The focus of the present study is to demonstrate the whole-mount immunostaining and visualization technique as an ideal method for 3D imaging of adipose tissue architecture and cellular component.
Adipose tissue is an important metabolic organ with high plasticity and is responsive to environmental stimuli and nutrient status. As such, various techniques have been developed to study the morphology and biology of adipose tissue. However, conventional visualization methods are limited to studying the tissue in 2D sections, failing to capture the 3D architecture of the whole organ. Here we present whole-mount staining, an immunohistochemistry method that preserves intact adipose tissue morphology with minimal processing steps. Hence, the structures of adipocytes and other cellular components are maintained without distortion, achieving the most representative 3D visualization of the tissue. In addition, whole-mount staining can be combined with lineage tracing methods to determine cell fate decisions. However, this technique has some limitations to providing accurate information regarding deeper parts of adipose tissue. To overcome this limitation, whole-mount staining can be further combined with tissue clearing techniques to remove the opaqueness of tissue and allow for complete visualization of entire adipose tissue anatomy using light-sheet fluorescent microscopy. Therefore, a higher resolution and more accurate representation of adipose tissue structures can be captured with the combination of these techniques.
Adipose tissue is an essential organ for energy storage and is characterized by dynamic remodelling and nearly unlimited expansion1. In addition to energy homeostasis, adipose tissue also plays an essential role in hormone secretion of over 50 adipokines to modulate whole-body metabolic function2. Adipose tissue has a diverse architecture comprising of various cell types including mature adipocytes, fibroblasts, endothelial cells, immune cells, and adipocyte progenitor cells3. Recent studies have shown that obesity and other metabolic dysfunction can significantly alter adipose tissue function and its microenvironment, which includes but is not limited to enlargement of adipocytes, infiltration of inflammatory cells (e.g., macrophages), and vascular dysfunction3.
Conventional morphological techniques such as histology and cryosectioning demonstrate several limitations in studying adipose biology such as lengthy chemical processing steps, which can lead to tissue shrinkage and structure distortion3,4. Furthermore, these 2D techniques are insufficient to observe intercellular interactions exerted by different cell types, as the sections obtained are limited to smaller regions of the entire tissue3. Compared to conventional methods of fluorescent imaging, whole-mount staining does not require additional invasive steps, such as embedding, sectioning, and dehydration; thus, this avoids the problem of diminishing antibody specificity. As such, it is a simple and efficient method for imaging adipose tissue, with better preservation of adipocyte morphology and overall adipose tissue structure5. Therefore, whole-mount staining as a quick and inexpensive immunolabeling technique was established to preserve adipose tissue 3D architecture1,6,7,8.
However, despite the preservation of adipose tissue morphology with use of whole-mount staining, this technique is still unable to visualize inner structures beneath the lipid surface of the tissue. Several recent studies9,10 have established tissue clearing techniques combined with whole-mount immunolabeling1,6 to allow for comprehensive 3D visualization within adipose tissue. In particular, dense neural and vasculature networks were visualized in recent studies9,10,11,12 with 3D volume imaging. Indeed, studying the neural and vascular plasticity of adipose tissue under different physiological conditions is essential to study its biology. Immunolabeling-enabled three-dimensional imaging of solvent-cleared organs (iDISCO+) tissue clearing is a process comprised of methanol pre-treatment, immunolabeling, and clearing of tissue opaqueness with organic chemical reagents dichloromethane (DCM) and dibenzyl ether (DBE)13,14. By making the adipose tissue entirely transparent, a more accurate representation of anatomy within the tissue such as blood vessels and neural fibers can be obtained9,10. IDISCO+ has advantages in that it is compatible with various antibodies and fluorescent reporters11,14, and it has demonstrated success in multiple organs and even embryoes14. However, its main limitation is a long incubation time, in which 18 to 20 days are needed to complete the entire experiment.
Another important application of whole-mount staining is the visualization of cell fate in combination with a lineage tracing system. Lineage tracing is the labelling of a specific gene/marker in a cell that can be passed on to all daughter cells and is conserved over time15. As such, it is a powerful tool that can be used to determine the fate of a cell's progeny15. Since the 1990s, the Cre-LoxP recombinant system has become a powerful approach for lineage tracing in living organisms15. When a mouse line that expresses Cre, a DNA recombinase enzyme, is crossed with another mouse line expressing a reporter that is adjacent to a loxP-STOP-loxP sequence, the reporter protein is expressed15.
For whole-mount staining, the use of fluorescent multicolor reporters is suitable for imaging of adipose tissue because it allows for minimal interference with intracellular activities of the adipocyte16. However, traditional reporters typically stain the cytoplasm, making it difficult to trace the lineage of white adipocytes, which have limited cytoplasmic content17. To overcome this problem, the use of membrane-bound fluorescent tdTomato/membrane eGFP (mT/mG) reporter marker is an ideal tool. Membrane-targeted tdTomato is expressed in Cre-negative cells18. Upon Cre excision, a switch to the expression of membrane-targeted eGFP occurs, making this reporter suitable for tracing the lineage of adipocyte progenitors17,18 (Supplementary Figure 1).
The purpose of this paper is to provide a detailed protocol for whole-mount staining and show how it can be combined with other techniques to study the development and physiology of adipose tissue. Two examples of applications described in this protocol are its use with 1) multicolor reporter mouse lines to identify various origins of adipocytes and 2) tissue clearing to further visualize the neural arborization in white adipose tissue (WAT).
All experimental animal protocols were approved by the Animal Care Committee of The Center for Phenogenomics (TCP) conformed to the standards of the Canadian Council on Animal Care. Mice were maintained on 12-h light/dark cycles and provided with free access to water and food. 7 month old C57BL/6J male mice were used in the whole-mount staining experiment.
NOTE: Sections 1 to 2 are in chronological order, with section 3 being an optional step right after section 1. Section 4 can be performed to analyze adipocyte size and blood vessel density after the completion of section 2.
1. Materials Preparation and Tissue Isolation
2. Whole-mount Staining of White Adipose Tissue
3. Tissue Clearing and Immunolabeling Using iDISCO+
NOTE: This protocol is based on previously published procedures9,10,19.
4. Examples of Data Analysis from Whole-mount Stained Tissue Images Using ImageJ
NOTE: See https://imageJ.nih.gov/ij/download.html for download and installation instructions.
Due to the fragility of adipose tissue, methods involving multiple processing steps and sectioning can lead to disfigurement of adipose tissue morphology3 (Figure 1A). However, whole-mount staining can preserve the morphology of adipocytes, ensuring accurate interpretation of results (Figure 1B).
Over-fixation of adipose tissue leads to fixative-induced autofluorescence. As shown in Figure 2A, green and red channels staining for tyrosine hydroxylase (TH) and PECAM-1 signals, respectively, overlap in identical regions of the tissue, indicating that autofluorescence may have occurred due to over-fixation in PFA for 3 days. In contrast, Figure 2B shows a representative image of whole-mount staining when proper fixation is performed, as the signal for TH staining occurs in different areas relative to PECAM-1 signal, demonstrating that this signal is not autofluorescence and is in fact a positive signal.
Whole-mount staining is an important visualization tool for Cre-loxP-based lineage tracing of adipocytes15, with mT/mG being the ideal reporter system18. Ng2, a marker for adipocyte progenitor cell population, is a plasma membrane proteoglycan. In this system, Ng2-Cre-positive cells express m-GFP, whereas m-Tomato is expressed in Ng2-Cre-negative cells (Figure 3).
Adipose tissue is an incredibly dynamic organ, capable of expanding and shrinking under different physiological conditions and demands20. Imaging whole-mount stained adipose tissue allows for quantification of imorphological changes under different experimental conditions. In particular, adipose tissue is highly vascularized, which is important in mediating metabolic homeostasis upon rapid changes in energy level1. For instance, C57BL/6J mice that undergo 24 h of fasting display significantly smaller adipocyte size (Figure 4), indicating lipolysis, and a trend in elevated blood vessel density compared to continuously fed mice (Figure 5). ImageJ software was utilized to quantify the size of the adipocytes and blood vessel density, as described above.
Tissue clearing is a relatively new technique developed to remove the opaqueness of adipose tissue to allow visualization deep within the tissue volume9,10 (Figure 6). Whole-mount staining on uncleared IWAT using confocal microscopy only showed sparse sympathetic innervation, since nerve fibers underneath the surface of the tissue could not be visualized (Figure 7A). However, dense neural arborization could be observed after tissue clearing and immunolabeling with the use of iDISCO+ as well as the use of light-sheet fluorescent microscopy (LSFM) (Figure 7B).
Figure 1: Comparison of adipose tissue morphology with conventional morphological techniques and whole-mount staining technique. (A) H&E stained adipose tissue on a paraffin-embedded section (left), with black arrowheads indicating distorted regions of adipocytes. Lectin (carbohydrate binding protein, white arrows) fluorescent dye injection, immunofluorescent staining of F4/80 (macrophage marker, yellow arrowheads), and DAPI nuclei staining of adipose tissue on cryosection (right). (B) White adipose tissue visualization using whole-mount staining with a step-size of 5 μm. The total Z-stack depth captured is around 100 μm. Adipocyte lipid droplets were stained with neutral lipid stain (grey), and blood vessels were stained with PECAM-1. Image was captured with microscopy with a step-size of 5 μm for Z-stacking (red). Please click here to view a larger version of this figure.
Figure 2: Visualization of neural fibers and blood vessels using whole-mount stained adipose tissue. (A) Representative microscopic images of undesirable results from whole-mount stained PWAT due to over-fixation in PFA for 3 days. Overlapping signals are indicated by white arrowheads. (B) Representative microscopic images of a positive result from whole-mount stained IWAT from control mouse. The images were captured at 100X magnification. Please click here to view a larger version of this figure.
Figure 3: Lineage tracing using mT/mG system in the adipose tissue. Representative images Cre-positive (mG) and Cre-negative (mT) cells in IWAT of a Ng2-Cre; mT/mG mouse. The images were captured at 200X magnification. Please click here to view a larger version of this figure.
Figure 4: Visualization and quantification of adipocyte size. (A) Representative images of adipocytes in PWAT of fed and 24-hour fasted C57BL/6J mice using neutral lipid staining. The images were captured at 200x magnification. (B) Adipocyte size comparison between fed and 24-h fasted C57BL/6J mice using ImageJ software. Values are expressed as mean ± SEM; 2-tailed unpaired Student’s t-test; ***p < 0.001. Please click here to view a larger version of this figure.
Figure 5: Visualization and quantification of blood vessel density. (A) Representative images of blood vessels in fed and 24-hour fasted C57BL/6J mice using PECAM-1 antibody. (B) Comparison of blood vessel density between fed and 24-hour fasted C57BL/6J mice using ImageJ software. Values are expressed as mean ± SEM; 2-tailed unpaired Student’s t-test. Please click here to view a larger version of this figure.
Figure 6: Clearing of adipose tissue using the iDISCO+ method. Prior to clearing, the tissue is opaque. The tissue becomes completely transparent at the end of the tissue clearing steps.
Figure 7: Whole-mount stained IWAT compared with tissue-cleared IWAT using iDISCO+ method. (A) Visualization of neural fibers using TH antibody (1:500) in whole-mount stained IWAT at 100x magnification with confocal microscopy with a step-size of 5 μm. The total Z-stack depth captured is around 100 μm. (B) Visualization of neural fibers using TH antibody (1:200) in whole-mount stained IWAT with iDISCO+ protocol at 1.6X magnification using LSFM with a step-size of 4 μm. The total Z-stack depth captured is around 8 mm. Please click here to view a larger version of this figure.
Supplementary Figure 1: Schematic diagram for mT/mG lineage tracing system. A dual fluorescent system that employs membrane-targeted eGFP and membrane-targeted tdTomato. Before Cre recombination, the mT is globally expressed. When the cell expresses Cre, the mT cassette is excised, and mG is expressed permanently. pA represents polyadenylation sequences after the stop codon. Please click here to view a larger version of this figure.
Supplementary Figure 2: Application of ImageJ software to quantify percentage area of blood vessel density. (A) “Image” tab commands and options to convert an image into a black-and-white threshold color. (B) Summary measurement of percentage area of vessel density using the “Analyze Particles” command. Please click here to view a larger version of this figure.
Supplementary Figure 3: Application of ImageJ software to quantify adipocyte area. “Analyze” tab: “Set Scale” and “Measurement” commands to measure the area of the adipocyte(s). Results display shows the area of each adipocyte measured. Please click here to view a larger version of this figure.
Although conventional techniques such as histology and cryosection offer benefits for observing intracellular structure, whole-mount staining provides a different perspective in adipose tissue research, which enables 3D visualization of cellular architecture of minimally processed tissue.
In order to successfully perform whole-mount staining, the following suggestions should be taken into consideration. Different adipose tissue depots can yield various immunostaining results; thus, the type of adipose tissue depot used should be determined first. For instance, brown adipose tissue (BAT) is denser relative to white adipose tissue (WAT) due to smaller, multilocular adipocytes21. This increased density of BAT makes it difficult for antibodies to permeate through the tissue. In addition, the size of the adipose tissue is also imperative for proper antibody staining, since too large/thick tissues can also result in insufficient antibody penetration. Hence, when obtaining the adipose tissue during animal dissection, the distal portion of the perigonadal fat tissue should be used, since this is the thinnest region and can allow for sufficient antibody penetration and consistent data. Alternatively, for denser tissues, use of a fluorescent reporter mouse line, such as mT/mG, may allow for better visualization of the marker of interest, since this avoids the issue of insufficient antibody penetration. Subsequently, 1% PFA is used for tissue fixation to preserve the natural distribution of proteins and ensure tissue permeabilization and antibody penetration22,23. However, over-fixation can decrease antigen recognition and produce autofluorescence24. This is due to the reaction between aldehyde groups on PFA and other aldehyde-containing fixative and tissue components, which creates fluorescent compounds24. To prevent the need for an antigen-retrieval step, it is recommended to leave tissues in a fixative for only an hour at room temperature and begin the next steps as soon as fixation is completed7. Like many other immunolabeling techniques, titrating the antibody concentration is an essential troubleshooting step to ensure the desired signal.
Indeed, there are several limitations of whole-mount staining despite the aforementioned significance and advantages. Whole-mount staining has stringent requirements regarding tissue type and size, because insufficient antibody penetration and uneven staining can occur in thicker tissues such as muscle and liver. Also, whole-mount staining is not the most accurate representation of certain antibodies such as tyrosine hydroxylase (TH), a marker for the sympathetic nervous system, as its signal is often masked by dense lipid content in adipose tissue (Figure 7). In addition, the uneven surface of adipose tissue also posed a challenge for confocal imaging, since signals located at different layers of the adipose tissue cannot be easily captured on a single image. Therefore, the quantification of signal intensity in images obtained from whole-mount staining yield inconsistent results for whole-nerve fiber arborization. The distal portion of PWAT is usually the least lipid-dense; hence, better images can be obtained from this area. However, whether this area is representative of the entire tissue for immunostaining in all types of antibodies will require further investigation.
By making tissue optically transparent, a clearing method reduces light scattering, enabling visualization of the deep structure25. iDISCO+ is an inexpensive protocol recently developed that combines whole-mount immunolabeling with volume imaging of various large cleared tissues9. Modified iDISCO+ methods with LSFM demonstrated by recent studies9,10 observed dense dendritic arborization on WAT that cannot be seen with conventional immunolabeling and confocal microscopy. Conventional confocal imaging uses an imaging beam system and pinhole for laser scanning. The scanning speed and penetrating depth are limitations of the microscope; thus, 3D tissue dynamics are often missed. In contrast, light-sheet microscopy has apparent advantages in that it uses sheet-scanning and only illuminates one optical section at a time, capturing all the fluorescence molecules within that section. Moreover, fluorophores in other sections are not excited, preventing photo-bleaching and phototoxic effects26. Thus, the amount of nerve innervation within the adipose tissue can be more accurately assessed with iDISCO+ and LSFM. Despite this, there are some limitations to the use of iDISCO+ and LSFM. For instance, users should note that only channels in the red and far-red channel are compatible with iDISCO+, because longer wavelengths of light are better able penetrate the sample27. Additionally, autofluorescence in the blue-green spectrum is quite high in large tissue samples, so imaging in the red and far-red spectra will help reduce any autofluorescence that occurs14. In regard to transgenic fluorescent reporter mice, iDISCO+ can be used to visualize reporter proteins. However, immunolabeling of the fluorescent reporter with a secondary antibody should be conducted, as the endogenous signal may fade during the tissue clearing process14. Nonetheless, this technique is extremely valuable for studying sympathetic nervous system-adipose tissue interactions and for investigating adipose plasticity under different physiological and metabolic conditions.
The authors have nothing to disclose.
This work was funded by grants from the Natural Sciences and Engineering Research Council (NSERC) of Canada, Pilot and Feasibility Study Grant of Banting & Best Diabetes Centre (BBDC), the SickKids Start-up Fund to H-K.S., Medical Research Center Program (2015R1A5A2009124) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning to J-R.K.
LipidTox | Life Technologies | H34477 | |
PECAM-1 primary antibody | Millipore | MAB1398Z(CH) | |
TH (tyrosine hydroxylase) primary antibody | Millipore | AB152, AB1542 | |
DAPI stain | BD Pharmingen | 564907 | |
Nikon A1R confocal microscope | Nikon | Confocal microscope | |
Ultramicroscope I | LaVision BioTec | Light sheet image fluorescent microscope | |
Alexa Fluor secondary antibodies | Jackson ImmunoResearch | Wavelengths 488, 594 and 647 used | |
Purified Rat Anti-Mouse CD16/CD32 | BioSciences | 553141 | |
Dichloromethane | Sigma-Aldrich | 270997 | |
Dibenzyl-ether | Sigma-Aldrich | 33630 | |
Methanol | Fisher Chemical | A452-1 | |
30% Hydrogen Peroxide | BIO BASIC CANADA INC | HC4060 | |
Dimethyl sulfoxide (DMSO) | Sigma-Aldrich | D2650 | |
Glycine | Sigma-Aldrich | J7126 | |
Heparin | Sigma-Aldrich | H3393 | |
Lectin kit I, fluorescein labeled | VECTOR LABORATORIES | FLK-2100 | |
F4/80 | Bio-Rad | MCA497GA | |
VECTASHIELD Hard Set Mounting Medium with DAPI | VECTOR LABORATORIES | H-1500 | |
Paraformaldehyde (PFA) | |||
Phosphate Buffer Saline (PBS) | |||
Triton-X | |||
Tween | |||
Animal serum (goat, donkey) |