Single Objective LSFM

Swept confocally aligned planar excitation (SCAPE) microscopy is a single objective light-sheet method that permits ultra-fast 3D imaging. We have developed a wide range of different SCAPE systems, capable of imaging structure and function across scales from sub-cellular to entire, freely moving organisms. We have also optimized SCAPE for challenging imaging applications such as neuronal activity in the awake, behaving mouse brain. In parallel, we have developed a range of dynamic image analysis methods to extract quantitative information from SCAPE datasets. I will describe our recent progress, applications and dissemination efforts.

Light-sheet microscopy provides considerable advantages for long-term volumetric imaging of living organisms. To achieve subcellular resolution and single-molecule sensitivity in common biological sample holders, including multi-well plates, we designed an epi-illumination SPIM (eSPIM) system with high spatial resolution and light collection efficiency. It is based on the Oblique Plane Microscopy (OPM) configuration, with a pair of mismatched objectives as the untiliting unit to ensure high detection numerical aperture. As a result, eSPIM has an identical sample interface as an inverted fluorescence microscope with no additional reflection elements. We demonstrated multicolor, fast, volumetric imaging of live cells and single-molecule super-resolution microscopy. Moreover, we imaged cells in parallel in multiwell plates over eight hours and recorded cellular responses to perturbations.

Biological processes often depend on the tight spatiotemporal coordination between cells across tissue-level length scales, extending over hundreds of microns in 3D. Functional understanding of such processes would be greatly aided by imaging tools that offer the combined speed and sensitivity needed to observe 3D cellular dynamics without compromising the normal biology. Light-field microscopy (LFM) is a fast, synchronous 3D imaging technique that captures both the 2D spatial and 2D angular information of light emitted from the sample, permitting computational reconstruction of the signal from a full volume in a single snapshot [1]. While the diffraction of light imposes a tradeoff between spatial resolution and z-depth coverage, LFM has successfully demonstrated synchronous imaging of depths covering hundreds of microns at cellular resolution, and more recently has achieved depths covering several microns at subcellular resolution. The full potential of LFM, however, has been hampered by conventional wide-field illumination where the entirety of sample is lit up, often beyond the volume of interest, creating significant extraneous background that degrades the image contrast, especially for thick and/or densely-labeled samples.

Light-sheet microscopy (also known as selective-plane illumination microscopy; SPIM) achieves high image contrast by illuminating only the optical plane of interest. Inspired by SPIM, we recently introduced an improved light-field-based imaging approach, termed selective-volume illumination microscopy (SVIM), where confining excitation to only the volume of interest reduces extraneous background, thereby sharpening contrast and effective resolution in thick (≥100 μm) living systems [2]. SVIM was implemented with two objective lenses: one to selectively illuminate the volume of interest, and a second objective, orthogonally aligned, to acquire the fluorescent light-field. This two-objective geometry requires that the sample be mounted within the mutual intersecting volume defined by the perpendicular objectives, complicating sample mounting and thus precluding wide-spread adoption.

We will present our latest development in SVIM that uses only one objective, eliminating the need for two orthogonally oriented objectives at the specimen, and greatly broadening its utility for biological research. By implementing an oblique one-photon illumination path or two-photon illumination, the volume of interest is selectively excited through the same objective used for high-numerical-aperture detection. This single-objective approach simplifies specimen mounting, yet still reduces extraneous background, resulting in improved contrast, resolution, and reconstruction quality over traditional LFM. Rapid synchronous imaging of multicellular volumes spanning hundreds of microns will be shown, including quantification of cellular structures and brain-wide neural activity in larval zebrafish, as well as imaging of multiple specimens mounted in a standard multi-well plate.

There are multiple large scale efforts, including the Human Cell Atlas and the Allen Brain Atlas, attempting to create spatial atlases of cell identity and location in entire organs. New highly-multiplexed molecular mapping methods that perform '-omics' using various readout strategies are continually evolving to achieve this lofty goal. Those strategies that rely on fluorescent microscopy require multiple rounds of repeated imaging of the same sample volume due to the spectral overlap of available fluorophores. Between sequential imaging rounds, an automated fluidics system performs chemical exchange of fluorescent labels between molecular targets. Post-acquisition, the obtained data are computationally decoded to determine the unique identity of individual molecules.

The requirements for sequential multiplexed molecular imaging using fluorescent microscopy are high numerical aperture (NA), large area 3D sample scanning for thin samples, multiple laser lines, a sealed sample chamber with fluidic input/output, and a fluidics controller. This unique set of requirements has led to a reliance on traditional fluorescent imaging methods, such as confocal or epi-fluorescence microscopy. To speed the imaging time for traditional methods, many groups intentionally under-sample in both the lateral and axial dimensions. Even with under-sampling, the imaging time dominates the total experimental time for larger samples.

Light sheet fluorescence microscopy is a natural choice to increase speed, sampling, and optical sectioning. However, sequential multiplexed molecular imaging's unique requirements rule out the existing single objective, oblique plane, open-top, and high NA dual objective light sheet approaches. A new light sheet approach that meets the stated requirements would transform the accuracy, speed, and application of fluorescent microscopy for sequential multiplexed molecular imaging. High-quality data generation would also enable advancements in computational decoding approaches, analogous to the computational advancements currently taking place in single-molecule localization microscopy.

Building on recent advancements in oblique plane microscopy (OPM [1]), we created a high NA OPM with an integrated fluidics unit for sequential multiplexed molecular imaging. Our approach translates the sample mounted in a fluidics chamber through a static light sheet and uses a solid immersion objective as the tertiary imaging objective [2]. We quantitatively evaluated our high NA OPM design's volume performance using calibration samples and a simplified forward model. Turning to real samples, we quantified the co-expression of RNA for the SARS-CoV-2 entry genes ACE2 and TMPRSS2 in alveolar type II cells across centimeter-scale fixed human lung tissue slices. As constructed, our high NA OPM can image ~1 cm x 1 cm x 50 μm x 1 color per hour at single-molecule resolution. The fluidic chamber optical window size places a practical constraint on the otherwise unconstrained lateral imaging size. We anticipate this increase in resolution, sampling, and speed versus traditional fluorescent microscopy methods will benefit spatial molecular and structural atlas efforts.