Self-interaction of NPM1 modulates multiple mechanisms of liquid-liquid phase separation.
Mitrea DM, Cika JA, Stanley CB, Nourse A, Onuchic PL, Banerjee PR, Phillips AH, Park CG, Deniz AA, Kriwacki RW. · 2018
Read Self-interaction of NPM1 modulates multiple mechanisms of liquid-liquid phase separation. by Mitrea DM, Cika JA, Stanley CB, Nourse A, Onuchic PL, Banerjee PR, Phillips AH, Park CG, Deniz AA, Kriwacki RW. (2018) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc5827731.txt.
Introduction
Mitrea and co-authors show one protein switching between two droplet modes. Nucleophosmin, or NPM1, helps build ribosomes in the nucleolus. It forms liquid droplets with RNA and partner proteins. The team finds it can also droplet with itself.
NPM1 carries floppy, charged regions that attract each other. That self-grip drives a second kind of phase separation. The two modes push against each other. Shape changes inside the protein flip which mode wins. The authors tie the switch to ribosome assembly flow. Readers get a neat case of one molecule holding two routes to the same droplet state.
Catalog record
| CREATOR | Mitrea DM, Cika JA, Stanley CB, Nourse A, Onuchic PL, Banerjee PR, Phillips AH, Park CG, Deniz AA, Kriwacki RW. |
|---|---|
| YEAR | 2018 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Self-interaction of NPM1 modulates multiple mechanisms of liquid–liquid phase separation By Diana M Mitrea, Jaclyn A Cika, Christopher B Stanley, Amanda Nourse, Paulo L Onuchic, Priya R Banerjee, Aaron H Phillips, Cheon-Gil Park, Ashok A Deniz, Richard W Kriwacki Published in: Nature Communications DOI: 10.1038/s41467-018-03255-3 ABSTRACT Abstract Nucleophosmin (NPM1) is an abundant, oligomeric protein in the granular component of the nucleolus with roles in ribosome biogenesis. Pentameric NPM1 undergoes liquid–liquid phase separation (LLPS) via heterotypic interactions with nucleolar components, including ribosomal RNA (rRNA) and proteins which display multivalent arginine-rich linear motifs (R-motifs), and is integral to the liquid-like nucleolar matrix. Here we show that NPM1 can also undergo LLPS via homotypic interactions between its polyampholytic intrinsically disordered regions, a mechanism that opposes LLPS via heterotypic interactions. Using a combination of biophysical techniques, including confocal microscopy, SAXS, analytical ultracentrifugation, and single-molecule fluorescence, we describe how conformational changes within NPM1 control valency and switching between the different LLPS mechanisms. We propose that this newly discovered interplay between multiple LLPS mechanisms may influence the direction of vectorial pre-ribosomal particle assembly within, and exit from the nucleolus as part of the ribosome biogenesis process. INTRODUCTION Membrane bilayers partition cells into a variety of compartments and organelles, including the cytoplasm, nucleus, endoplasmic reticulum, Golgi apparatus, and mitochondria, which perform specialized biological functions. Cells are further compartmentalized through formation of membrane-less organelles (MLOs) such as P bodies and stress granules in the cytoplasm and nucleoli, Cajal bodies and nuclear speckles in the nucleus1–3. Specific sets of macromolecules self-assemble to form MLOs through a process termed phase separation, creating highly dynamic, condensed partitions with specialized functions1. Proteins and/or ribonucleic acids (RNAs) termed scaffold molecules4 drive phase separation into dense liquids or hydrogels that create compositionally variable microenvironments that enable spatial and temporal control of biochemical processes2. The resulting condensed scaffolds recruit client4 proteins and nucleic acids to modulate the biochemical processes localized within these MLOs. Temporal and spatial organization of protein, deoxyribonucleic acid (DNA), and RNA macromolecules within the nucleolus orchestrate complex biological processes, including ribosome biogenesis5,6, stress signal integration6,7, and regulation of gene transcription8,9. Spatial organization is achieved through sub-compartmentalization into coexisting, immiscible liquid phases10. Transcription of pre-ribosomal RNA at active nucleolar organizer regions (NORs)11,12 lowers the critical concentration for phase separation of nucleolar proteins, such as Fibrillarin and Nucleophosmin (NPM1)10,13,14, and nucleates11 assembly of the dense fibrillar component (DFC, Fibrillarin-rich) and, subsequently, the surrounding granular component region (GC, NPM1-rich). The immiscibility and spatial separation of these nucleolar sub-compartments is dictated by differences in their viscoelastic properties, especially in the surface tensions of the Fibrillarin- and NPM1-rich phases with respect to the surrounding nucleoplasm10. Genetic ablation or mRNA knock-down of nucleolar protein markers, such as Ki-6715, NPM116,17, nucleolin18, and fibrillarin19,20, alter nuclear and/or nucleolar morphology and disrupt ribosome biogenesis. Interestingly, though, depletion of any one of these aforementioned proteins does not prevent assembly of the nucleolus, suggesting that multiple scaffolding mechanisms underlie nucleolar assembly and possibly other membrane-less bodies. We previously demonstrated that localization of NPM1 within nucleoli in live cells depends upon its ability to independently undergo liquid–liquid phase separation (LLPS) with (1) nucleolar proteins displaying multivalent arginine-rich linear motifs (R-motifs) and (2) ribosomal RNA (rRNA). The first of these mechanisms is driven by interactions between acidic tracts (A-tracts) in the N-terminal oligomerization domain (OD; A1-tract)13,21 and intrinsically disordered region (IDR; A2- and A3-tracts)13 of NPM1 and multivalent R-motifs within partner proteins, while the second relies upon binding of the C-terminal nucleic acid-binding domain (CTD) and the adjacent basic, disordered segment (B2-tract) of NPM1 to rRNA13,22–24. Figure 1a, b illustrates the clustered charges and the sub-domain organization in NPM1. Either of these heterotypic mechanisms support LLPS in vitro. In this work we hypothesized that these two mechanisms are not mutually exclusive, but operate simultaneously within the nucleolar matrix. We further hypothesized that interactions between acidic- and basic-tracts within the IDR (NPM1IDR) can modulate their accessibility for inter-molecular interactions with phase separation partners. We tested these hypotheses using in vitro phase separation assays as well as a variety of structural methods with wild-type NPM1 (NPM1WT) and mutants that either lacked critical domains/regions or had mutations within them. Our results show that, in fact, intra-IDR interactions do modulate LLPS by both heterotypic mechanisms, and, unexpectedly, that inter-IDR interactions mediate homotypic LLPS by NPM1 under conditions that reflect physiological crowding25. The model that emerges is that NPM1 contributes to the liquid-like features of the GC region of the nucleolus through both heterotypic and homotypic LLPS mechanisms. This multiplicity of LLPS mechanisms may enable NPM1 to “buffer” the liquid-like features of the nucleolus as the process of ribosomal subunit assembly progresses, altering the accessibility of ribosomal proteins and RNA for interactions with NPM1. [Figure: Electrostatic interactions drive NPM1:SURF6-N LLPS. a Net charge per residue distribution using CIDER (http://pappulab.wustl.edu/CIDER/analysis/); top: the central intrinsically disordered region (IDR) in NPM1 exhibits polyampholityc properties; IDR contains two strong acidic tracts (A2 & A3) interleaved with moderately charged basic tracts (B1 & B2); bottom: the intrinsically disordered nucleolar protein SURF6 contains multivalent R-motifs throughout its primary structure. Data are shown for residues 1–182, corresponding to the construct used in this study (SURF6-N); b Schematic representation of the NPM1 constructs used in this study; c Confocal microscopy images of phase separation by 20 µM NPM1WT with 20 µM SURF6-N in buffer with variable concentrations of NaCl, as indicated. Screening of electrostatic interactions at high [NaCl] disrupts LLPS droplets. NPM1 is labeled with AlexaFluor 488 and SURF6-N with AlexaFluor 647. Scale bar, 10 µm] RESULTS NPM1 and SURF6-N form heterotypic liquid-like droplets NPM1 is a multifunctional protein, which interacts with over 130 proteins26, many of them annotated to be localized within the nucleolus and shown to display multivalent R-motifs13. One of these, Surfeit locus protein 6 (SURF6), is a non-ribosomal protein that co-localizes with NPM1 in the granular component of the nucleolus; genetic deletion of SURF6 disrupts ribosome biogenesis27 and reduces cell viability28. While the specific role of SURF6 in the nucleolus is unknown, it was proposed to serve as a scaffold within the nucleolar matrix29, albeit prior to the discovery in 2011 that nucleoli have liquid-like features30. SURF6 is predicted to be disordered and displays multivalent R-motifs within its sequence (Fig. 1a and Supplementary Fig. 1d, e). A N-terminal segment of SURF6, spanning residues 1–182 (SURF6-N) exhibits the 2D NMR signature of a disordered protein (Supplementary Fig. 1c) and undergoes heterotypic LLPS with NPM1. The condensation mechanism is strongly dependent on electrostatic interactions, which can be screened by increasing the solution ionic strength (Fig. 1c). We hypothesized that electrostatic complementarity of acidic-tracts within the IDR of NPM1 and R-motif-containing basic-tracts in SURF6-N drive heterotypic LLPS. We tested this hypothesis through studies of phase separation by NPM1 deletion mutants (Fig. 1b) with SURF6-N. C-terminal basic segment of NPM1 influences heterotypic LLPS We prepared two truncation mutants, one that lacks the CTD which is required for DNA and RNA binding22–24, NPM1N240, and another that lacks the CTD plus the B2-tract, NPM1N188 (Fig. 1b). While its affinity for RNA is unknown, the folded CTD binds to G-quadruplex DNA sequences with affinity in the micromolar range22,23,31 and the disordered B2-tract enhances affinity and binding kinetics23,31,32. We determined the threshold