Liquid-Liquid Phase Separation: Unraveling the Enigma of Biomolecular Condensates in Microbial Cells

Abstract

Numerous examples of microbial phase-separated biomolecular condensates have now been identified following advances in fluorescence imaging and single molecule microscopy technologies. The structure, function, and potential applications of these microbial condensates are currently receiving a great deal of attention. By neatly compartmentalizing proteins and their interactors in membrane-less organizations while maintaining free communication between these macromolecules and the external environment, microbial cells are able to achieve enhanced metabolic efficiency. Typically, these condensates also possess the ability to rapidly adapt to internal and external changes. The biological functions of several phase-separated condensates in small bacterial cells show evolutionary convergence with the biological functions of their eukaryotic paralogs. Artificial microbial membrane-less organelles are being constructed with application prospects in biocatalysis, biosynthesis, and biomedicine. In this review, we provide an overview of currently known biomolecular condensates driven by liquid-liquid phase separation (LLPS) in microbial cells, and we elaborate on their biogenesis mechanisms and biological functions. Additionally, we highlight the major challenges and future research prospects in studying microbial LLPS.

Keywords: biomolecular condensates; cellular noise; crowded environments; liquid-liquid phase separation; membraneless organelles; multivalent interactions.

FIGURE 1

Schematic view of a phase diagram. Phase separation is a function of molecular concentration under environmental conditions such as temperature, ionic strength, pH, etc. At a concentration below Csat, the system remains in the one-phase regime. As the concentration increases, two-phase regimes will coexist in the system, and the required concentration is effected by the environmental change as represented in the y-axis. Within the coexistence line (black), molecules often condense into smaller droplets and fuse into bigger droplets to lower the surface tension. These processes are usually reversible. When the concentration continuously increases, the droplets may irreversibly turn into gel-like or solid condensates.

FIGURE 2

A model for the control of biomolecular condensates. (A) Multivalent interactions that drive LLPS. Scaffold molecules (red) that undergo LLPS are in stoichiometric excess (often in a crowding environment) and enriched for defined modular domains or intrinsically disordered regions. Client molecules (green) are recruited by binding to the free cognate sites in the scaffold. The critical scaffold/client or scaffold/scaffold interactions include electrostatic, cation-π, and π-π contacts. (B) Model of yeast Taf14-mediated transcriptional condensate. The Taf14 protein contains two main domains, an N-terminal YEATS (Yaf9, ENL, AF9, Taf14, Sas5) domain (yellow) that recognizes lysine acylation modification, as well as a C-terminal ET domain (green) that is reported as a protein-protein interaction domain and recognizes peptide substrates. The disordered regions of Taf14 were predicted by PONDR (Xue et al., 2010). Taf14 works as a scaffold protein that promotes phase separation of condensates and concentrates different transcriptional machinery to form Taf14-containing complexes, thereby enhancing transcription efficiency (Chen et al., 2020). (C) Model of Caulobacter RNase E BR-body assembly. The domain architecture for the RNase E protein is shown, and the disordered regions were predicted by PONDR (Xue et al., 2010). The N-terminal catalytic DNaseI domain (blue) and C-terminal disordered regions (yellow and red) are highlighted. The disordered regions contain positive-charged patches (Arg-rich RNA binding sites, yellow) and negative-charged patches (facilitating multivalent interactions with RNA, red), causing self-assembly of BR-bodies into condensates through electrostatic interactions (Al-Husini et al., 2018, 2020).

FIGURE 3

Schematic illustrations of CO₂-fixing phase-separated liquid organelles in prokaryotic or eukaryotic cells. (A) Carboxysome-based Rubisco condensate found in the prokaryotic cyanobacterium Synechococcus elongatus PCC7942. As a scaffold protein, CcmM peptide (red) binds the Rubisco large subunit (RbcL, green) and Rubisco large subunit (RbcS, yellow) via salt bridges and van der Waals contacts to form the CcmM-Rubisco complex. It includes the condensates in the carboxysome covering with protein shells. As shown in the cryo-EM structure (6hbc, Wang et al., 2019), CcmM fills a pocket between the RbcL dimers and the loop of RbcS. (B) Pyrenoid-based Rubisco condensate found in the eukaryotic microalgae Chlamydomonas reinhardtii. The pyrenoid matrix is predominantly composed of Rubisco-EPYC1 complexes, forming by the multivalent interactions of EPYC1 peptide (orange) and Rubisco (green and yellow) (Freeman Rosenzweig et al., 2017; Wunder et al., 2018; Meyer et al., 2020; Barrett et al., 2021). Cryo-EM supported a structural model (7jsx, He et al., 2020), showing that EPYC1 binds close to the equator of the Rubisco cylinder and forms a codependent network of the specific low-affinity bonds (Mackinder et al., 2016; He et al., 2020). (C) Alignment of the Rubisco-binding regions from both CcmM and EYPC1 peptides by using Clustal Omega (Sievers and Higgins, 2021) and ESPript 3.0 (Robert and Gouet, 2014).

FIGURE 4

Overview of biological functions of biomolecular condensates in microbial cells. The image shown is representative of nine main functions of LLPS in microbial condensates, which could be further summarized into two categories: ^∗,condensates play a role in regulating metabolic flux. ^∗∗,condensates play a role in acting against noise and stress.