Imaging the kidney: from light to super-resolution microscopy

Abstract

The important achievements in kidney physiological and pathophysiological mechanisms can largely be ascribed to progress in the technology of microscopy. Much of what we know about the architecture of the kidney is based on the fundamental descriptions of anatomic microscopists using light microscopy and later by ultrastructural analysis provided by electron microscopy. These two techniques were used for the first classification systems of kidney diseases and for their constant updates. More recently, a series of novel imaging techniques added the analysis in further dimensions of time and space. Confocal microscopy allowed us to sequentially visualize optical sections along the z-axis and the availability of specific analysis software provided a three-dimensional rendering of thicker tissue specimens. Multiphoton microscopy permitted us to simultaneously investigate kidney function and structure in real time. Fluorescence-lifetime imaging microscopy allowed to study the spatial distribution of metabolites. Super-resolution microscopy increased sensitivity and resolution up to nanoscale levels. With cryo-electron microscopy, researchers could visualize the individual biomolecules at atomic levels directly in the tissues and understand their interaction at subcellular levels. Finally, matrix-assisted laser desorption/ionization imaging mass spectrometry permitted the measuring of hundreds of different molecules at the same time on tissue sections at high resolution. This review provides an overview of available kidney imaging strategies, with a focus on the possible impact of the most recent technical improvements.

Keywords: immunohistochemistry; kidney biopsy; podocytes; proximal tubule; stem cells.

FIGURE 1

(A) 3D reconstruction of a glomerulus from a 50-μm-thick kidney section, showing de novo podocyte regeneration in a transgenic mouse model of adriamycin-induced nephropathy. The fluorescent reporter GFP (green) labels Pax2+ renal progenitors, whereas Tomato Red labels all other cell types. An x-plane (red) has been added to virtually dissect the glomerulus in two parts. In the upper part, only green signal is shown; in the lower part, all three colours are shown. (a’) Representative 2D z-section stacks of the 3D reconstruction shown in (A). Asterisks indicate the different number of Pax2+ progenitor-derived podocytes counted inside the glomerular tuft in each 2D image. The z-stack analysis used to perform the 3D reconstruction of the glomerulus showed that the detection of regenerated podocytes was highly dependent on which section was used for 2D analysis and its thickness. (Adapted from Romoli et al. [6]. This article is licenced under a Creative Common Attribution 4.0 International License. http://creativecommons.org/licenses/by/4.0/.) (B) Representative image of a glomerular capillary loop as seen with TEM. BM, basement membrane; FP, foot processes; E, endothelial cell (TEM ×5000) (Adapted from Liapis [10] by permission of the publisher Taylor & Francis, http://www.tandfonline.com.) (C) SEM of a healthy glomerulus reveals octopus-like podocytes wrapping the capillary loops with interdigitating foot processes capillary loops (SEM ×140 000). (Adapted from Liapis [10] by permission of the publisher Taylor & Francis, http://www.tandfonline.com.) (D) The Jablonski diagram shows a comparison between one- and two-photon absorption. Conventional one-photon excitation uses ultraviolet or visible light to excite fluorescent molecules, whereas two-photon excitation depends on the simultaneous absorption of two photons with double λ (infrared light). Compared with one-photon excitation, the fluorescence excitation (orange) of two-photon excitation is spatially restricted to a small point within the focal plane, because the fluorescence excitation occurs only where the density of illuminating photons is highest. This reduces tissue photobleaching compared with the confocal approach. Representative MPM images of glomeruli (G) in vivo in (E) control or (F) PAN-treated Munich–Wistar–Fromter rat kidneys. The freely filtered dye, Lucifer Yellow, injected intravenously, labelled the Bowman’s space and early proximal tubule (PT) and allowed negative labelling of podocytes and parietal cells (which do not take up the dye). Cell nuclei were labelled using Hoechst33342 (green). The intravascular space (plasma) was labelled red by 70-kDa dextran rhodamine B injected intravenously. Unlike a normal glomerular structure (E), podocytes develop numerous pseudocysts (asterisk) after PAN treatment (F). Scale bars: 20 μm. (Adapted from Peti-Peterdi [17] with permission from Elsevier.)

FIGURE 2

(A) The Jablonski diagram (left) shows the process of stimulated emission in STED microscopy. In normal fluorescence, a fluorophore can absorb a photon from the excitation light (blue arrow) and jump from the ground state to the excited state. Spontaneous fluorescence emission (with longer wavelength) (yellow arrow) brings the fluorophore back to the ground state. Stimulated emission of the excited molecules (red arrow) causes the emitted light to be of sufficiently longer wavelength and shorter fluorescent lifetime so that it can be separated from normal fluorescence. Schematic drawing (right) of a STED microscope: the excitation laser (blue) and STED laser (red) are focused into the sample through the objective. A phase mask is placed in the light path of the STED laser to create a specific doughnut-shaped pattern at the objective focal point. With this configuration, a diffraction-limited spot is excited (blue spot) while a superimposed, red-shifted STED laser with a doughnut shape (red spot) depletes all emission laterally, leaving only a central focal spot with a dimension less than the diffraction limit (yellow spot). Only emitted photons from the centre of the doughnut are collected. (B) A z-stack and 3D rendering of a glomerulus in a cleared tick slice (500 μm thick) of kidney tissue acquired with confocal microscopy. The optical transparency and antibody penetration depth were sufficient for imaging thick samples with confocal microscopy, enabling a global view of protein expression in the whole glomerulus. Magnifications of the boxed area show the comparison between confocal and STED acquisitions. The sample was stained for nephrin (green) and podocin (red). Localization of podocin and nephrin can clearly be resolved at the nanometre scale with super-resolution STED imaging. (b’) Depth coding profile of the same glomerulus in (B). (C) Volumetric representation of a 3D STED z-stack. The combination of optical clearing, immunostaining and higher-resolution imaging permitted visualization of the spatial distribution of proteins in the slit diaphragm (in c’, 3D projections of C). All images were deconvolved with SVI Huygens software. STED and confocal images were acquired with a Leica TCS SP8 STED 3X.

FIGURE 3

(A) A schematic drawing shows the basic principle of SIM. SIM uses a wide-field microscope setup with a fine-striped illumination pattern. The interaction between illumination patterns and structures in the sample produces moiré fringes, allowing the capture of high-frequency information (corresponding to fine details in the sample) at lower spatial frequencies. By acquiring multiple images with illumination patterns of different phases and orientations, a high-resolution image can be reconstructed. Because the illumination pattern itself is also limited by the diffraction of light, SIM is only capable of doubling the spatial resolution. (B, C) Representative images of nephrin-stained glomeruli after SIM reconstruction. The asterisk indicates the glomerular capillary lumen and arrows indicate the plan view areas on the glomerular capillary. The magnification of the boxed area in (B) shows a regular nephrin staining pattern with a nanometric resolution of the slit diaphragm. Scale bar = 10 µm. Depth coding profile of a 4.5 µm z-stack acquisition (C). (Adapted from Siegerist [36]. This article is licenced under a Creative Common Attribution 4.0 International License. http://creativecommons.org/licenses/by/4.0/.) (D) A schematic drawing shows the basic principle of localization-based microscopy techniques, such as STORM. All these techniques are based on the possibility to cyclically switch on and off individual fluorescent molecules that are too close to be resolved. In these approaches, molecules within a diffraction-limited region can be activated at different time points so that they can be individually imaged and subsequently localized by computationally finding their centres. (E–I) STORM imaging of slit diaphragm proteins to orient the localization of them relative to the GBM (Adapted from Suleiman [39] with permission). (E–G) Double-colour imaging of the GBM protein agrin (blue) and podocyte proteins nephrin (E, red), podocin (F, red) and Cd2ap (G, red) shows that these proteins cluster adjacent to the GBM. (H) Triple-colour imaging of agrin (blue), nephrin (red) and synaptopodin (green) shows that synaptopodin clusters are located between the nephrin clusters. (I) Triple-colour imaging of agrin (blue), synaptopodin (green) and the cytoskeletal protein α-actinin-4 (magenta) shows that synaptopodin and α-actinin-4 clusters have a similar pattern. Scale bars: 200 nm.