Dynamic clustering of dynamin-amphiphysin helices regulates membrane constriction and fission coupled with GTP hydrolysis

  1. Tetsuya Takeda  Is a corresponding author
  2. Toshiya Kozai  Is a corresponding author
  3. Huiran Yang  Is a corresponding author
  4. Daiki Ishikuro  Is a corresponding author
  5. Kaho Seyama  Is a corresponding author
  6. Yusuke Kumagai  Is a corresponding author
  7. Tadashi Abe  Is a corresponding author
  8. Hiroshi Yamada  Is a corresponding author
  9. Takayuki Uchihashi
  10. Toshio Ando  Is a corresponding author
  11. Kohji Takei  Is a corresponding author
  1. Dentistry and Pharmaceutical Sciences, Okayama University, Japan
  2. Kanazawa University, Japan
  3. JST, Japan
  4. Nagoya University, Japan
5 figures, 8 videos and 1 additional file

Figures

Figure 1 with 2 supplements
GTP hydrolysis is required and sufficient for membrane constriction by dynamin-amphiphysin helical complexes.

(A) Electron micrographs of lipid tubules induced by dynamin-amphiphysin helical complex before GTP addition (No GTP) and at different time points after addition of 1 mM GTP (GTP 1 s, GTP 5 s, GTP 10 s, GTP 30 s and GTP 1 min). More than thirty samples from three individual experiments were examined and representative images are shown. Scale bar is 200 nm. (B) Electron micrographs of lipid tubules induced by dynamin-amphiphysin helical complex without guanine nucleotide (No GTP) or with a transition states analogue of GTPase reaction, by adding 1 mM each of slowly hydrolyzable GTP analogue (GTPγS), nonhydrolyzable GTP analogue (GMP-PNP), GDP combined with vanadate (GDP + vanadate), GDP (GDP), or GTP (GTP) for 10 min. More than thirty samples from three individual experiments were examined and representative images are shown. Scale bar is 200 nm.

https://doi.org/10.7554/eLife.30246.003
Figure 1—figure supplement 1
Purified dynamin and amphiphysin forms ring-shaped complexes.

(A) SDS-PAGE of purified dynamin (Dynamin) and amphiphysin (Amphiphysin). (B) Negative staining electron micrographs of dynamin alone (Dynamin), dynamin and amphiphysin (Dynamin + Amphiphysin) and amphiphysin alone (Amphiphysin) incubated in cytosolic buffer with physiological salt concentration and PH (see Materials and methods). Scale bar is 100 nm.

https://doi.org/10.7554/eLife.30246.004
Figure 1—figure supplement 2
Stoichiometry dependency of dynamin and amphiphysin in ring complex formation, liposome tubulation and membrane fission.

Dynamin (2.3 μM) were mixed with 1.15 μM, 2.3 μM and 4.6 μM of amphiphsyin (1:0.5, 1:1 and 1:2 in molar ratio) and formation of the ring-shaped complexes (A), tubulation of liposomes (B) and membrane fission (C) were analyzed by EM. Scale bars are 100 nm and 500 nm for (A) and (B), (C), respectively.

https://doi.org/10.7554/eLife.30246.005
Figure 2 with 1 supplement
GTP hydrolysis induces clustering of dynamin-amphiphysin complexes on lipid nanotubes.

(A) Electron micrographs of a lipid nanotube (Nanotube) and those with dynamin-amphiphysin complexes before GTP addition (No GTP) and at different time points after GTP addition (GTP 1 s, GTP 20 s, GTP 30 s, and GTP 1 min). Clusters of dynamin-amphiphysin helical complexes are indicated (white brackets). More than thirty samples from three individual experiments were examined and representative images are shown. Scale bar is 100 nm. (B) Electron micrographs of lipid nanotubes after addition of dynamin-amphiphysin complexes without guanine nucleotide (No GTP) or with a transition states analogue of GTPase reaction, by adding 1 mM each of slowly hydrolysable GTP analogue (GTPγS), nonhydrolyzable GTP analogue (GMP-PNP), GDP combined with vanadate (GDP + vanadate), GDP (GDP) and GTP (GTP) for 10 min. More than 30 samples from three individual experiments were examined and representative images are shown. Clusters of dynamin-amphiphysin helical complexes are indicated (white brackets). The average pitch of helices in the clusters is 15.00 ± 2.2 nm (mean pitch ± s.e.m., n = 63 from 7 nanotubes) in GDP + vanadate, while the average pitch of the helical complexes is 20.0 ± 0.5 nm (mean pitch ± s.e.m., n=81 from 9 nanotubes) in No GTP control. Scale bar is 100 nm.

https://doi.org/10.7554/eLife.30246.006
Figure 2—source data 1

Measuring pitch size of dynamin-amphiphysin helices on the lipid nanotube without GTP and in the presence of GDP and vanadate in panel B.

https://doi.org/10.7554/eLife.30246.008
Figure 2—figure supplement 1
Stoichiometry dependency of dynamin and amphiphysin in cluster formation

Dynamin (2.3 μM) were mixed with 1.15 μM, 2.3 μM and 4.6 μM of amphiphsyin (1:0.5, 1:1 and 1:2 in molar ratio) and their structure on lipid nanotubes in the absence of nucleotides (no nucleotides) and in the presence of 1 mM GDP and vanadate (GDP-vanadate) were analyzed by EM. Clusters formed by a few helical turns of dynamin-amphiphysin complexes are indicated (white brackets). Scale bar is 100 nm.

https://doi.org/10.7554/eLife.30246.007
Figure 3 with 2 supplements
Dynamic clustering of dynamin-amphiphysin helices during GTP hydrolysis.

(A) HS-AFM images captured at 1 frame/s of dynamin-amphiphysin helical complexes on membrane tubules before (0 s and 21 s) and after GTP addition at different time points (42 s, 63 s, 84 s, 100 s, 115 s and 131 s). Dynamin-amphiphysin helices (arrowheads) are assembled into three distinct clusters (1, 2 and 3 at 131 s). The pitch of dynamin-amphiphysin helices on the lipid tubule was 22.0 ± 0.7 nm (mean pitch ± s.e.m., n = 36 from 3 time points) before GTP addition and 15.7 ± 0.3 nm (mean pitch ± s.e.m., n = 36 from 9 time points) after GTP addition. (B) Particle tracking of dynamin-amphiphysin helices before (5–21 s) and after addition of 1 mM GTP (38–54 s, 38–86 s and 38–118 s) from Videos 4 and 5, respectively. Particle tracking of the complexes in the cluster 2 (light blue, dark blue and magenta) and cluster 3 (red, yellow and green) are shown. (C) Dynamin-amphiphysin helices tend to constrict during clustering. Average heights before (0 ≤ t ≤ 100 s) and after clustering (101 s ≤ t) are 38.4 ± 0.2 nm and 36.5 ± 0.2 nm for cluster 1, 32.6 ± 0.2 nm and 32.6 ± 0.2 nm for cluster 2, 27.9 ± 0.1 nm and 26.8 ± 0.2 nm for cluster 3, respectively. The heights were measured from the substrate surface. The marks *** indicate p<0.001 and n.s. is not significant, respectively.

https://doi.org/10.7554/eLife.30246.009
Figure 3—source data 1

Measuring pitch size of dynamin-amphiphysin helices on the lipid tubule before and after cluster formation.

Pitch sizes of dynamin-amphiphysin helices before cluster formation (0, 3 and 6 s) and after cluster formation (111–119 s) were measured based on the Video 3 for panel A.

https://doi.org/10.7554/eLife.30246.012
Figure 3—source data 2

Measuring heights of dynamin-amphiphysin helices on the lipid tubule before and after cluster formation.

Average heights of dynamin-amphiphysin helices composing clusters 1, 2 and 3 before cluster formation (0 ≤ t ≤ 100 s) and after cluster formation (101 ≤ t s) were calculated based on Figure 3—figure Supplement 2—source data 1 for panel C.

https://doi.org/10.7554/eLife.30246.013
Figure 3—figure supplement 1
HS-AFM imaging of LUV and its tubulation by dynamin-amphiphysin complex.

(A) HS-AFM images of a LUV. Scale bar is 50 nm. (B) Lipid tubules induced from LUVs in the presence of dynamin-amphiphysin complexes. Scale bar is 200 nm.

https://doi.org/10.7554/eLife.30246.010
Figure 3—figure supplement 2
Dynaimin-amphiphysin helices constrict during cluster formation.

Maximum heights of each helices plotted as a function of time (left panels) and their average helix heights before (0–100 s) and after clustering (101sec-) are shown for clusters 1, 2 and 3 (A, B and C, respectively). ***: p<0.001, **: p<0.01, *: p<0.1, n.s.: not significant.

https://doi.org/10.7554/eLife.30246.011
Figure 3—figure supplement 2—source data 1

Measuring heights of dynamin-amphiphysin helices on the lipid tubule during cluster formation.

Height changes of each helices composing clusters 1, 2 and 3 were measured based on the Video 3. Average heights of each dynamin-amphiphysin helices composing clusters 1, 2 and 3 before cluster formation (0 ≤ t ≤ 100 s) and after cluster formation (101 ≤ t s) were also calculated.

https://doi.org/10.7554/eLife.30246.014
Membrane fission occurs at the protein-uncoated regions flanking dynamin-amphiphysin clusters.

(A) Clips of HS-AFM images captured at 0.42 frames/s showing membrane fission by dynamin-amphiphysin complexes (0 s, 125.3 s, 185.5 s and 227.5 s) in Video 6. Membrane fission occurred at flanking regions of a dynamin-amphiphysin cluster. Corresponding height profiles along the red line (shown in the 0 s image) passing through the two fission points (arrows marked with FP.1 and FP.2) are shown below, together with clustered dynamin-amphiphysin helical complexes (red arrowheads). (B) Height profiles at fission points (FP.1 and FP.2) over time before (Video 7) and after constriction (Video 6). The heights of the lipid tubules from the substrate surface were measured at the fission points. (C) Clips of HS-AFM images showing clustering dynamin-amphiphysin complexes and membrane constriction at flanking regions of the cluster (arrow). HS-AFM images are shown in pseudo color. Scale bar is 40 nm.

https://doi.org/10.7554/eLife.30246.020
Figure 4—source data 1

Measuring height changes of lipid tubules during constriction and fission by dynamin-amphiphysin helices.

Heights of the lipid tubule undergoing constriction and fission were measured along the red line in Figure 4A at 0 s, 125.3 s, 185.5 s and 227.5 s based on the Video 6 for panel A.

https://doi.org/10.7554/eLife.30246.021
Figure 4—source data 2

Measuring heights of fission points (FP.1 and FP.2) over time before (Video 7) and after GTP addition (Video 6) for panel B.

https://doi.org/10.7554/eLife.30246.022
Amphiphysin contributes to generation of uniformly-sized vesicles by controlling dynamin-amphiphysin clusters.

(A) Representative EM images of membrane vesicles generated by dynamin-amphiphysin complexes (Dynamin + Amphiphysin) or dynamin alone (Dynamin) after addition of GTP. Size distribution of generated vesicles are shown in the right panel. The average sizes of vesicles were 70.0 ± 0.6 nm (mean diameter ± s.e.m., n > 30, N = 3) for dynamin-amphiphysin complexes and 204.6 ± 1.1 nm (mean diameter ± s.e.m., n > 45, N = 3) for dynamin alone. Scale bar is 200 nm. (B) Representative EM images of membrane constriction induced by dynamin-amphiphysin complexes (Dynamin + Amphiphysin) and dynamin alone (Dynamin) in the presence of GDP and vanadate. Distribution of intervals between constriction sites (arrows) are quantified in the right panel. The average intervals of constriction sites induced are 150.3 ± 9.8 nm (mean intervals ± s.e.m., n = 25 from 7 tubes) by dynamin-amphiphysin complexes and 193.5 ± 15.8 nm (mean intervals ± s.e.m., n = 46 from 15 tubes) by dynamin alone. Scale bar is 200 nm. (C) Clustering of dynamin-amphiphysin complexes (Dynamin + Amphiphysin) and dynamin alone (Dynamin) on lipid nanotubes in the presence of GDP and vanadate. Clusters of dynamin-amphiphysin helices are indicated (white brackets). Distribution of cluster size were shown as scattered plot in the right panel. Average size of the clusters formed by dynamin-amphiphysin complexes and dynamin alone are 34.2 ± 1.7 nm (mean cluster size ± s.e.m., n = 36 from 7 tubes) and 59.3 ± 4.7 nm (mean cluster size ± s.e.m., n = 30 from 5 tubes) respectively. Scale bar is 100 nm.

https://doi.org/10.7554/eLife.30246.026
Figure 5—source data 1

Measuring diameters of vesicles generated by dynamin-amphiphysin complexes or dynamin alone after GTP addition for panel A.

https://doi.org/10.7554/eLife.30246.027
Figure 5—source data 2

Measuring distances between membrane constriction sites induced by dynamin-amphiphysin complexes and dynamin alone in the presence of GDP and vanadate for panel B.

https://doi.org/10.7554/eLife.30246.028
Figure 5—source data 3

Measuring size of clusters formed by dynamin-amphiphysin complexes and dynamin alone in the presence of GDP and vanadate for panel C.

https://doi.org/10.7554/eLife.30246.029

Videos

Video 1
HS-AFM imaging of a LUV.
https://doi.org/10.7554/eLife.30246.015
Video 2
HS-AFM imaging of a lipid tubules induced from LUVs by dynamin-amphiphysin complexes.
https://doi.org/10.7554/eLife.30246.016
Video 3
HS-AFM imaging of cluster formation by dynamin-amphiphysin helical complexes.
https://doi.org/10.7554/eLife.30246.017
Video 4
Particle tracking of dynamin-amphiphysin helical complexes before GTP addition (frames from 5 s to 21 s in Video 3).
https://doi.org/10.7554/eLife.30246.018
Video 5
Particle tracking of dynamin-amphiphysin helical complexes after GTP addition (frames from 38 s to 124 s in Video 3).
https://doi.org/10.7554/eLife.30246.019
Video 6
HS-AFM imaging of constriction and fission of lipid tubules by dynamin-amphiphysin complexes.
https://doi.org/10.7554/eLife.30246.023
Video 7
HS-AFM imaging of lipid tubules before constriction and fission by dynamin-amphiphysin complexes.
https://doi.org/10.7554/eLife.30246.024
Video 8
HS-AFM imaging of constriction and fission of lipid tubules by dynamin-amphiphysin complexes.
https://doi.org/10.7554/eLife.30246.025

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Tetsuya Takeda
  2. Toshiya Kozai
  3. Huiran Yang
  4. Daiki Ishikuro
  5. Kaho Seyama
  6. Yusuke Kumagai
  7. Tadashi Abe
  8. Hiroshi Yamada
  9. Takayuki Uchihashi
  10. Toshio Ando
  11. Kohji Takei
(2018)
Dynamic clustering of dynamin-amphiphysin helices regulates membrane constriction and fission coupled with GTP hydrolysis
eLife 7:e30246.
https://doi.org/10.7554/eLife.30246