–liquid phase separation (LLPS). To overcome the limitations of biochemical experiments and enable systematic evaluation of K174Q mutation on EB1 phase separation, we performed ultra-coarse-grained computational modeling. The modeling also showed a decrease in phase-separation propensity for the K174Q mutant compared to wild-type EB1 (Supplementary Figure S1B), which echoed the result from the in vitro phase separation assay. To model the thermomechanics and the dynamic changes of K174Q on phase separation, we established wild-type and K174Q condensates to depict the depolymerization curves over time. As shown in Supplementary Figure S1C and D, the maximum cluster size and deformation rate of K174Q condensates were more extensively perturbed compared to those of wild-type EB1 condensates, suggesting the perturbation on intermolecular interaction by K174Q mutation. To validate the above modeling results, a pull-down assay was conducted. As shown in Figure 1F, K174Q mutation weakened the intermolecular interaction of EB1 in vitro. Thus, we conclude that K174Q mutation attenuates the EB1 phase-separation propensity in vitro. To characterize the relationship among microtubule plus-end tracking activity, microtubule dynamics, and EB1 phase separation, we introduced wild-type EB1, the K174Q mutant, and the K174R mutant into COS-7 cells. In contrast to wild-type EB1 and the K174R mutant that exhibited a typical comet-like structure, the K174Q mutant presented as stunted comet in the expressing cells (Supplementary Figure S2A and B), which suggests that phase-separation propensity is essential for EB1 microtubule plus-end tracking activity in cells. To further evaluate the effect of EB1 phase separation and the K174Q mutation on microtubule dynamics, we performed total internal reflection fluorescence microscopy (TIRFM) with purified wild-type, K174Q, and K174R mutant EB1 proteins. As shown in Supplementary Figure S2C, wild-type EB1 and the K174R mutant promoted the microtubule growth rate and increased the catastrophe frequency, while the K174Q mutant attenuated microtubule plus-ends tracking activity and the regulation of microtubule plus-end dynamics. Thus, these results suggest that EB1 phase separation determines microtubule plus-ends dynamics and K174 is essential for the EB1 functions. EB1 serves as a dynamic platform to recruit most of +TIPs to microtubule plus-ends and regulate microtubule plus-end plasticity; the recruited +TIPs include SxIP-motif-containing proteins, such as TIP150, MCAK, and CLASP1/2, and CAP-GLY domain-containing proteins, such as p150Glued and CLIP-170 (Jiang et al., 2009; Gudimchuk and McIntosh, 2021). To exclude the possibility that K174Q mutation alters the ability of EB1 to bind to +TIP proteins, a pull-down assay was performed. As shown in Supplementary Figure S3, bindings of EB1-interacting partners, such as p150Glued, CLIP170, CLASP1/2, TIP150, and MCAK, to the K174Q mutant were comparable to wild-type EB1 and the K174R mutant, in agreement with previous reports that the C-terminus of EB1 serves as the interacting domain (Honnappa et al., 2009; Xia et al., 2012). To quantify the binding affinity between wild-type EB1 or the K174 mutants and the interacting +TIPs protein, we performed isothermal titration calorimetry (ITC) assays using purified MCAK and CLASP2 peptides, in which the equilibrium dissociation constant Kd was calculated. Our results showed comparable binding affinities for wild-type EB1, the K174R mutant, and the K174Q mutant (Supplementary Figure S4). To evaluate the functional relevance of K174 of EB1 in mitosis, we carried out an immunoprecipitation assay using wild-type and K174Q mutant EB1-GFP proteins from synchronized HeLa cells in mitosis. As shown in Supplementary Figure S5, wild-type EB1 and the K174Q mutant exhibited similar binding behaviors to MCAK, CLASP1, STIM1, or SKAP (Supplementary Figure S5). Thus, we conclude that attenuation of EB1 phase separation by K174Q mutation is irrelevant to the binding partners containing SxIP motifs. Given the fact that recent studies suggested an essential role of EB1 phase separation in mitosis (Meier et al., 2023; Song et al., 2023) and a high level of lactylation of EB1 was identified during mitosis (Figure 1G), we set to introduce the K174Q mutant into HeLa cells to evaluate its effect on mitotic progression (Figure 1H). To this end, HeLa cells, in which endogenous EB1 was depleted by an shRNA, were transfected with the shRNA-resistant wild-type EB1-GFP or the K174 mutants for live cell imaging. As shown in Figure 1I, control-transfected cells progressed through mitosis with accurate chromosome segregation. Mitotic cells expressing EB1-K174Q presented with noticeable mitotic defects, such as misorientation and misalignment (Figure 1I), consistent to our previous study (Song et al., 2023). Specifically, K174Q mutation significantly increased the number of cells exhibiting aberrant spindle orientation and chromosome alignment (Figure 1I and J). Therefore, we conclude that the EB1 phase-separation propensity and related chromosome segregation in mitosis are subjected to post-translational regulation. During cell division, subcellular architecture and compartments are reorganized to guide the quality control of cell renewal (Liu et al., 2020). Mounting evidence demonstrated that phase-separation function of SKAP is essential to govern the end-on capture of spindle microtubules in mitosis (Zhang et al., 2022). In addition, phase separation of spindle regulators such as BuGZ and NuMA is essential for stability and chromosome segregation (Jiang et al., 2015; Sun et al., 2021; Ma et al., 2022). Given the coordination of EB1 with NuMA complex in spindle position during mitosis (Song et al., 2021), it would be of great interest to see whether and how spindle plasticity regulators such as NuMA and BuGZ synergize with EB1 to guide spindle plasticity and cell division quality control. In addition, it is important to conduct systematic analyses of EB1 post-translational modification in follow-up studies to delineate how EB1 post-translational modifications orchestrate chromosome movements. In sum, our findings reveal that Lys174 serves as a regulatory niche for EB1 phase separation, which regulates dynamic microtubule plus-end hubs to orchestrate the robustness of mitotic spindle. These results highlight an uncharacterized molecular mechanism that underlies EB1 post-translational modification cross-talks in mitosis. We reason that post-translational regulation of LLPS-driven EB1 condensates provided a mechanism for the accurate and robust regulation of microtubule plus-end dynamics in space and time during cell division. [We thank our groups for critical discussion during the course of this study. This work was supported by grants from the Ministry of Science and Technology and the National Natural Science Foundation of China (2022YFA1303100-32090040, 32370772, 92253305, 32090043), Anhui Provincial Natural Science Foundation Grant (2408085J014, 202303a07020002, 202303a07020003), the Ministry of Education (IRT_17R102, 20113402130010; YD2070006001), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB19040000), and the Fundamental Research Funds for the Central Universities (WK2070000194). X.L., X.S., and X.Y. conceived the project. M.L. and M.D. designed and performed most of the biochemical and cellular experiments. X.H. and K.J. performed EB1 TIRF experiments in vitro. Z.H. performed molecular modeling. Y. Li, X.W., Y Lin, Y.Z., and T.U. assisted in recombinant protein engineering and purification. K.Z. contributed reagents. Z.W. edited the manuscript. X.S., X.L., and X.Y. wrote the manuscript. All authors have commented and approved the manuscript.]

Source: Journal of Molecular Cell Biology

DOI: 10.1093/jmcb/mjae059