A light-sheet microscope for imaging living biomimetics

The ability to accurately simulate the structure and function of organs is essential in the analysis of diseases, the modeling of human development, and the development of therapeutic treatments. As such, considerable efforts have been made towards creating 3D biomimetic cultures, which provide a more accurate representation of a cell's natural environment than 2D cultures. However, the intricate structure of 3D cultures makes them difficult to image. A method must be employed that enables deep imaging with a wide field of view while minimizing phototoxicity for long-term imaging.
Light-sheet microscopy (LSM) is an ideal imaging method for living samples as it provides optical sectioning, a wide field of view, and minimal phototoxicity. Lately, a new type of microscopy called lattice light-sheet microscopy (LLSM) has emerged. This technique involves the use of a light-sheet created by coherent superposition of Bessel beams, providing a compromise between light-sheet thickness and sidelobe energy density.
The MOSAIC (multimodal optical scope with adaptive imaging correction) microscope integrates LLSM with other microscopy modes into a single instrument. It was developed by Eric Betzig’s group in Janelia Research Campus. Furthermore, the MOSAIC is equipped with adaptive optics to correct for aberrations and provides benefits such as high axial resolution and minimal photobleaching and photodamage.
This thesis focuses on the construction of the LLSM imaging mode in the MOSAIC microscope, its optimization process, and its application for imaging 3D biomimetic cultures. The design behind the MOSAIC relies on modular custom-made components, making the building process straightforward. However, the lattice light-sheet (LLS) mode in the MOSAIC is complex due to the presence of other imaging modes that share components with the LLS optical path. This complexity can lead to an increase in optical aberrations in the instrument.
During the construction of the LLS, some complexities and issues were encountered that were resolved and documented. Most notably, these include design errors in the custom-made components, the absorption of fluorescence dye by the detection objective, and the shift in the light-sheet when aligning with water compared to medium.
Optimizations were carried out to ensure optimal imaging with the instrument. The optimizations include calibrating the spatial light modulator (SLM) for uniform excitation. Additionally, system aberrations are corrected with pupil wavefront correction (PWFC) using the SLM and phase retrieval algorithm using the deformable mirror. The instrument was also optimized for long-term imaging of living samples by incorporating temperature, CO₂, and evaporation controls in the sample chamber.
Three imaging experiments were performed that demonstrated our capability to image both 2D and 3D cultures, showcasing the versatility of the instrument in capturing a wide range of spatiotemporal dynamics, including subcellular vesicle movement and multi-cellular structure of a biomimetic skin culture.
The results showed that living samples can be maintained for a substantial time of 6 hours, but the image quality was impacted by cell degradation and bleaching. We found that the scanning direction of the sample is crucial in imaging 2D versus 3D samples. For 2D samples, a simple translation along the x-axis (horizontal) was sufficient. While for 3D samples, scanning along the xz-axis (detection objective optical axis) was necessary to image deeper structures. Although scanning along the xz-axis provides a limited field of view (FOV), a tiling technique was used to address it. The methods for creating 3D renderings from the images obtained by different scanning directions differ and require different pre-processing. Moreover, the sample mounting process for 3D samples was challenging but was accomplished using a fibrinogen/thrombin glue.
The PSF (point spread function) was measured in the LLS and compared to that of a spinning disk microscope. The measured full width at half maximum (FWHM) of the PSF in the LLS were found to be 325 nm, 316 nm, and 738 nm in the x, y, and z axes, respectively, matching the expected (theoretical) FWHM of 285 nm in the x and y axes and 700 nm in the z-axis.
Interestingly, because of the configuration of the objectives in the LLS, the PSF is rotated relative to the sample coordinates. This leads to an apparent improvement in z-resolution in the sample coordinates. This improved z-resolution was quantified by deriving an equation that indicates a resolution of 649 nm in the sample coordinates.
Tracking of clathrin-coated pits (CCPs) was employed as a demonstration of the wealth of data (e.g., position distributions and movement speeds) that can be extracted with 3D LLS imaging and subsequent image analysis.
The MOSAIC is a useful instrument for imaging 3D cultures, but its design still has some limitations that need to be addressed in future research and development. The results of this thesis provide valuable insights and information for using and improving the design of the MOSAIC microscope.