An evolutionarily nascent architecture underlying the formation and emergence of biomolecular condensates

Abstract

Biomolecular condensates are implicated in core cellular processes such as gene regulation and ribosome biogenesis. Although the architecture of biomolecular condensates is thought to rely on collective interactions between many components, it is unclear how the collective interactions required for their formation emerge during evolution. Here, we show that the structure and evolution of a recently emerged biomolecular condensate, the nucleolar fibrillar center (FC), is explained by a single self-assembling scaffold, TCOF1. TCOF1 is necessary to form the FC, and it structurally defines the FC through self-assembly mediated by homotypic interactions of serine/glutamate-rich low-complexity regions (LCRs). Finally, introduction of TCOF1 into a species lacking the FC is sufficient to form an FC-like biomolecular condensate. By demonstrating that a recently emerged biomolecular condensate is built on a simple architecture determined by a single self-assembling protein, our work provides a compelling mechanism by which biomolecular condensates can emerge in the tree of life.

Keywords: CP: Cell biology; CP: Molecular biology; biophysics; condensates; evolution; low-complexity regions; nucleolus; phase separation; scaffolds; self-assembly.

Figure 1.. TCOF1 structurally defines the nucleolar fibrillar center

(A) Transmission electron microscopy (TEM) of a HeLa cell nucleolus with the GC, DFC, and FC labeled. (B) Strategy for endogenous tagging of TCOF1 with the FKBP degron, degradation of TCOF1 with dTAG-13, and subsequent TEM. (C) TEM of WT HeLa cells and TCOF1-degron HeLa cells in DMSO and dTAG-13 treatment. Scale bar: 1 μm. Inset shows a close up of the region indicated (arrowhead). Inset scale bar: 0.2 μm. (D) Immunofluorescence of endogenous nucleolar markers across varying mEGFP-TCOF1 levels. Nuclei are outlined (dotted line). Scale bars: 5 μm. (E) Model of TCOF1 structurally defining the FC and how TCOF1 assembly impacts nucleolar sub-compartments. See also Figures S1–S3.

Figure 2.. TCOF1 displays single-component assembly in cells

(A) Schematic illustrating expected $\mathrm{\Delta }{\mathrm{G}}^{\text{transfer}}$ and dense phase size trends for multi-component or single-component assembly behavior. (B) Quantification of $\mathrm{\Delta }{\mathrm{G}}^{\text{transfer}}$ and dense phase size vs. average nuclear intensity for mEGFP-Fibrillarin. Cells with different expression levels are shown (left). Nuclei are outlined (dotted line). Scale bars: 5 μm. n = 66 nuclei for $\mathrm{\Delta }{\mathrm{G}}^{\text{transfer}}$, n = 81 for dense phase area. Logarithmic and linear fits with 95% confidence intervals are shown. (C) Same as (B) but for mEGFP-TCOF1. n = 46 nuclei for $\mathrm{\Delta }{\mathrm{G}}^{\text{transfer}}$, n = 167 for dense phase area. See also Figure S4.

Figure 3.. TCOF1 structurally defines the FC through self-interacting S/E-rich LCRs

(A) Quantification of $\mathrm{\Delta }{\mathrm{G}}^{\text{transfer}}$ vs. average nuclear intensity for WT TCOF1 and TCOF1 mutants with repeat region truncations. WT TCOF1 data are shown on each plot for clarity. n= 26 nuclei for WT TCOF1, n = 48 for TCOF1Δ83–1170, n = 25 for TCOF1Δ83–785, n = 27 for TCOF1Δ83–502, n = 24 for TCOF1Δ83–286. Logarithmic fits and 95% confidence intervals are shown. (B) Quantification of $\mathrm{\Delta }{\mathrm{G}}^{\text{transfer}}$ vs. average nuclear intensity for WT TCOF1 and TCOF1ΔSE83–785. n = 46 nuclei for WT TCOF1, n = 29 for TCOF1ΔSE83–785. (C) Localization of GC marker MPP10 with mEGFP-WT TCOF1 or mEGFP-TCOF1DSE. Nuclei are outlined (dotted line). Scale bars: 5 μm. Line profiles are of regions indicated by the finely dotted line for mEGFP (green) and MPP10 channels (red). (D) Fluorescence microscopy of co-assembly experiments in cells. Nuclei are outlined (dotted line). Scale bars: 5 μm. See also Figures S5 and S6.

Figure 4.. Emergence of the fibrillar center can be explained by emergence of TCOF1

(A) Analysis of presence of TCOF1 and FC-containing (tripartite) nucleoli in species (green). Tree (left) illustrates evolutionary relationships between species. (B) Schematic of approach to express TCOF1 in cells derived from zebrafish embryos. (C) Fluorescence microscopy of cells derived from uninjected zebrafish embryos or from zebrafish embryos injected with mEGFP-TCOF1 mRNA or mEGFP-TCOF1Δ83–1170 mRNA. Nuclei are outlined (dotted line). Scale bars: 2 μm. Line profiles are of regions indicated by the finely dotted line for mEGFP (green) and Fibrillarin channels (red). (D) Fluorescence microscopy of nucleoli of cells derived from zebrafish embryos injected with mCherry-zfMpp10 alone (top) to mark the GZ or co-injected with mCherry-zfMpp10 and mEGFP-WT TCOF1 (bottom). Scale bars: 1 μm. Line profiles are of regions indicated by the finely dotted line for mEGFP (green), Fibrillarin (red), and mCherry (blue). (E) Model for the emergence and underlying structure of the FC. See also Figure S7 and Table S1.