We report on the recently emerging (laser) light-sheet-based fluorescence microscopy field (LSFM). The techniques used in this field allow to study and visualize biomedical objects nondestructively in high resolution through virtual optical sectioning with sheets of laser light. Fluorescence originating in the cross-section of the sheet and sample is recorded orthogonally with a camera. In this paper, the first implementation of LSFM to image biomedical tissue in three dimensions—orthogonal-plane fluorescence optical sectioning microscopy (OPFOS)—is discussed. Since then many similar and derived methods have surfaced, (SPIM, ultramicroscopy, HR-OPFOS, mSPIM, DSLM, TSLIM, etc.) which we all briefly discuss. All these optical sectioning methods create images showing histological detail. We illustrate the applicability of LSFM on several specimen types with application in biomedical and life sciences.
Serial (mechanical) histological sectioning (SHS) creates physical slices of fixed, stained, and embedded tissues which are then imaged with an optical microscope in unsurpassed submicrometer resolution. Obtaining these slices is however extremely work intensive, requires physical (one-time and one-directional) slicing and thus destruction of the specimen. A 2D sectional image reveals lots of histologically relevant information, but a data stack and its 3D reconstruction are even more essential for the morphological interpretation of complex structures, because they give additional insight in the anatomy. The SHS method requires semiautomatic to manual image registration to align all recorded 2D slices into order to get realistic 3D reconstructions. Often dedicated image processing of the sections is needed because of the geometrical distortions from the slicing.
A valuable alternative to achieve sectional imaging and three-dimensional modeling of anatomic structures can be found in the little known and relatively recent field of microscopy called (laser) light-sheet-based fluorescence microscopy or LSFM. These nondestructive methods generate registered optical sections in real-time through bio(medical) samples ranging from microscopic till macroscopic size. LSFM can reveal both bone and soft tissue at a micrometer resolution, thus showing a large amount of histological detail as well.
The first account of the LSFM idea was published by Voie et al. in 1993 and applied to image the inner ear cochlea of guinea pig [
Surprisingly, all these problems can be avoided by combining two old techniques. Voie was the first to combine the Spalteholz method of 1911 [
OPFOS utilizes yet a third method in conjunction with the two previous techniques when the specimen contains calcified tissue or bone. In this case, the calcium first needs to be removed before the Spalteholz procedure is applied. Bone cannot be made transparent, as the calcium atoms strongly scatter light.
Since 1993, many OPFOS-like derived methods were developed for tissue microscopy, all based on light sheet illumination. “LSFM” has become a broadly accepted acronym to cover the whole of these techniques. In the discussion, we will give a short overview of this OPFOS-derived LSFM microscopy family. First, we will explain in detail the specimen preparation and the optical arrangement of the original OPFOS setup. The remainder of this paper will serve to demonstrate some applications of OPFOS.
In most LSFM methods, the biomedical tissue samples are severely limited in size, though for instance the LSFM implementations of Ultramicroscopy, HR-OPFOS, and TSLIM (cf. the discussion section) are capable of imaging macroscopic samples up to tens of millimeters [ Euthanasia: living animals cannot be used in combination with clearing solutions. In general, LSFM is thus mainly used Perfusion: before dissecting a sample to the required dimensions, transcardial perfusion with phosphate buffered saline is useful as coagulated blood is difficult to clear with Spalteholz fluid [ Fixation: immersion in 4% paraformaldehyde (10% formalin) for 24 h or more for preservation and fixation of the specimen. Bleaching: optional bleaching in 5% to 10% hydrogen peroxide for one hour up to several days can be performed when the sample contains dark pigmented tissue (e.g., black skin and fish eyes) [ Decalcification: when the specimen contains calcified or mineralized tissue, such as cartilage or bone, decalcification is in order. A 10% demineralized water solution of dihydrate ethylenediaminetetraacetic acid (EDTA) slowly diffuses calcium atoms from the sample through a chelation process. Low-power microwave exposure (without heating) drastically accelerates the decalcification process from a month to several days [ Dehydration: immersion in a graded ethanol series (f.i., 25%, 50%, 75%, 100%, and 100% each for 24 h) removes all water content from the sample [ Hexane or benzene: the optional immersion in a graded series of hexane or benzene is said to improve dehydration further [ Clearing: to achieve large volume imaging in inherently less transparent samples, clearing is needed. The specimens are to be immersed in clearing solution, either through a graded series (f.i., 25%, 50%, 75%, 100%, and 100% each for 24 h) when the hexane or benzene step was skipped [ Staining: the required fluorescence can originate from auto-fluorescence from lipofuscins, elastin, and/or collagen [
In what follows, the original OPFOS setup is discussed as it was introduced by Voie et al. in 1993 [
The setup is represented in Figures
Schematic drawing of the (HR-)OPFOS setup: light from a green (GL) or blue laser (BL) passes through a Keplerian beam expander (BE) with spatial filter, a field stop (FS), and a cylindrical achromat lens (CL) which focuses the laser along one dimension within the transparent and fluorescent object (O). A two-axis motorized object translation stage (OTS) allows scanning of the specimen and imaging of different depths. The fluorescence light emitted by the object is projected onto a CCD camera by a microscope objective lens (OL) with fluorescence color filter (CF) in front. The focusing translation stage (FTS) is used to make the objective lens focal plane coincide with the laser focus.
A 3D setup representation of an (HR-)OPFOS setup with two-sided cylindrical lens sheet illumination and with two laser wavelengths (green and blue). The blue laser is active here.
An essential requirement for OPFOS is the generation of a laser light sheet. In practice, it is impossible to generate a perfect plane or sheet of light; however, using a cylindrical lens a sheet can be approximated. A Gaussian laser beam is first expanded and collimated by a Keplerian beam expander. The broadened beam then travels along the
The hyperbolic focus profile of a cylindrical lens is shown. OPFOS records 2D images in an approximated planar sheet defined by the confocal parameter zone
The height of the beam in
In summary, an OPFOS image has a slicing thickness
As a first illustration of the above described OPFOS setup, we show an application in hearing research of the middle ear [
An OPFOS cross-section of
2D virtual cross-sections (
A 3D OPFOS reconstruction of Gerbil showing a surface mesh of the stapes hearing bone, a blood vessel running through it, and the tensor tympani muscle attaching to the stapes head. The blood vessel wall and inner cavity are both separately modeled. Voxel size
In neurology, morphological brain atlases are a useful tool. To this end, sectional imaging with histological detail of mice (C57 black
Three OPFOS cross sections of
In morphological studies, functionality of a musculoskeletal system requires the visualization of both skeleton and muscles. For example, the authors gained insight into the feeding mechanisms of newly born seahorses (
3D reconstruction of the head of a one-day-old seahorse. The OPFOS image data is functionally segmented to study the morphology of the sternohyoideus muscle, cf. zoom (oblique view of the muscle). Natural autofluorescence of the head was achieved using 488 nm laser light sheet sectioning. Voxel size
Organogenesis and evolutionary morphology can benefit from OPFOS as well. The technique allows to discern the main structural elements of the head of an African clawed tadpole (
(a) A transverse OPFOS cross section through a tadpole head with indications of the different tissue types. (b) A 3D reconstruction of the entire functionally segmented OPFOS image data stack (sensory organs, muscles, cartilage and neuronal structures in different colors) (frontal view). Voxel size
(a) A photograph of the tadpole after bleaching. (b) The photograph is superposed with the OPFOS surface mesh of the tadpole head and body. (c) Color-coded functional segmentation of individual organs, cf. Figure
The elaborate specimen preparation required in OPFOS and other LSFM techniques is a major disadvantage. The method is considered nondestructive; however, dehydration removed all water content and decalcification did the same with calcium. It is clear that shrinkage is thus unavoidable and in the same order of magnitude as serial histological sectioning [
The accuracy of measurements based on OPFOS sections depends greatly on the quality of the transparency of the sample and thus on the bleaching, dehydration, and decalcification process. Dark or dense regions in the sample, remaining water content or calcium atoms, refract or scatter laser light, leading to out-of-focus illumination and blurring. Furthermore, the illuminating light sheet entering the sample from one side can be partially absorbed in dense regions resulting in loss of excitation light and fluorescence on the far side of the region. Remaining pigment or zones of less(er) transparency also create this kind of shadows. These stripes or shadow line artifacts are a typical drawback of OPFOS-like techniques. Solutions for these stripes have been implemented, cf. the following section.
Finally, it is important to keep the distance and the amount of refractive material constant between the laser light sectioning plane and the observation lens when sectioning different depths. By translating the refraction-index-matched sample within the Spalteholz-filled specimen chamber orthogonally to the light sheet [
Optical sectioning with a plane of light was initiated in 1903 by Siedentopf and Zsigmondy [
In 2007, Dodt et al. published a new LSFM setup in
Whenever using cylindrical lenses for light sheet generation, the resulting parabolic focus can only be approximated as a plane over a length described by the confocal parameter, cf. the section on OPFOS setup. The minimal beam waist thickness of the parabolic focus widens near the edges of the confocal parameter with a factor
In 2008, three new LSFM versions were developed. Holekamp et al. fixed the light sheet illumination unit to the observation objective [
Thin-sheet laser imaging microscopy (TSLIM) by Santi et al. incorporates many improved features of the previous devices [
Finally, Mertz and Kim developed the HiLo LSFM system [
The long-lasting lack of a commercial LSFM device is responsible for the many different implementations of the basic method and for the unfamiliarity of researchers with the technique in certain fields [
We have shown with several applications that the OPFOS (and derived) methods, better known as light-sheet-based fluorescence microscopy or LSFM, are a valuable addition for sectional imaging and three-dimensional modeling of anatomic structures. LSFM has the major advantage that the virtual slices are automatically and perfectly aligned, making it easy to generate 3D models from them. Microscopy techniques are either focusing on flexibility, imaging depth, speed, or resolution. LSFM has all these benefits according to device manufacturers and the LSFM scientific community. Specimens containing both bone and soft tissue and ranging from microscopic till small macroscopic in size can be studied with LSFM, with application in biomedical and life sciences. This microscopy method is relatively new, conceptually simple but powerful. Researchers can easily build their own setup, and even the first commercial devices are becoming available.
The authors thank A. Voie, P. Santi, and U. Schröer for sharing information on OPFOS, the LSFM field, and the commercial Ultramicroscope device. They gratefully acknowledge the financial support of the Research Foundation—Flanders (FWO), La Fondation belge de la Vocation, the University of Antwerp, and the GOA project (01G01908) of Ghent University.