Abstract
Biological membranes, including plasma membrane (PM) and organelle membranes, restrict the flux of ions, molecules and organelles. However, the barrier function of biological membranes is frequently compromised by various perturbations, including physical membrane damage and protein- or chemical-induced pore formation. Recent evidence suggests that, upon PM damage, protein gelation and solid condensation are utilized to restrict ion/molecule/organelle flux across the damaged membranes by zoning the cytoplasm. In addition, membrane permeabilization dramatically alters intramembrane and extramembrane ion/molecule concentrations via the flux across the permeabilized membrane. The changes in ion/molecule concentration and their downstream pathways induce protein phase transition to form zones for biological processes or protein sequestration. Here, we review the mechanisms and functions of protein phase transition after biological membrane permeabilization.
Abbreviations
- DAMP
damage associated molecular pattern
- KCC
K-Cl cotransporter
- LLOMe
l-leucyl-l-leucine methyl ester
- LLPS
liquid-liquid phase separation
- NKCC
Na-K-2Cl cotransporter
- PM
plasma membrane
Cells and most membranous organelles are surrounded by lipid bilayers. While lipid bilayers are permeable to small hydrophobic molecules and partially permeable to small uncharged polar molecules, they are essentially impermeable to ions and large uncharged polar molecules (1,). The impermeability of the plasma membrane (PM) and organelle membranes to ions and uncharged polar molecules is fundamental to cellular integrity and homeostasis. However, the membrane impermeability is altered by various factors such as membrane lipid composition, temperature and pH (1,2,) (Fig. 1). Physical stresses and pore-forming proteins/chemicals assemble size- or ion/molecule-selective pores on the membranes (1,2,). According to the electrical and/or concentration gradients, ions and molecules are passively exchanged across the permeabilized membranes, leading to dramatic changes in intramembrane and extramembrane ion/molecule concentration. Additionally, cell rupture in hypotonic conditions causes severe cytoplasm outflows, generating shear (3). Recent studies have shown that these chemical and physical changes in the cytoplasmic/extracellular environment and their downstream pathways induce protein phase transitions and translocation of protein condensates. The protein phase transition and relocation of protein condensates enable the formation of cytoplasmic zones for specific biological processes, protein sequestration and regulation of cytoplasmic flows upon membrane permeabilization.
Protein phase transition upon membrane permeabilization. PM/organelle membrane permeabilization triggers protein phase transition via flux of ions/molecules across the membrane. Damage associated molecular patterns (DAMPs) that are released from PM-permeabilized cells induce protein phase transition in neighboring cells via receptor-mediated signaling. Ruptured cells release cytoplasmic prions, and neighboring cells are infected with the extracellular prions via endocytosis and endolysosomal ruptures.
Protein Phase Transition in Cellular Wound Healing
The PM is wounded by various stresses such as physical stimuli, pore-forming proteins, and surfactants, resulting in PM pore formation. The membrane pores freely pass ions, molecules and organelles smaller than the pores, which ultimately leads to cell death. Multiple mechanisms have been shown to repair the damaged PM to avoid cell death, which include exocytic membrane patch, endocytic internalization of the damaged membrane and ESCRT-mediated scission of the injured membrane (4).
Flux of intra/extracellular fluids is also mitigated by protein condensates. The protein condensates or assemblies of protein condensates and membranous organelles work as zones that physically restrict cytoplasmic flows. Filamentous fungi form multi-nucleated long hypha with connected cytoplasm. Because of their strong turgor pressure, cell rupture causes rapid cytoplasmic outflows. In Mucoromycete cells, the cytoplasmic outflows are mitigated by cytoplasmic gelation (3,) (Fig. 2A). The shear that is generated by cytoplasmic outflows triggers unfolding and fibrillar aggregation of Gellin proteins, and the Gellin phase transition and trap of membranous organelles into the aggregates gelate the cytoplasm. The cytoplasmic zones containing the Gellin aggregates and the trapped organelles are highly viscous, which reduces the cytoplasmic outflows from the ruptured PM (3).
Mitigation of cytoplasmic outflows by protein condensates. (A) Cytoplasmic gelation by Gellin aggregation. Due to high turgor pressure, cytoplasmic outflows generate strong shear. The shear unfolds Gellin proteins and induces Gellin aggragation. The assemblies of Gellin aggregates and membranous organelles reduces cytoplasm fluidities and mitigate cytoplasmic outflows. (B) Septal pore plugging by Woronin bodies. ATP efflux triggers Woronin body translocation to the septal pores. (C) Plugging of sieve plate pores by phloem protein fibres. (D) Sieve tube plugging by forisome swelling. Ca2+ influx from the ruptured PM induces forisome swelling. The swelled forisome plug the sieve tube cells and mitigate cytoplasmic flows.
In contrast to Mucoromycetes, Pezizomycotina in Ascomycota has septa in hyphae. In the absence of stress, the septa have pores, and cytoplasmic components are exchanged across the pores non-selectively (Fig. 2B). Upon hyphal injury, the septal pores are plugged by Woronin bodies, dense-core vesicles specific to Pezizomycotina, zoning the cytoplasm at septa to block cytoplasmic outflows (5,) (Fig. 2B). Woronin bodies are assembled from the peroxisomal membrane and surrounded by HEX-1 crystal lattice. While the shape of lipid bilayer is flexible, HEX-1 crystal-coated Woronin bodies have rigid structures, which would ensure the plugging even under high turgor pressures (6,). The plugging was suggested to be induced by the efflux of cytoplasmic ATP, while how the ATP depletion translocates the Woronin bodies is unclear (7,). In addition to HEX-1, several proteins form condensates around septal pores and contribute to Woronin body-dependent and -independent septal pore plugging and septation (8,9).
Pore plugging by protein condensates is also observed in sieve tubes of plants. The sieve tubes are constituted by stacks of sieve tube cells, and the holes between the cells allow ions and molecules to pass through. When sieve tube cells are injured by insect attack or physical stress, cytoplasm leaks out even in uninjured cells due to cytosolic connectivity. Upon the injury, fibrillar condensates of phloem proteins plug the holes between cells and mitigate cytoplasmic outflows (10,) (Fig. 2C). The cytoplasmic outflows from the sieve tubes were also suggested to be restricted by forisomes in Fabaceae (Fig. 2D). Forisomes are large fibrillar protein assemblies, and a high concentration of Ca2+ reversibly induces swelling of the forisomes without ATP hydrolysis (11,). The swelled forisomes block outflows of cytoplasmic components by plugging sieve tubes, although it is controversial (12,13).
Altogether, restriction of cytoplasmic outflows seems to be a general mode of action conducted by protein phase transition-mediated cytoplasmic zoning in multiple eukaryotic systems after PM rupture.
Non-healing Protein Phase Transition upon Molecule/Ion Flux across the PM
Protein phase transition also regulates non-healing biological processes upon ion/molecule flux across the PM. Osmotic stress alters human cell volume via ASK3 and WNK1 kinases-dependent mechanisms (Fig. 3A). WNK1 activates SPAK and OSR1 kinases, which phosphorylate Na-K-2Cl cotransporter (NKCC) and K-Cl cotransporters (KCCs) in the PM (14,). The phosphorylation of NKCC increases influx of Na+, K+ and Cl− from the extracellular fluids, though the phosphorylation of KCCs decreases efflux of K+ and Cl− from the cytosol. Therefore, active WNK1 increases cytosolic Na+, K+ and Cl− concentration, leading to water influx to the cytosol and cell volume increase. In contrast, ASK3 kinase works as a negative regulator of the cell volume increase by inactivating WNK1 (15,). Upon hyperosmotic shock, ASK3 and WNK1 form distinct liquid droplets, triggering recovery of cell shrinkage (16–18,). Although the phase-separated ASK3 is inactive, WNK1 droplets contain its substrates, SPAK and OSR1, and serve as zones for WNK1 signaling cascades. The liquid–liquid phase separation (LLPS) of ASK3 and WNK1 is triggered by enhancement of molecular crowding through efflux of cytoplasmic water upon hyperosmotic shock. The liquid condensates dissociate in isotonic conditions; thus water flux reversibly regulates phase separation of ASK3 and WNK1. While H2O is permeable across the lipid bilayers, aquaporin passes H2O more efficiently (19,). The H2O permeability is regulated by expression level, localization, and opening and closing of aquaporins (19,), suggesting that the aquaporin expression and structures affects ASK3/WNK1 condensation. The fluidity of ASK3 condensates is maintained by Na+ influx through PM-localized TRPM4 channels (18). TRPM4 is required for Na+ influx in earlier phase of hyperosmotic adaptation than WNK1-mediated NKCC activation. Inhibition of the Na+ influx from TRPM4 leads to the formation of irreversible solid ASK3 condensates, impairing cell size adaptation. Therefore, protein phase transition in osmoadaptation is regulated at least by H2O and Na+ flux across the PM.
Non-healing protein phase transition upon PM permeabilization. PM permeabilization triggers non-healing protein phase transition upon molecule/ion flux across the PM. (A) Cell volume regulation by WNK1 and SPAK/OSR1 droplets. Upon hyperosmotic shock, WNK1 forms lipid droplets with SPAK and OSR1 kinases. WNK1 activates SPAK/OSR1, which phosphorylate Na-K-2Cl cotransporters (NKCCs) and K-Cl cotransporters (KCCs) in the PM. Phosphorylation of NKCC increases the influx of Na+, K+, and Cl−. Phosphorylation of KCCs decreases the efflux of K+ and Cl−. The increase in cytosolic ion concentrations triggers water influx into the cytosol and cell volume increase. (B) Inflammasome formation by NLRP3 oligomerization. K+ efflux triggers NLRP3 oligomerization, which induces inflammasome assembly by recruiting ASC proteins, leading to self-processing and activation of pro-caspase-1. (C) Gln1 cytoophidium assembly. Upon proton influx, a glutamine synthase Gln1 forms cytoplasmic fibrillar assemblies. Gln1 condensation was proposed to save energy consumption. (D) Sup35 condensation by proton influx. Increase of cytoplasmic proton concentration induces phase transition of an essential translation termination factor, Sup35. Sup35 condensation protects its functional domain from irreversible aggregation under cytosolic acidification.
PM damage by pore-forming toxins triggers NLRP3 inflammasome assembly and pyroptosis via K+ efflux in myeloid cells (20,21,) (Fig. 3B). In the unstressed state, K+ concentration is higher in the cytosol than in the extracellular fluids. K+ efflux from the PM pores induces the decline of cytoplasmic K+ concentration, which triggers NLRP3 oligomerization (20–22,). NLRP3 oligomerization induces inflammasome assembly by recruiting ASC proteins. The recruited ASC form fibrillar assemblies that function as zones for self-processing and activation of pro-caspase-1, leading to pyroptosis and inflammation (20,23,). PM damage also triggers NLRP3 inflammasome assembly in neighboring cells. Extracellular ATP that is released from the injured cells open channels of P2X7 receptors in the neighboring cells, resulting in K+ efflux and NLRP3 inflammasome assembly (21,24,). Moreover, lysosomal rupture also induces K+ efflux and NLRP3 inflammasome assembly through cathepsin release to the cytosol, although how the cytosolic cathepsin induce cytoplasmic K+ efflux is unclear (20,25,26).
Budding yeast typically lives in a weak acid environment, such as wort and grape juice. Because yeast cytoplasm is maintained at neutral pH by PM and vacuolar proton pumps, proton concentration in the cytoplasm is often lower than that in the extracellular region (27,). Therefore, an increase in the proton permeability by protonophores causes proton influx into the cytoplasm (28,). Inhibition of proton pump by energy depletion also induces cytosolic acidification possibly because of proton leakage across lipid bilayers (28,29,). Cytosolic acidification triggers phase transition of dozens of proteins by altering their net charges and/or allosteric modulation (28,). For example, a glutamine synthase Gln1 forms cytoplasmic fibrillar assemblies, cytoophidia, upon proton influx (30,) (Fig. 3C). Cytoophidium assembly locks the protein structures to active or inactive forms, leading to hyperactivation or repression of the proteins (31,). Gln1 cytoophidia are inactive, and the Gln1 condensation was proposed to save energy consumption or regulates glutamine metabolisms under energy depletion (30).
Proton influx in yeasts also induces phase transition of an essential translation termination factor, Sup35 (32,) (Fig. 3D). Sup35 has a prion domain and forms inactive amyloidogenic prions (33). The prion domain of Sup35 also drives reversible LLPS and subsequent gelation of non-prion form of Sup35 upon cytoplasm acidification. Increase of cytosolic proton concentration triggers Sup35 LLPS by reducing charges of acidic amino acid clusters in the prion domain. The reversible Sup35 condensation protects its functional domain from irreversible aggregation under cytosolic acidification, thus the condensates function as zones for storage of functional Sup35.
Together, ion/molecule flux across the permeabilized PM trigger protein phase transition in the permeabilized cells and the neighboring cells. The protein condensates serve as zones for protein modification and processing, regulation of enzyme activity and storage of functional proteins.
Protein Condensation upon Organelle Membrane Permeabilization
Organelle membrane permeabilization also induces protein phase transition in the cytoplasm via ion/molecule flux. Macroautophagy, hereafter autophagy, degrades cytoplasmic components via sequestration into autophagosomes and subsequent delivery to lysosomes or vacuoles. The ER is considered to be a main source of membrane lipids for isolation membranes, later autophagosomal membranes (34,35,). Upon autophagy induction, the mammalian ULK complex is recruited to the ER and initiate autophagosome formation via ULK kinase signaling (35,). Ca2+ efflux from the ER Ca2+ channels triggers LLPS of the cytosolic FIP200, a component of ULK complex, on the ER surface, forming zones for the autophagosome initiation signaling (36) (Fig. 4A).
Protein phase transition upon organelle membrane permeabilization. Organelle membrane permeabilization triggers protein phase transition via ion/molecule flux across the membrane. (A) Ca2+ efflux from the ER triggers FIP200 condensation, which initiates autophagosome formation. (B) Endolysosomal membrane damage induces the influx of cytoplasmic ubiquitin ligases that ubiquitilate endolysosomal proteins. Autophagy adaptor protein p62 is recruited to ubiquitilated endolysosomal proteins and forms condensates, leading to autophagic clearance of damaged endolysosome. (C) Upon apoptotic stress, BAX/BAK proteins form macropores on the mitochondrial outer membrane, which triggers mitochondrial DNA (mtDNA) release. Released cytosolic mtDNA binds to cGAS to generate cGAMP, leading to STING activation.
The endolysosomal membrane is damaged by various factors, such as pathogenic bacteria, endocytosed amyloid fibrils, l-leucyl-l-leucine methyl ester (LLOMe) produced by monocytes. When the endolysosomal membrane is damaged, the membrane is repaired by ESCRT-mediated scission of the injury sites. Failure of the repair leads to autophagic clearance of the damaged endolysosome. The endolysosomal damage induces ubiquitilation of proteins in endolysosomal lumen via entry of cytoplasmic ubiquitin ligases from the injured membranes (37–39,) (Fig. 4B). The ubiquitilated proteins recruit autophagy adaptors, such as p62/SQSTM1 (39,). The recruited p62 forms condensates on the endolysosomal membrane, which is required for autophagic clearance of the damaged endolysosome (40).
Upon apoptosis induction, BAX/BAK proteins form macropores on the mitochondrial outer membrane (Fig. 4C). The BAX/BAK pores release mitochondrial DNA (mtDNA) to the cytosol, and the cytosolic mtDNA activates the innate immune response via cGAS/STING pathway (41,). The cytosolic DNA binds to cGAS protein and activates cGAS-mediated cGAMP production. The produced cGAMP binds to STING proteins and trigger the type-1 interferon signaling. The binding of DNA to cGAS induces the LLPS of cGAS-DNA, enhancing cGAMP production and downstream STING signaling (42).
Collectively, ion/molecule flux across the permeabilized organelle membranes also triggers protein phase transition, which is utilized for determination of the location of biological processes, as well as sensing of organelle membrane permeabilization.
Cell-to-Cell Transmission of Cytoplasmic Prions Via Membrane Ruptures
Prions self-propagate their misfolded conformation to native variants of the same proteins, leading to loss- and/or gain-of-function of the proteins. Many prions have structural features of amyloids represented by β-sheet-rich fibrillar aggregates. Amyloid formation alters protein activity by conformational change and/or aggregation-dependent decrease in accessibility to other molecules (43,). Furthermore, the interactors of amyloid-forming proteins and chaperones are also sequestered into the amyloids by co-aggregation (44,45). Thus, amyloidogenic prions work as cytoplasmic zones to modulate activity of native variants of the same proteins, their interactors, and chaperones.
Cytoplasmic prions are transmissible from cell to cell. While exosome- or intercellular nanotube-mediated prion transmission is achieved without membrane permeabilization, membrane pore formation also contributes to cell-to-cell prion transmission (46,). Upon PM rupture, prions are released to extracellular fluids, and the released prions are incorporated into other cells by endocytosis. Since endocytosed prions are surrounded by the endolysosomal membrane, ruptures of the endolysosomal membrane are required for their translocation to the cytosol (46,). Several pathogenic amyloids have potency to disrupt endolysosomal membrane, which would be critical for cell-to-cell transmission of amyloidogenic prions (47). While PM pores might also play a part in prion uptake, it is unclear whether amyloidogenic prions could pass non-lethal sized PM pores.
Cell-to-cell transmission of amyloidogenic prions has been suggested to be associated with various diseases including neurodegenerative diseases and cancers (46,48). Therefore, cell-to-cell prion transmission via membrane pore formation might contribute to progression of these diseases.
Conclusions
We here introduced the diverse functions of protein phase transition upon membrane permeabilization. Membrane permeabilization and damage induces various biological processes, such as cytokine production and release, exosome secretion and cellular senescence (49–51). The perspectives of protein phase transitions may provide better understandings of these processes associated with membrane permeabilization.
Funding
This work was supported by the JSPS KAKENHI (20H03440 to KK and SS; 22 J21762 to KS), and JST COI-NEXT (JPMJPF2205 to KK).
Conflict of interest
The authors declare that they have no competing interests.
Acknowledgments
We thank Yohsuke Moriyama, Yuta Yamazaki, and Nurhanani Razali for critical reading of the manuscript.
