Superresolution fluorescence microscopy for 3D reconstruction of thick samples.
Sangjun ParkWooyoung KangYeong-Dae KwonJaehoon ShimSiyong KimBong-Kiun KaangSungchul HohngPublished in: Molecular brain (2018)
Three-dimensional (3D) reconstruction of thick samples using superresolution fluorescence microscopy remains challenging due to high level of background noise and fast photobleaching of fluorescence probes. We develop superresolution fluorescence microscopy that can reconstruct 3D structures of thick samples with both high localization accuracy and no photobleaching problem. The background noise is reduced by optically sectioning the sample using line-scan confocal microscopy, and the photobleaching problem is overcome by using the DNA-PAINT (Point Accumulation for Imaging in Nanoscale Topography). As demonstrations, we take 3D superresolution images of microtubules of a whole cell, and two-color 3D images of microtubules and mitochondria. We also present superresolution images of chemical synapse of a mouse brain section at different z-positions ranging from 0 μm to 100 μm.
Keyphrases
- single molecule
- atomic force microscopy
- deep learning
- living cells
- high resolution
- optical coherence tomography
- convolutional neural network
- air pollution
- computed tomography
- energy transfer
- cell death
- cell therapy
- stem cells
- high throughput
- magnetic resonance imaging
- high speed
- machine learning
- magnetic resonance
- bone marrow
- cell free
- dual energy
- fluorescence imaging