Dual-view oblique plane microscopy for cleared tissue imaging

Session

Session 6 - Imaging Across Scales

Authors

Liuba Dvinskikh (3, 4), Hugh Sparks (3), Raffaele Sarnataro (1), Darren Ennis (4), David Carreno Yugueros (2), Paula Cunnea (4), Iain McNeish (4), Chris Dunsby (3, 4)

Affiliations

1. Centre for Neural Circuits and Behaviour, University of Oxford
2. Department of Infectious Diseases, Imperial College London
3. Department of Physics, Imperial College London
4. Department of Surgery and Cancer, Imperial College London

Keywords

oblique plane microscopy, light-sheet fluorescence microscopy, cleared tissue imaging

Abstract text

The optical sectioning of light-sheet fluorescence microscopy offers a non-destructive alternative to the physically-sectioning histopathology-based techniques for visualizing the 3D tissue architecture. However, scattering in biological samples due to non-uniform refractive index distribution limits the imaging depth. Optical clearing [1] aims to reduce the refractive index mismatch between biomolecules and the surrounding medium, with the final refractive index of cleared tissue typically ranging between 1.38-1.56 [1]. 

We present a dual-view oblique plane microscope (dOPM) [2] system, based around a silicone immersion primary objective for better imaging of optically cleared tissues compared to using a water immersion primary objective. The doublets combined to form the second tube lens in the remote refocusing relay were selected using stock optics tube lens design software [3]. The spatial resolution was characterized using a series of fluorescent bead phantoms adjusted to various refractive indices. Within the central 200 µm × 200 µm region of the overlapping dual-view fused volume, the average raw bead image FWHM was measured to be 0.47 ± 0.06 µm, 0.59 ± 0.06 µm and 1.2 ± 0.1 µm in X,Y and Z respectively for a refractive-index matched bead sample, showing minimal degradation over a >250 µm thickness (Fig. 1a). We further characterize how the remote refocusing performance at further refractive index mismatch (up to n = 1.5) can be partially compensated for using the adjustable correction collar.  

We apply the system to imaging a range of biological samples. Combined with tiled acquisition and image stitching, the microscope enables multicolour imaging of immunofluorescence labelled mm-wide and ∼300 µm thick cleared mouse ovarian cancer and gut tissue samples with subcellular resolution. We also image whole Drosophila melanogaster fruit fly brains (Fig. 1b), achieving a ∼7-10x fold speed improvement compared to a standard confocal imaging technique employed. The system offers a platform for fast and high-resolution, multicolour volumetric imaging across spatial scales, integrated into a commercially available base.

Figure 1: (a) X-, Y- and Z-FWHM of 1D Gaussian fits for images of 170 nm fluorescent beads, binned over each ∼20 µm slice across a ∼250 µm imaging range, for a total of 3157 beads. Error bars represent standard deviation. (b) Depth-encoded projection of a Drosophila melanogaster fly brain expressing mitochondria-localized GFP, stitched from 6 tiles. Scalebar: 100 µm

References

[1] Richardson et al. (2015) Cell 162 (2) DOI: 10.1016/j.cell.2015.06.067

[2] Sparks et al. (2020) Biomed. Opt. Express 11 (12) DOI: 10.1364/BOE.409781

[3] Hong et al. (2022) Opt. Express 30 DOI: 10.1364/OE.450320