three-dimensional gaussian psf  (MathWorks Inc)

Journal: Frontiers in Cellular Neuroscience

Article Title: Heterogeneity and Development of Fine Astrocyte Morphology Captured by Diffraction-Limited Microscopy

doi: 10.3389/fncel.2021.669280

Figure Lengend Snippet: Assessing fine astroglial morphology in diffraction-limited microscopy by measuring astroglial volume fraction (VF) and segment density. (A) Example of a single EGFP-expressing astrocyte imaged using diffraction-limited two-photon excitation (2PE) fluorescence microscopy (single focal plane through the soma). Note the blurry periphery representing small astroglial processes. (B) Left panel: example of an electron micrograph of the hippocampal neuropil with clearly delineated astrocyte fragments highlighted in purple (left panel). For further experimental details see Medvedev et al. . Middle panel: same section only showing astrocyte fragments. The fraction of the section that is occupied by astroglial segments is the product of their number (n segments ) and average size (s segments ) divided by the total area. Right panel: diffraction-limited microscopy was illustrated by applying a Gaussian filter with a FWHM of 500 nm. Astrocyte fragments cannot be clearly separated and counted. The fraction of occupied area/volume can be calculated by normalizing the average intensity (I average ) to the value corresponding to 100% (I 100% ). (C) An example of a single section at full resolution cutting through simulated astrocyte processes of various sizes and orientations in a simulated image stack (see text and methods). (D) Emulation of diffraction-limited fluorescence imaging of the image stack from (C) . Note that this is not a smoothed version of (C) but instead representative of diffraction-limited imaging of the entire stack. Also, note the now blurry and overlapping astrocyte processes. (E) The VF obtained from simulated diffraction-limited microscopy (I average /I 100% ) strongly depends on the original fraction of volume occupied by astrocyte processes (linear fit, n = 190 separate sets of astrocyte processes). (F) The number of detectable astrocyte segments was determined (see text and methods) in simulations of diffraction-limited microscopy (D) . The number of detected segments was smaller than the number of simulated astrocyte processes but showed a highly significant positive correlation (linear fit, same data set as in (E) . (G) Example of super-resolution expansion microscopy (ExM) of an EGFP-expressing astrocyte (single focal plane from a stack). Note the high level of detail. (H) Simulation of diffraction-limited microscopy of ExM data (same cell as in G ). Note the appearance of out-of-focus structures and blurring. (I) The number of astrocyte segments determined in ExM was lower after simulation of diffraction-limited imaging but strongly correlated with the number obtained directly from ExM data (linear fit, n = 10 independent experiments).

Article Snippet: The generated tissue blocks were then convolved with a three-dimensional Gaussian PSF with an FWHM of 0.45 μm in the x-y plane and 0.90 μm in the z-direction to simulate diffraction-limited imaging (MATLAB, Mathworks).

Techniques: Microscopy, Expressing, Fluorescence, Imaging