Spine membrane topology produces anomalous diffusion. a, three-dimensional representation of Monte Carlo simulations.
Theoretical spines were produced by a vertical concatenation of a truncated sphere (head radius 450 nm) with a cylinder (neck radius 200 nm and neck length 750 nm)
placed in the middle of a flat field (4 × 4 µm, half displayed in the figure). The step size for spine shaping was 10 nm.
As a starting condition, particles (red dots) were randomly distributed in the flat field. The blue trajectory represents the successive positions of a particle during
a 775-ms run, starting in the flat field and ending in the spine head. Over time, particles fill the theoretical spine; see right panel (t = 30 s). b,
the black curve represents the number of particles in a spine (same features as in a) over time.
The red curve shows the corresponding fit (time constant t is 5.78 s; efficient diffusion coefficient is 0.023 µm2/s; time exponent: a = 0.88). c,
distribution of log(Deff/Dth) (where Deff = efficient diffusion coefficient and Dth = theoretical diffusion coefficient) for 256 experiments on flat
and stubby spines and 512 experiments on mushroom-shaped spines. In mushroom-shaped spines, Deff is 10-1000 times smaller than Dth
(medians: flat, -0.042; stubby, -0.271; mushroom-shaped, -1.41; stubby versus flat, p < 0.05; mushroom-shaped versus flat, p < 0.001). d and e,
log(Deff/Dth) plotted versus spine features, respectively, neck radius/neck length and neck radius/head radius. Data were grouped (bin sizes, 0.2 (d) and 0.1 (e))
and mean ± S.E. plotted.
This figure was published in "Dynamin-dependent Membrane Drift Recruits AMPA Receptors to Dendritic Spines",
Frédéric Jaskolski, Belen Mayo-Martin, David Jane and Jeremy M. Henley,
J Biol Chem. 2009 May 1; 284(18): 12491–12503.
doi: 10.1074/jbc.M808401200. Graphics a) and b) were not created using QtiPlot.