Dynamics of membrane tubulation coupled with fission by a two-component module

Abstract

Membrane tubulation coupled with fission (MTCF) is a widespread phenomenon but mechanisms for their coordination remain unclear, partly because of the lack of assays to monitor dynamics of membrane tubulation and subsequent fission. Using polymer cushioned bilayer islands, we analyze the membrane tubulator Bridging Integrator 1 (BIN1) mixed with the fission catalyst dynamin2 (Dyn2). Our results reveal this mixture to constitute a minimal two-component module that demonstrates MTCF. MTCF is an emergent property and arises because BIN1 facilitates recruitment but inhibits membrane binding of Dyn2 in a dose-dependent manner. MTCF is therefore apparent only at high Dyn2 to BIN1 ratios. Because of their mutual involvement in T-tubules biogenesis, mutations in BIN1 and Dyn2 are associated with centronuclear myopathies and our analysis links the pathology with aberrant MTCF. Together, our results establish cushioned bilayer islands as a facile template for the analysis of membrane tubulation and inform of mechanisms that coordinate MTCF.

Keywords: BAR domain–containing proteins; dynamin; membrane fission; membrane tubulation; polymer cushions.

Fig. 1.

Dynamics of BIN1-induced membrane tubulation. (A) Time-lapse images from Movie S1 of a bilayer island exposed to BIN1-mEGFP. The membrane is colored in gray and inverted in contrast. Fluorescence (B) and scanning electron microscopic (C) images showing BIN1 tubules on the bilayer island after 10 min. The membrane is colored in gray and inverted in contrast. Tubules are marked by white arrows. (D) Time-lapse images from Movie S2 showing coiling of BIN1 tubules. BIN1-mEGFP is colored in gray and inverted in contrast. (E) Time-lapse images from Movie S4 showing initiation and growth of BIN1 tubules, marked by white arrows. BIN1-mEGFP is colored in gray and inverted in contrast. (F) Plot showing growth rate of BIN1 tubules. Time is normalized to when tubules became apparent. Data represent the mean ± SD of nine tubules. (G) Plot showing the radius of BIN1 tubules. Data represent the mean ± SD radius of 37 separate and uncoiled tubules. (H) Time-lapse images from Movie S5 showing the formation of BIN1-mEGFP oligomers on the bilayer island. The membrane is colored in gray and inverted in contrast. (I) Fluorescence profile showing that the BIN1 oligomer, marked by the white arrow in (H), coincides with high membrane fluorescence. (J) Plot showing kinetics of BIN1 oligomerization and membrane tubulation fitted to a segmental linear regression equation. Data represent the mean ± SD of fluorescence intensities of five independent events.

Fig. 2.

CNM-linked BIN1 mutants are defective in membrane binding and tubulation. (A) Structure of the BIN1 N-BAR domain (PDB: 2FIC) rendered as a space filling model and colored based on electrostatics using ChimeraX (45). The structure marks the tip residues. (B) Fluorescence images showing the distribution of WT and the indicated mutants on a bilayer island. The membrane is colored in gray and inverted in contrast. White arrows in WT, R145C, and R154Q panels mark membrane tubules. Data represent the mean ± SD radius of 47 and 33 separate and uncoiled tubules for R145C and R154Q, respectively. (C) Plot showing membrane density of WT and CNM-linked BIN1 mutants measured as the BIN1-mEGFP to membrane fluorescence ratio. Data represent the mean ± SD of fluorescence ratio on 6 to 18 ROIs on multiple bilayer patches. Statistical significance was estimated using Mann–Whitney’s test where **** denotes P < 0.0001. (D) Membrane tubulation defined as the increase of membrane fluorescence over bilayer fluorescence for WT and mutants. Data represent the mean ± SD of fluorescence on 6 to 18 ROIs on multiple bilayer patches. Statistical significance was estimated using Mann–Whitney’s test where **** denotes P < 0.0001 and ** denotes P = 0.0016.

Fig. 3.

BIN1 and Dyn2 comprise a minimal two-component module that can manage MTCF. (A) Fluorescence image showing the distribution of Dyn2-mCherry on the bilayer island. Dyn2-mCherry is colored in yellow, and the membrane is colored in gray and inverted in contrast. White arrows mark preexisting buds on the island. (B) Plot showing membrane density of Dyn2 measured as the ratio of Dyn2-mCherry and membrane fluorescence on bilayers, buds, and tubules. Data represent the mean ± SD of 11 bilayer patches, 5 buds, and 9 tubules. Fluorescence (C) and scanning electron microscopic (D) images of a bilayer island exposed to BIN1 with excess Dyn2 and GTP. The membrane is colored in gray and inverted in contrast. (E) Time-lapse images from Movie S9 showing recovery after bleaching the fluorescent lipid probe in a large area of the bilayer island exposed to BIN1 with Dyn2 and GTP. The dotted line represents the boundary of the bleached region. The membrane is colored in gray and inverted in contrast. (F) Plots showing fluorescence recovery of the lipid probe on the bilayer (black) and on foci (red). Data represent the mean ± SD of fluorescence on 10 foci and the underlying bilayer. (G) Fluorescence image showing the distribution of BIN1ΔSH3-mEGFP and Dyn2-mCherry. Bilayers were imaged after 10 min incubation with 0.2 μM BIN1ΔSH3-mEGFP and 1.5 μM Dyn2-mCherry with 1 mM GTP followed by a wash-off with buffer. (H) Fluorescence image showing the distribution of BIN1-mEGFP and Dyn2-mCherry in presence of GTP. Bilayers were imaged after 10 min of incubation with 0.2 μM BIN1-mEGFP and 0.2 μM or 0.4 μM Dyn2-mCherry with 1 mM GTP followed by a wash-off with buffer.

Fig. 4.

Pathway to MTCF. (A) Time-lapse images from Movie S8 of bilayer islands exposed to the indicated concentrations of BIN1-mEGFP and Dyn2-mCherry with 1 mM GTP. White arrows mark tubular intermediates. (B) Time-lapse images of a bilayer island exposed to the indicated proteins in the absence of GTP. Black arrows mark single tubules and green arrows mark coiled regions. See Movie S10. (C) Plot showing the growth rate of BIN1 tubules under conditions described in (B). Data represent the mean ± SD of 10 tubules. (D) Time-lapse images of a bilayer island exposed to the indicated proteins in the presence of GTP. Black arrows mark single tubules, green arrows mark coiled regions and yellow arrows mark fission. See Movie S11. (E) Plot showing the growth rate of BIN1 tubules under conditions described in (D). Data represent the mean ± SD of six tubules. (F) Plot showing the fission time, defined as the time interval between appearance and fission of a tubule. Data represent the mean ± SD of 12 events.

Fig. 5.

Mechanistic basis for MTCF. (A) Fluorescence images of membrane nanotubes after being exposed to 0.1 μM Dyn2 with GTP for 10 min. The membrane is colored in gray and inverted in contrast. (B) Plot showing the fraction of nanotubes showing at least one cut in presence of 0.1 μM Dyn2 (black) or 0.1 μM Dyn2(S619L) (red) and 1 mM GTP with increasing concentrations of BIN1. Data represent the mean ± SD from six separate fields of nanotubes. (C) Results from a coupled liposome cosedimentation and PLiMAP experiment showing in-gel fluorescence (Fluor) and Coomassie brilliant blue (CBB) staining of Dyn2 and BIN1 in the supernatant (S) and pellet (P) fractions. (D) Plot showing densitometric (black) and fluorescence (red) quantitation of Dyn2 in the pellet fraction. Data represent the mean ± SD from two independent experiments. (E) Fluorescence images of membrane nanotubes after being exposed to 0.1 μM Dyn2(S619L) with GTP for 10 min. The membrane is colored in gray and inverted in contrast. (F) Time-lapse images from Movie S14 showing the bilayer island exposed to the indicated concentrations of proteins with 1 mM GTP. Insets show magnified images of intermediates adjusted in contrast for clarity. White arrows mark tubules and yellow arrows mark vesicles.