The present study describes decellularization-based methodologies for visualizing and quantifying intramuscular adipose tissue (IMAT) deposition through intact muscle volume, as well as quantifying metrics of individual adipocytes that comprise IMAT.
Fatty infiltration is the accumulation of adipocytes between myofibers in skeletal muscle and is a prominent feature of many myopathies, metabolic disorders, and dystrophies. Clinically in human populations, fatty infiltration is assessed using noninvasive methods, including computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound (US). Although some studies have used CT or MRI to quantify fatty infiltration in mouse muscle, costs and insufficient spatial resolution remain challenging. Other small animal methods utilize histology to visualize individual adipocytes; however, this methodology suffers from sampling bias in heterogeneous pathology. This protocol describes the methodology to qualitatively view and quantitatively measure fatty infiltration comprehensively throughout intact mouse muscle and at the level of individual adipocytes using decellularization. The protocol is not limited to specific muscles or specific species and can be extended to human biopsy. Additionally, gross qualitative and quantitative assessments can be made with standard laboratory equipment for little cost, making this procedure more accessible across research laboratories.
The accumulation of adipocytes between myofibers within the skeletal muscle is a prominent feature of disparate conditions, from type 2 diabetes to sarcopenia to musculoskeletal injury1,2,3,4,5,6,7. Comprehensive assessment of this intramuscular adipose tissue (IMAT) is critical to understanding the pathogenesis of these conditions, as IMAT deposition is strongly correlated with insulin resistance3,8,9,10 and poor skeletal muscle function11,12,13,14,15. Although these associations have been noted for decades, the mechanisms associated with and the origin of IMAT still remain an area of intense investigation. This is partly because most studies assessing skeletal muscle fatty infiltration have been performed in humans, where mechanistic investigations are limited16,17. However, more recently, small animal models, including mice, have been utilized to help pinpoint cellular regulation of IMAT development and signaling18,19,20. This work aims to provide a new tool for use with small animal models for qualitatively visualizing and quantifying skeletal muscle fatty infiltration.
Clinically in human populations, fatty infiltration is assessed using noninvasive methods, including computed tomography (CT)6,21, magnetic resonance imaging (MRI)16,17,22,23, and ultrasound (US)17,24. These imaging techniques typically identify a defined region of interest (ROI) in a muscle and acquire image slices within that region, although comprehensive approaches have also been employed25,26,27. These image slices are subjected to qualitative grading6 and quantified via pixel thresholding28. Similar approaches have been utilized in animals previously29,30; however, they are costly and require access to small animal imaging systems. Spatial resolution via CT and MRI use also presents a major issue, as they are unable to delineate IMAT adipocytes from skeletal muscle fibers within a voxel and instead rely on subjective separation of primarily muscle regions and primarily IMAT regions31,32. As such, the inability to accurately identify fat or muscle tissue also presents inaccurate quantification of representative amounts of these tissues.
Due to these limitations, current techniques to assess skeletal muscle fatty infiltration in small animal models most commonly rely on histology as an inexpensive and accessible alternative33,34. Standard staining procedures, including hematoxylin and eosin (H&E), oil red O (ORO), and immunostaining for adipocyte markers such as perilipin, allow for simple detection and visualization of adipocytes comprising fatty infiltration within the muscle. However, histology approaches are rarely comprehensive, and typically, IMAT qualitative or quantitative assessment is limited to a single section34. Lipid extraction has also been used to quantify total muscle lipids35; however, this technique fails to distinguish between intramyocellular lipid (IMCL) and intramuscular adipose tissue (IMAT) stores36. In summary, current methodologies to visualize and quantify fat in muscle remain limited either by financial costs or the specific detection of IMAT.
Here, we describe a detailed method for assessing skeletal muscle fatty infiltration both by qualitative visualization and multi-scale quantification. This methodology employs a simple decellularization technique that removes myocellular structures, including IMCL, but keeps the larger IMAT adipocyte-derived lipid droplets intact. Validation of the specificity of this technique has been published37, including using lipid extraction to show the depletion of IMCL with decellularization, µCT to show the retention of IMAT patterning with decellularization, and histology to show the similar size distribution of IMAT lipid droplets compared with those identified with decellularization. Once decellularized, muscles can be stained with lipid-soluble dyes for qualitative visualization of the pattern and the extent of fatty infiltration and/or quantitative imaging of individual IMAT lipid droplets. Dyes can subsequently be extracted with isopropanol, and the optical density (OD) of the resulting solution can be used to estimate the IMAT lipid volume. The stringent validation of this technique has been published elsewhere37. This article provides a detailed protocol to use this methodology with mouse muscles and provides troubleshooting tips to support the adoption of this method in other applications, such as muscles from other species or other tissues.
The care and sacrifice of mice were performed in accordance with the National Institutes of Health Guide for the Use and Care of Laboratory Animals. All work was approved by the Animal Studies Committee at Washington University in the St. Louis School of Medicine. Male C57BL/6J mice aged 2-3 months (see Table of Materials) were used to generate the example images included in this protocol. All steps described below are performed at room temperature.
1. Muscle decellularization
2. Visualization of IMAT with oil red O
3. Visualization of IMAT lipid droplets with BODIPY
NOTE: Confocal imaging is most effective with thin muscles like the EDL or diaphragm (~2 mm thickness). Alternatively, comparable thickness strips of muscles like the TA could be used.
4. Estimation of total lipid volume by lipid extraction
5. Quantification of IMAT lipid droplet metrics from confocal images
NOTE: This section require access to ImageJ version 1.47 (see Table of Materials) or later and basic ImageJ skills40.
Qualitative visualization of skeletal muscle fatty infiltration
Properly decellularized muscles are white and semi-transparent (section 1; Figure 1). When decellularized muscles are stained with ORO to visualize IMAT (section 2), IMAT lipid droplets are apparent within the clear muscle structures as red spheres (Figure 1). Healthy mouse hindlimb muscles has little natural IMAT, evidenced by little to no red, ORO positive lipid (Figure 1A). By comparison, hindlimb muscles injected with cardiotoxin (CTX; Figure 1B) or glycerol (GLY; Figure 1C) 14 days prior to decellularization have an increased accumulation of IMAT, with a larger concentration of IMAT following CTX compared with GLY as previously noted37.
Incomplete decellularization can be identified immediately following initial SDS treatment or after washout of the ORO staining as semi-opaque, light pink fibers (Figure 2B compared with Figure 2A). Incomplete clearance of ORO can be identified following ORO washout as a pink or red uniform background, rather than distinct fiber lines (Figure 2C). Figure 2A,B also contains epimuscular adipose (asterisks), a clump of lipid droplets outside the decellularized muscle. Figure 2C also demonstrates the muscle folding that can occur if muscles are not spread out during decellularization. Incomplete decellularization, incomplete ORO clearance, and residual epimuscular adipose all artifactually increase the (OD) of extracted lipid, but do not necessarily impede qualitative assessment of fatty infiltration if they are recognized as an artifact.
Quantitative imaging of skeletal muscle fatty infiltration
BODIPY fluorescently labeled lipid droplets can be imaged via confocal microscopy for a more detailed assessment of individual lipid droplet metrics and distribution (Figure 3). This process is semi-automated, as previously described, including Matlab code37. Good BODIPY staining results in bright elliptical shapes delineated from neighboring shapes when imaged in plane (Figure 3A). Thresholding and shape segmentation provides a good first pass for generating ROIs for each lipid droplet (Figure 3B), but manual edits are needed to correct errors. The most pervasive error is lipid droplets that are deep in the tissue and thus are not bright on any slices (Figure 3B; pink double arrows). This can be remedied by adding ROIs by hand using the Oval tool in ImageJ (Figure 3C). The second is identifying a group of lipid droplets as a single ROI (Figure 3B; red asterisk). This can be corrected by deleting and replacing the original ROI with multiple new ROIs (Figure 3D). Finally, a single lipid droplet is identified as a unique ROI in multiple slices, so the duplicate ROIs must be consolidated into a single ROI (Figure 3E; blue arrow). This is most easily done using a data processing tool such as Matlab, but could also be done by hand by identifying the largest ROI and deleting the rest. Average values for each metric in the C57BL6/J and 129Sv mouse can be found in Biltz et al.37.
Figure 1: Example oil red O (ORO) staining of decellularized muscles. ORO-stained muscles 14 days post-injection with saline (SAL), cardiotoxin (CTX), or glycerol (GLY). Mouse extensor digitorum longus (EDL) and tibialis anterior (TA) muscles have little IMAT (red spheres) with SAL treatment (A), but accumulate IMAT in response to CTX (B) and GLY (C) treatment. There is a complete decellularization and ORO washout, evidenced by distinct ORO positive lipid droplets in a transparent white muscle background. Scale bars = 500 µm. Please click here to view a larger version of this figure.
Figure 2: Examples of poor ORO staining results. Incomplete decellularization or incomplete ORO clearance leads to a semi-opaque pink/red background. Compared with the transparent white background of fully decellularized mouse diaphragm muscle (A), incomplete decellularization is characterized by light pink/red fiber tracks (B), and incomplete ORO clearance is characterized by a diffuse pink/red background (C). Scale bars: upper panels = 1 mm; lower panels = 500 µm. Please click here to view a larger version of this figure.
Figure 3: Examples of individual lipid droplet identification with fluorescent BODIPY staining and confocal microscopy Individual BODIPY stained lipid droplets can be identified and quantified in decellularized muscles using confocal microscopy. Individual slices through a confocal stack show lipid droplets in plane as bright green ellipses, and lipid droplets out of plane as fainter shapes (A; blue arrows). Thresholding combined with watershed object segmentation and ROI identification can map BODIPY-stained ROIs (B). Thresholding may miss some fainter lipid droplets (B; pink double arrow), requiring identification by hand (C). Watershed segmentation may group several lipid droplets together (B; red asterisk), requiring deletion of the ROI and re-estimation by hand (D). The same lipid droplet is identified in multiple slices requiring image registration (E) to delete the duplicate ROIs. Scale bars = 100 µm. Please click here to view a larger version of this figure.
This manuscript describes methods to qualitatively visualize and quantify skeletal muscle fatty infiltration in small animal models that can be applied to further understanding the pathogenesis of intramuscular adipose tissue (IMAT) development and pathological expansion. The use of whole-muscle decellularization and lipid-soluble staining allows for a cost-effective, reproducible, and simple methodology to comprehensively assess the presence of IMAT in whole muscles.
The basis for this protocol is that decellularization of muscle with SDS removes the cellular components of myofibers, including the small lipid droplets of IMCL, but spares the large lipid droplets in intramyocellular adipocytes. SDS has been used extensively42 in tissue engineering to decellularize matrices. Tissues such as adipose and skeletal muscle typically require additional mechanical dissociation and/or alcohol extraction to remove the residual adipocyte lipid42,43. We have previously shown that this is because while decellularization with SDS eliminates IMCL, it spares the large lipid droplet in adipocytes37. Imaging of osmium tetroxide-stained intact muscle pre- and post-decellulariztion with µCT verified that the spatial pattern of IMAT was not disrupted by decellularization. Further, intramuscular triglyceride quantification in a decellularized muscle with negligible IMAT was ~5% of the intact muscle values, verifying the removal of IMCL. Therefore, this methodology retains IMAT lipid droplets in their original anatomical distribution through a semi-transparent muscle matrix.
Proper decellularization is the most critical step in this protocol. If the decellularization is incomplete, IMAT lipid droplets will be difficult to visualize and residual IMCL will cause high background staining with either ORO or BODIPY (Figure 2). Common errors by inexperienced users are inadequate SDS coverage per muscle (within each well), such that each muscle is not completely covered in SDS solution, not using a rocker to agitate the solution during decellularization, and not performing solution changes frequently enough. In this manuscript, we have recommended the amount of SDS needed per unit muscle mass, but the user will still need to ensure that muscles are completely covered by solutions, as each muscle has a unique geometry. Users are also recommended to change the solutions liberally (as much as twice per day) to ensure that decellularization is complete. Good quality staining of IMAT lipid droplets has been achieved after as many as 4 days of SDS treatment. For high-quality ORO staining results, adequate fixation and ORO solution prep are also important. Similar to the SDS treatment described above, adequate coverage of 3.7% formaldehyde solution for each muscle sample is needed. If the muscle is removed from the fixative too early, lipid droplets will only weakly stain with ORO. A total of 1-2 h should be sufficient, but overnight fixation is recommended to ensure the fixative penetrates to the center of the muscle and fully fixes all lipid droplets. An additional challenge with ORO staining is that when the alcohol concentration is reduced to 60%, a particulate begins to form. This particulate can settle on the surface and become stuck on the border of muscle. The best way to avoid this is to make a fresh working solution for each staining and use both 40 mesh µm and 0.22 µm filters. Then, maintaining agitation with the rocker and limiting the staining time to 10 min will help keep any particulate that forms from settling. If the problem persists, making a fresh ORO stock solution may help. If some artifact remains stuck to the decellularized muscle surface, a stereo microscope, forceps, and surgical scissors can be used to remove this artifact. Failing to eliminate artifacts will impact the image quality of muscles and overestimate IMAT content during the lipid extraction portion in preparation for OD reading.
Overall, this technique is straightforward and offers several advantages over gold standard methods for visualizing and quantifying skeletal muscle fatty infiltration. Noninvasive techniques, such as CT, MRI and US, which are used extensively in humans and sometimes in animal models, have limited spatial resolution and are unable to distinguish lipid droplets from muscle fibers. Thus, a pixel or voxel of intermediate signal intensity is assigned as “muscle” or “fat”, while in actuality it is likely a mix of myofibers and adipocytes. More commonly, fatty infiltration in animal muscle is assessed by histology, most frequently by ORO in muscle cryosections. However, this is typically only performed in a single representative section and is difficult to quantify due to lipid scatter over the section. By contrast, ORO staining of an entire decellularized muscle provides a comprehensive assessment of IMAT with similar costs and effort as intact morphology. Furthermore, in addition to enhancing visualization, ORO staining of decellularization enables the quantification of fatty infiltration by lipid extraction. For a deeper dive into the features of fatty infiltration, a fluorescent stain, BODIPY, can be used in conjunction with confocal microscopy. This enables the reconstruction of individual IMAT lipid droplets to map the 3D landscape, which is not possible with histology unless sections are analyzed over the length of the muscle. While a confocal microscope is not standard lab equipment, it is more likely to be accessible in a university or industry setting than small animal MRI or CT. Furthermore, much of this process can be automated, reducing the time cost compared with sequential histology. Optimizing the settings on the confocal microscope is an additional consideration for BODIPY staining. These are unique to each microscope. The critical value is laser intensity, which must be high enough to detect the lipid droplets on the distant surface of the muscle while also not saturating the signal from the lipid droplets on the near side. Because of this, it is suggested that using BODIPY staining with confocal microscopy is best suited on thinner muscles, including the EDL or diaphragm.
Several limitations of this approach warrant discussion. First, while it is anticipated that this technique has broad applicability beyond injury models (cardiotoxin and glycerol) in mice presented here, new applications (e.g., the mdx model) may require optimization, as the size and composition of the muscle (e.g., fibrosis) could affect decellularization, requiring increased SDS concentration or incubation times. Other disease models with altered muscle mass would also require analysis of both absolute and normalized (to muscle mass) metrics of fatty infiltration to determine the absolute amount of lipid or percentage of lipid relative to the muscle volume to provide a more meaningful outcome measure. Furthermore, this technique is anticipated to be broadly applicable to larger animal models and human biopsies, but this may require optimization for each new application. Second, in this strategy, the entire muscle must be dedicated to this assay and cannot be used to assess another pathological feature. Studies that aim to assess longitudinal changes in IMAT are better served with noninvasive imaging techniques and studies whose primary aim requires the muscle for other purposes (histology, quantitative polymerase chain reaction, western blotting) are better served by histological assessment, as the remainder of the frozen muscle can be allocated to other assays. However, this assay is well suited to pair with in vivo testing, such as treadmill running, or ex vivo contractile testing ,since these measures can be made before decellularization44. Third, although the use of BODIPY stain with confocal microscopy provides high-resolution visualization and quantification of lipid droplets, it cannot conclusively identify lipid droplets as individual adipocytes, as the cell membrane is removed and endogenous adipocyte proteins are lost. Multilocular adipocytes, representing immature adipocytes or a “brown/beige” phenotype, may be identified as multiple lipid droplets. Finally, the protocol does not work well on previously frozen muscle. These limitations are probably most profound for human biopsies, as while the entire biopsy can be decellularized, the spatial distribution of IMAT in the biopsy is not likely to be more representative of the whole muscle than a histological slice. However, since this technique is relatively insensitive to unfrozen biopsy handling conditions (e.g., hours on ice in PBS), the biopsy could be divided later for various assays, including a portion for decellularization, which would provide a better resolution of individual lipid droplets.
In conclusion, a novel method for qualitative and quantitative analysis of skeletal muscle fatty infiltration has been developed by staining and imaging the retained lipid of decellularized constructs. This methodology offers improvements over gold-standard approaches, in that it enables comprehensive imaging of three-dimensional fatty infiltration within muscle and quick, cheap quantification with ORO staining. For more detailed measures, a second lipid-soluble BODIPY stain provides a more detailed quantification of lipid droplet number, volume, and distribution pattern, as imaged by confocal microscopy. Together, these measures provide researchers with a way to precisely measure skeletal muscle fatty infiltration at the level of the individual lipid droplets without sampling or expensive noninvasive imaging.
The authors have nothing to disclose.
This work was supported by R01AR075773 to GAM.
0.22 µm Syringe Filter | Fisher Scientific | SLGP033RS | |
1 mL LuerLock Syringes | Fisher Scientific | 14823434 | |
12 mm Coverslips | Fisher Scientific | 12545F | |
12 well plates | Fisher Scientific | 08-772-29 | |
24 well plates | Fisher Scientific | 08-772-1H | |
2-Propanol (Isopropanol) | Sigma Aldrich | I9516 | 0.5 mg/mL stock solution can be stored at room temperature for 1 month. Working solution must be made fresh. |
37% Formaldehyde Solution | Sigma Aldrich | 8187081000 | |
40 µm Mesh Filter | Fisher Scientific | 87711 | |
6 well plates | Fisher Scientific | 08-772-1B | |
96 well plates | Fisher Scientific | 08-772-2C | |
BODIPY 493/503 | Fisher Scientific | D-3922 | |
C57BL/6J Mice | Jackson Laboratory | 000664 | |
Confocal Imaging Dish | VWR | 734-2905 | |
Confocal Microscope | Leica | TCS SPEII | |
Dissecting/stereo Microscope | Zeiss | 4107009123001000 | |
Dissection scissors | Fine Science Tools | 14060-09 | |
Dumont #5 forceps | Fine Science Tools | 11254-20 | |
Ethanol | Fisher Scientific | 033361.K2 | |
ImageJ | NIH | ||
Matlab | Mathworks | ||
Oil Red O Powder | Sigma Aldrich | O0625 | |
Plate reader | Bio-tek | Synergy II | |
Rocker/Shaker | Reliable Scientific | 55D | |
Sodium Dodecyl Sulfate (SDS) | Sigma Aldrich | L3771 | 1% Solution can be stored at room temperature for 1 month |
Transfer pipettes | Fisher Scientific | 137119D | |
Vannas spring scissors | Fine Science Tools | 15000-00 |