Chloride intracellular channel (CLIC) proteins function as fusogens

Abstract

Chloride Intracellular Channel (CLIC) family members uniquely transition between soluble and membrane-associated conformations. Despite decades of extensive functional and structural studies, CLICs' function as ion channels remains debated, rendering our understanding of their physiological role incomplete. Here, we expose the function of CLIC5 as a fusogen. We demonstrate that purified CLIC5 directly interacts with the membrane and induces fusion, as reflected by increased liposomal diameter and lipid and content mixing between liposomes. Moreover, we show that this activity is facilitated by acidic pH, a known trigger for CLICs' transition to a membrane-associated conformation, and that increased exposure of the hydrophobic inter-domain interface is crucial for this process. Finally, mutation of a conserved hydrophobic interfacial residue diminishes the fusogenic activity of CLIC5 in vitro and impairs excretory canal extension in C. elegans in vivo. Together, our results unravel the long-sought physiological role of these enigmatic proteins.

Fig. 1. CLIC5 directly interacts with membranes.

a Sample turbidity of liposome samples was measured as 410 nm absorbance (A₄₁₀) in the presence or absence of CLIC5. A representative image of the samples is provided (right). b Schematic representation of the liposome co-floatation assay (upper panel; Created with BioRender.com). Representative SDS-PAGE analysis and R18 fluorescence measurement are provided for each fraction, indicating the presence of CLIC5 in direct interaction with the liposomes. Molecular weight markers in kDa are indicated. c Control liposome co-floatation assay experiments. DOC2B interacts with the liposomes in a calcium-dependent manner, while 14-3-3 shows no liposomal association. Molecular weight markers in kDa are indicated. d, f Averaged fluorescence emission spectra following excitation at F₂₈₀ at the indicated incubation times for CLIC5 (d) and calcium-bound DOC2B (f). The peak at F₃₄₄ corresponds to tryptophan emission, while the peak at F₅₀₅ represents dansyl-PE emission due to the occurrence of FRET. e, g FRET ratio (F₅₀₅/F₃₄₄) at the indicated time-points for CLIC5 (e) and calcium-bound DOC2B (g). For all experiments, data are presented as mean ± SEM, n = 3–4 independent experiments. Two-sided student’s t test was performed for data analysis, ****P < 0.0001, ***P = 0.002, *P = 0.0137.

Fig. 2. CLIC5 increases the hydrodynamic radius of liposomes.

a Liposomes were incubated with 25 μM CLIC5, as described in the materials and methods section. Following 45 min of incubation, DLS analysis revealed that a population with a significantly increased radius emerged only in the presence of CLIC5. b–d Control experiments with calcium-bound DOC2B (b), apo DOC2B (c), and 14-3-3 (d). Only calcium-bound DOC2B resulted in an effect reminiscent of that of CLIC5, as expected.

Fig. 3. CLIC5 facilitates mixing between liposomal membranes.

a Dose-response analysis of liposomal membrane mixing by CLIC5 using the R18 fluorescence unquenching assay. The time-dependent fluorescence increase reflects membrane mixing between labeled and unlabeled liposomes. b pH-dependence of CLIC5-mediated lipid mixing. R18 unquenching was measured in the presence of 25 μM CLIC5 following exposure to different pH values as indicated. c R18 unquenching following calcium-dependent lipid mixing by DOC2B, serving as a positive control. d Dithionite scrambling assay. CLIC5 does not scramble NBD-PE labeled liposomes, as reflected by the plateau at 50% prior to detergent solubilization with Triton X-100. For all experiments, data are presented as mean ± SEM, n = 2–3 independent experiments.

Fig. 4. CLIC5 induces liposomal content mixing.

a Dose-response analysis of liposomal content mixing by CLIC5 using the carboxyfluorescein fluorescence unquenching assay. The time-dependent fluorescence increase reflects the dilution of carboxyfluorescein due to the fusion of loaded and unloaded liposomes. b DOC2B fails to induce liposomal content mixing in the carboxyfluorescein unquenching assay, regardless of the presence of calcium, consistent with its inability to form a fusion pore. c Liposomal content leakage analysis using the fluorescence monitoring of encapsulated NBD-glucose. No leakage from the liposome lumen following incubation with CLIC5 could be detected, indicating that the increase in carboxyfluorescein fluorescence (a) does not result from content leakage. For all experiments, data are presented as mean ± SEM, n = 3–4 independent experiments.

Fig. 5. CLIC5-F34D exhibits impaired membrane fusion capability.

a CLIC5-F34D shows reduced interaction with the membrane, as assessed by the co-floatation assay. Representative SDS-PAGE analysis and R18 fluorescence measurement are provided. Molecular weight markers in kDa are indicated. b Averaged fluorescence emission spectra following excitation at F₂₈₀ for CLIC5-F34D (red) and CLIC5-WT (black; as in Fig. 1d) after incubation of 75 min. The peak at F₃₄₄ corresponds to tryptophan emission, while the peak at F₅₀₅ represents dansyl-PE emission due to the occurrence of FRET. c FRET ratio (F₅₀₅/F₃₄₄) comparison between CLIC5-WT and CLIC5-F34D, reflecting reduced membrane interaction of CLIC5-F34D. d–f CLIC5-F34D exhibits a markedly reduced effect on liposomal diameter as assessed by DLS (d), membranal lipid mixing monitored by R18 fluorescence unquenching (e), and carboxyfluorescein-mediated liposomal content mixing (f). For all experiments, data are presented as mean ± SEM, n = 3–4 independent experiments. Two-sided student’s t-test was performed for data analysis, ***P = 0.001.

Fig. 6. CLIC5 undergoes structural rearrangements upon acidification.

a Crystal structures of CLIC5-Δloop-WT (left) and CLIC5-Δloop-F34D (right). 2F_o-F_c electron density maps, contoured at 1σ, are provided for position 34, and residues within 5 Å are shown as sticks. b The difference in deuteration levels at the indicated time points between pH 5.5 and 7.5 for CLIC5-WT (upper panel) and CLIC5-F34D (lower panel). Due to the difference in the exchange rate at different pH values (exchange 100x slower at pH 5.5 than at 7.5), time points of equal exchange (20 s or 60 s at pH 7.5 and 2000s or 6000 s at pH 5.5, respectively) must be compared. c The HDX difference between the pH levels tested at the final time point is mapped onto the structure of CLIC5 (PDB 6Y2H). d The difference in deuteration levels at the indicated time points between CLIC5-WT and CLIC5-F34D at pH 7.5 (upper panel) and pH 5.5 (lower panel). (e) The HDX difference between CLIC5-WT and CLIC5-F34D at the final time point is mapped onto the structure of CLIC5 (PDB 6Y2H).

Fig. 7. Excretory canal defects in the nematode C. elegans with point mutation F38D in the exc-4 gene.

a Logo plot of the fractional abundance of residues flanking CLIC5-F34D among human CLIC1-6 and C. elegans EXC-4. b Closeup view of CLIC5 F34 (left) and EXC-4 F38 (right), colored according to Consurf conservation score analyses. c Representative images of L4 stage worms expressing an excretory canal fluorescent marker (green). Control worm (upper panel) with an excretory canal extending from the pharynx to the tail, and exc-4 mutant worm (lower panel) with an excretory canal extending only to the vulva and harboring visible cysts along the pharyngeal region are shown. Yellow arrows indicate the end of the canal in each worm. Magnified pharyngeal regions are presented on the right. Bars = 50 μm (left panels) and 20 μm (right panels). d Outgrowth scores of control vs. exc-4 mutant worms. Significance was calculated with the two-sided Mann-Whitney U test, ****P < 0.0001. e Percentage of worms with cysts along the excretory canal in control and exc-4 mutant worms.

Fig. 8. Model for CLICs-induced membrane fusion.

CLICs are schematically represented as circles, composed of the TRX (blue) and α (red) domains. A triggering event (e.g., acidic pH or introduction of ROS) leads to increased exposure of the inter-domain interface, favoring membrane interaction and, finally, fusion. Created with elements adapted from BioRender.com.