Chapter Two - Imaging White Adipose Tissue with Confocal Microscopy

https://doi.org/10.1016/B978-0-12-411619-1.00002-1Get rights and content

Abstract

Adipose tissue is composed of a variety of cell types that include mature adipocytes, endothelial cells, fibroblasts, adipocyte progenitors, and a range of inflammatory leukocytes. These cells work in concert to promote nutrient storage in adipose tissue depots and vary widely based on location. In addition, overnutrition and obesity impart significant changes in the architecture of adipose tissue that are strongly associated with metabolic dysfunction. Recent studies have called attention to the importance of adipose tissue microenvironments in regulating adipocyte function and therefore require techniques that preserve cellular interactions and permit detailed analysis of three-dimensional structures in fat. This chapter summarizes our experience with the use of laser scanning confocal microscopy for imaging adipose tissue in rodents.

Introduction

Adipose tissue is comprised of a vast range of cellular and noncellular components. By volume, adipocytes are the most prominent cell within a given white fat depot. However, by cell number, it is likely that mature adipocytes are in the minority due to the presence of a large network of supporting cells (Granneman, 2008). These include cells that comprise an extensive vascular system in fat, including fibroblasts, preadipocytes, and cells with mesenchymal and hematopoietic stem cell capacity (Crandall et al., 1997, Nishimura et al., 2007, Zuk et al., 2002). Obesity research has called significant attention to the presence of a wide range of inflammatory leukocytes and lymphocytes in fat that change with obesity. These include myeloid cells (macrophages, neutrophils, etc.), lymphocytes (T cells, B cells), eosinophils, mast cells, and NK cells (Lee et al., 2009, Nagasaki et al., 2009, Nishimura et al., 2009, Ohmura et al., 2010). Quantitation and localization of these cells is a critical part of understanding the link between obesity and inflammation. A significant need to understand this association has spurred the development of techniques that permit a detailed examination of the diverse components and architecture of adipose tissue.

Traditional histologic techniques for adipose tissue analysis have included electron microscopy (transmission and scanning) and freeze fracturing. Many laboratories analyze adipose tissue with light microscopy (LM) techniques such as immunohistochemistry and in situ hybridization (Cinti, Zingaretti, Cancello, Ceresi, & Ferrara, 2001). However, because of the high lipid content in fat, sectioning of frozen or paraffin-embedded samples is often inconsistent and can distort adipose tissue architecture. This can lead to biased assessments of adipocyte size. More importantly, this limits our capacity to appreciate the diversity of nonadipocyte cell types in fat, and limits our ability to observe their cell–cell interactions. For those reasons, we and others have developed techniques that permit the imaging of whole-mount tissue samples in a way that maintains native architecture (Cho et al., 2007, Lumeng et al., 2008). Here, we present a detailed description of the adipose tissue structures that can be imaged with confocal microscopy in rodents, along with detailed protocols.

The mature white adipocyte is primarily composed of a single large lipid droplet that is ~ 100 μm in diameter in mice (Suzuki, Shinohara, Ohsaki, & Fujimoto, 2011). Nuclear and other subcellular components are localized within a very thin cytoplasmic layer that lines the lipid droplet and forms the ghost-like remnant of the adipocyte seen in traditional paraffin-embedded sections. Immature adipocytes contain multiple small lipid droplets and are described as having a “multilocular” appearance. As the adipocyte matures, these lipid droplets fuse and form the round “unilocular” droplet. The fluorescent stains BODIPY and Nile Red are lipid-soluble compounds that help visualize lipid aggregation (Table 2.1).

The adipocyte plasma membrane contains numerous receptors (e.g., insulin receptors) involved in cell signaling that can regulate lipid uptake and fatty acid trafficking. Of these, Caveolin-1 is enriched in the plasma membrane and is commonly found in lipid rafts (Jasmin, Frank, & Lisanti, 2012). Because Caveolin-1 is abundant on the cell surface, it provides an excellent target for staining and imaging the plasma membrane of adipocytes. Lipid droplets are surrounded by PAT proteins (i.e., perilipin, ADRP, TIP47), which regulate both storage and release of lipids. Perilipin is a useful marker of lipid droplet structures in white fat. Stimulation by adrenergic agonists changes the conformation of perilipin, which allows access of lipases, like hormone-sensitive lipase, to the lipid droplet. This results in the mobilization of triglycerides (Greenberg et al., 1991). Perilipin is also useful for identifying dead or dying adipocytes where loss of perilipin staining is noted (Feng et al., 2011). For reagents useful in visualizing these structures, refer to Table 2.1.

The death of adipocytes results in marked remodeling of the adipose tissue microenvironment. H&E sections and immunohistochemistry studies have revealed that areas with adipocyte death create regions called crown-like structures (CLSs) that are described as accumulations of proinflammatory macrophages and extracellular matrix material (Cinti, 2005, Spencer et al., 2010) (Fig. 2.1). Dying adipocytes leave behind Perilipin-negative lipid droplets that also lack Caveolin-1 staining (Feng et al., 2011, Lumeng et al., 2008, Lumeng et al., 2007). CLSs are a hallmark of adipose tissue inflammation and fibrosis in human and rodent adipose tissue.

A major cellular component of adipose tissue and CLSs is a population of adipose tissue macrophages (ATMs). Total ATMs can be detected in adipose tissue using a variety of macrophage-specific surface stains such as Mac-2 and F4/80 (Table 2.2). An example of how CLSs can be visualized is by using a combination of macrophage stain and perilipin stain, where Mac-2 and/or F4/80 will reveal a circular organization of macrophages that is void of perilipin stain. Resident CD11c/MGL-1+ M2 ATMs are seen in interstitial spaces between adipocytes and have morphologic characteristics that are distinct from CD11c+ ATMs (Lumeng et al., 2007, Xu et al., 2003). In contrast, CD11c+ “classically activated” M1 ATMs are rare in lean mice, but are abundant in obese mice. These are enriched in CLS and are frequently found to contain triglyceride-laden lipid droplets. Resident ATMs lack lipid accumulation and are enriched for markers of M2 polarization such as CD206 and CD301/MGL-1. In addition to ATMs, CLS have been shown to be sites of accumulation of numerous other lymphocytes and leukocytes that include T cells (adipose tissue T cells, or ATTs), B cells, mast cells, and eosinophils (Nishimura et al., 2009, Ohmura et al., 2010, Tsui et al., 2011). The trafficking of these cells to fat, and the mechanism by which they are enriched in CLSs, is still unclear.

Hypercellular clusters are known to reside on the surface of numerous adipose tissue depots. Milky spots (MSs) are found primarily in omental adipose tissue depots and are composed of macrophages and B and T lymphocytes (Fig. 2.2A–C). They have been shown to participate in the clearing of debris from the peritoneum and may play a role in adaptive immunity (Rangel-Moreno et al., 2009). Fat-associated lymphoid clusters (FALCs) have recently been described in mesenteric fat, as well as in gonadal fat depots in mice (Moro et al., 2010, Morris et al., 2013) (Fig. 2.2D–G). These contain a unique population of LinKit+Sca1+ innate lymphocytes. If FALC are identical or related to milky spots is not clear, but they both appear to participate in phagocytosis and immune surveillance in several contexts. While they appear to resemble lymph nodes, there is little evidence of associated lymphatic vessels.

The sizes of these uncapsulated structures range between 100 and 500 μm in diameter and are in direct contact with adipocytes (Rangel-Moreno et al., 2009). They appear to expand in concert with obesity, adipose tissue inflammation, and also in response to aging (Lumeng et al., 2011). The localization of such structures is a challenge in tissue sections as they are relatively rare on the surface of fat. However, whole-mount techniques facilitate the localization and characterization of FALCs and MSs. To image these structures, combination stains for nuclei (DAPI), T cells (CD4), and macrophages (F4/80) can be implemented (Table 2.2).

Adipose tissue contains an extensive vascular network that participates in the transport of nutrients and leukocytes in and out of fat. Many of the vascular structures are tightly associated with CLS and FALCSs and are believed to facilitate cellular trafficking of leukocytes and lymphocytes (Nishimura et al., 2008). The formation of the primitive fat organ is dependent on the development of an extensive vascular bed (Crandall et al., 1997). Groups have found that the expansion of vascular networks occurs in concert, and even precedes adipogenesis (Han et al., 2011, Hausman and Richardson, 1983, Kimura et al., 2002). Adipose tissue growth, referring to both the expansion of number (hyperplasia) and size (hypertrophy), is tightly linked with angiogenesis. It has been demonstrated that limiting angiogenesis can also block adipogenesis (Brakenhielm et al., 2004, Liu and Meydani, 2003). In parallel with this idea, proangiogenic therapies can promote adipogenesis (Tabata et al., 2000). A general stain for adipose tissue vasculature is isolectin, which binds tightly to the surface of vascular endothelial cells (Cho et al., 2007, Nishimura et al., 2008) (Fig. 2.3).

The features in adipose tissue mentioned are complex, three-dimensional structures, thus the limitations of standard LM techniques are apparent. Furthermore, limiting sampling of fat to cross-sections will underrepresent structures such as FALCs and will not fully capture the three-dimensional tortuous route of many adipose tissue blood vessels. In addition, due to the high lipid content of fat, LM often results in significant auto-fluorescence, which is further promoted by light diffraction. The amount of auto-fluorescence can be determined simply by viewing a specimen that is unstained. Often the resolution of images taken with traditional microscopes is compromised because of the fluorescence signals that may arise from other optical layers that are not within the plane of focus. Groups have attempted to alleviate this issue by cryosectioning methods and by flattening tissue fragments (Paddock & Eliceiri, 2014). However, such methods can significantly disturb the unique architecture of adipose tissue and may lead to highly variable results.

Laser scanning confocal imaging is an imaging technique that can address many of the limitations of traditional fluorescence microscopy. Because confocal imaging allows for visualizing a very narrow plane of focus, much of the interference that results from auto-fluorescence and out-of-focus blur can be removed. With proper fixation techniques and appropriate staining procedures whole-mount samples of adipose tissue can be imaged with confocal microscopy much more rapidly than with conventional LM techniques. Other advantages of confocal microscopy and its sophisticated optics include the use of specific excitation wavelengths, as well as the ability to employ detectors that exclude auto-fluorescence from other emission spectra. This feature comes into play when there is cross-fluorochrome excitation in neighboring light channels. In addition, z-stack series of images can be easily combined to assemble three-dimensional reconstruction of many of the unique structures within adipose tissue.

Section snippets

Reagents and buffers

  • 16% paraformaldehyde (PFA) 16% EM Grade (Electron Microscopy Sciences, Hatfield, PA; Cat. # 15710)

  • Phosphate-buffered saline (PBS), pH 7.4 (GIBCO Invitrogen, Carlsbad, CA; Cat. # 10010-023)

  • Bovine serum albumin (BSA) (Sigma-Aldrich, St. Louis, MO; Cat. # A7030)

  • Saponin (Sigma-Aldrich, St. Louis, MO; Cat. # 47036)

  • Glycerol (Sigma-Aldrich, St. Louis, MO; Cat. # G5516)

  • Tris base (Tris (Hydroxymethyl)Aminomethane) (EMD, Darmstadt, Germany; Cat. # 9230)

  • Fixing buffer: 1% PFA in PBS, pH 7.4 (v/v)

  • Blocking

Methods

In order to best preserve structure and cellular components within adipose tissue and eliminate background fluorescence, perfusion fixation is recommended. Lower concentrations of fixative are typically employed to minimize auto-fluorescence. In addition, a gentle postfixation is beneficial immediately after tissue is removed. All buffers should be used at room temperature unless stated otherwise.

This protocol is routinely applied for confocal imaging of mouse adipose tissue from different

Summary

The protocols and images captured above provide a new depth of insight into the architecture and cell–cell interactions in adipose tissue with better resolution than LM techniques. While limited to static measurements, it is possible to use similar imaging modalities to evaluate dynamic leukocyte trafficking events into adipose tissue (Nishimura et al., 2008). We hope that these protocols provide a starting point for many researchers to explore and identify novel markers for use in adipose

Acknowledgments

This work was supported by NIH Grants DK-090262 and DK-092873. This work used services at the University of Michigan NORC supported by NIH Grant DK-089503 and the Michigan Diabetes Research and Training Center funded by P60-DK-020572 from the National Institute of Diabetes and Digestive and Kidney Diseases. Special thanks to the Morphology and Image Analysis Core for training and equipment use.

References (34)

  • J. Han et al.

    The spatiotemporal development of adipose tissue

    Development

    (2011)
  • J.-F. Jasmin et al.

    Caveolins and caveolae

    (2012)
  • Y.Y. Kimura et al.

    Time course of de novo adipogenesis in matrigel by gelatin microspheres incorporating basic fibroblast growth factor

    Tissue Engineering (United States)

    (2002)
  • J. Lee et al.

    Lean, but not obese, fat is enriched for a unique population of regulatory T cells that affect metabolic parameters

    Nature Medicine

    (2009)
  • L. Liu et al.

    Angiogenesis inhibitors may regulate adiposity

    Nutrition Reviews

    (2003)
  • C.N. Lumeng et al.

    Obesity induces a phenotypic switch in adipose tissue macrophage polarization

    Journal of Clinical Investigation

    (2007)
  • C.N. Lumeng et al.

    Phenotypic switching of adipose tissue macrophages with obesity is generated by spatiotemporal differences in macrophage subtypes

    Diabetes

    (2008)
  • Cited by (41)

    • Epigenetic regulation of inflammatory factors in adipose tissue

      2021, Biochimica et Biophysica Acta - Molecular and Cell Biology of Lipids
      Citation Excerpt :

      Although the latter study suggests a causal role of DNMT3A in promoting inflammation, further studies are required to delineate the cellular origin of the pro-inflammatory signal, given that FABP4 is expressed in several non-adipogenic cells, including macrophages. In addition to mature adipocytes, various stromal and vascular cell types compose the adipose tissue, including adipocyte progenitors, endothelial cells and adipose tissue–resident immune cells [29]. Of these resident immune cells, macrophages are the majority.

    • Visualization and analysis of whole depot adipose tissue neural innervation

      2021, iScience
      Citation Excerpt :

      It has been suggested that WAT lacks parasympathetic innervation (Giordano et al., 2006), indicating that precise control of vasodilation is either not required in WAT or regulated entirely by sympathetic and sensory innervation. Neurovascular staining of whole mount scWAT was performed by co-staining with pan-neuronal markers and IB4, a marker for vasculature, as it binds to erythrocytes and endothelial cells (Martinez-Santibanez et al., 2014; Ernst and Christie, 2006; Peters and Goldstein, 1979; Ismail et al., 2003; Gorakshakar and Ghosh, 2016) and effectively marks vessels smaller than 50 μm in diameter. Whole mount axi-scWAT tiled Zmax projections exposed a dense vascular and lymphatic network residing in the scWAT depot (Figure 6A).

    View all citing articles on Scopus
    View full text