Abstract
Hepatocellular vacuolation can be a diagnostic challenge since cytoplasmic accumulations of various substances (lipid, water, phospholipids, glycogen, and plasma) can have a similar morphology. Cytoplasmic accumulation of phospholipids following administration of cationic amphiphilic drugs (CAD) can be particularly difficult to differentiate from nonphosphorylated lipid accumulations at the light microscopic level. Histochemical methods (Sudan Black, Oil Red-O, Nile Blue, etc.) can be used to identify both nonphosphorylated and/or phosphorylated lipid accumulations, but these techniques require non-paraffin-embedded tissue and are only moderately sensitive. Thus, electron microscopy is often utilized to achieve a definitive diagnosis based upon the characteristic morphologic features of phospholipid accumulations; however, this is a low throughput and labor intense procedure. In this report, we describe the use of immunohistochemical staining for LAMP-2 (a lysosome-associated protein) and adipophilin (a protein that forms the membrane around non-lysosomal lipid droplets) to differentiate phospholipidosis and lipidosis, respectively in the livers of rats. This staining procedure can be performed on formalin-fixed paraffin embedded tissues, is more sensitive than histochemistry, and easier to perform than ultrastructural evaluation.
Introduction
Cytoplasmic accumulation of phospholipids following administration of cationic amphiphilic drugs (CAD) can be particularly difficult to differentiate from nonphosphorylated lipid accumulations at the light microscopic level even though their subcellular/organelle locations, mechanisms of formation, and ultrastructural morphology are quite different. It has been well established that lysosomes play an integral role in the etiology of CAD-induced phospholipidosis (Reasor, 1984; Kodavanti, 1990). Subcellular localization studies demonstrated that lysosomes are the primary site for phospholipid accumulation (Matsuzawa, 1980; Kacew, 1987). Electron microscopy is usually required to elucidate the morphologic features associated with the storage of this material (Kacew, 1997). The common feature to all forms of this alteration is distention of membrane-bound cytoplasmic vacuoles.
These intravesicular, single-membrane, electron-dense, concentric configurations contain primarily stored phospholipids, which results from an inability of the cell to catabolize this substrate (Hostetler, 1984). The most common ultrastructural manifestations are concentric lamellar bodies. These lamellae arise from the homogenous lysosomal matrix that undergoes further transformation into amorphous granular or membranous material (Reasor, 1981). The presence of clear cytoplasmic vacuoles is another morphologic manifestation of CAD storage. These distended membrane-bound vesicles contain predominantly electron-lucent material consistent with phospholipid as determined by acid hematein histochemical staining (Benitz 1968; Thelmo, 1978; Frisch, 1979; Ruben, 1989).
To identify phospholipid and nonphosphorylated lipid histochemically, non-paraffin-embedded tissue samples are required, which are often not available. Thus, an alternative method to differentiate nonphosphorylated lipid accumulations from phospholipid or other lysosomal accumulations, which uses formalin-fixed, paraffin-embedded tissue samples would be useful. One approach is to immunohistochemically label the membrane proteins most commonly associated with these different lipid accumulations: LAMP-2 protein for phospholipid and adipophilin protein for nonphosphorylated lipids.
The lysosome-associated membrane protein LAMP-2 and structurally similar but genomically distinct (located on a different gene with only 37% sequence homology) LAMP-1 are type 1 glycoproteins localized primarily on the outer membrane of the lysosome and are recognized as major constituents of the lysosomal membrane (Chen, 1988; Fambrough, 1988; Howe, 1988; Viitala, 1988; Cha, 1990; Fukuda, 1991; Peters, 1994; Konecki, 1995; Cuervo, 2000). Roles for LAMP-2 in cell-cell or cell-extracellular matrix adhesion and in maturation of autophagic vacuoles have also been proposed (Lippincott-Schwartz, 1987; Carlsson, 1988; Licheter-Konecki, 1999; Saitoh, 1992; Tanaka, 2000). LAMP proteins can also be detected in lower amounts in endosomes and in the plasma membrane (Furuno and Himeno, 1989; Akasaki, 1993; Furuta, 1999). The lamp2 gene undergoes alternative splicing resulting in different isoforms with different tissue distribution (Gough, 1995; Hatem, 1995; Konecki, 1995). In humans, hLAMP-2B predominates in brain and muscle, and hLAMP-2A is the predominant form in placenta, lung, liver, kidney, pancreas, and prostate gland. LAMP-1 is an integral membrane protein that is transported from the trans-Golgi network to endosomes and then lysosomes. Upon cell activation, LAMP-1 is transferred to the plasma membrane and promotes adhesion of human peripheral blood mononuclear cells to vascular endothelium (Himeno, 1989; Arterburn, 1990; Zot 1990). In contrast, LAMP-2 is ubiquitously distributed and evolutionarily conserved. It is involved in the direct uptake of cytosolic proteins in lysosomes and also participates in the endocytic pathway (Lippincott-Schwartz, 1987; Carlsson, 1988; Saitoh 1992; Cuervo, 1996; Licheter-Konecki, 1999; Tanaka 2000).
Originally, the term adipose-differentiation-related protein (ADRP) was coined to reflect this protein’s early appearance during adipocyte differentiation and its apparent specific expression in adipocytes (Jiang, 1992; Eisinger, 1993). However, this protein was renamed adipophilin to emphasize that it is a more general lipid-binding adipose-loving protein (Heid, 1998). Adipophilin functions, together with intermediate filaments, in the packaging, transport and deposition of cytosolic lipid droplets (Heid 1996, 1998; Gao, 1999, 2000). Adipophilin has been strongly conserved during evolution as evidenced by its high sequence homology and immunological cross-reactivity over a range of species (human, bovine, pig, dog, rat, and mouse; Heid, 1998). Heid et al. (1998) demonstrated that adipophilin occurs in a wide range of cells including fibroblasts, endothelial cells, and epithelial cells at the surface of lipid droplets.
Adipophilin not only encircles droplets mainly composed of triglycerides, but also droplets rich in cholesterol stored as steroid hormone precursors in adrenocortical cells (Weiss, 1988). Expression of adipophilin in tissues is restricted to certain cell types, such as lactating mammary epithelial cells, adrenal cortex cells, testicular Sertoli and Leydig cells, and steatotic hepatocytes in alcoholic liver cirrhosis (Heid, 1998). Heid et al. proposed adipophilin as a new marker for the identification of specialized differentiated cells containing lipid droplets and for diseases associated with fat-accumulating cells. Lee et al. (1994) suggested one potential diagnostic use for adipophilin is to assess early or minor fatty change in liver biopsies, since microvesicular fat inclusions in hepatocytes may be difficult to distinguish morphologically from glycogen accumulation or certain degenerative cellular changes (Bannasch 1968; Lee, 1994). In addition, conventional fat-staining histochemical methods such as Sudan Black and Oil Red-O staining require non-paraffin-embedded tissue and are only moderately sensitive (Lee, 1994). However, glycogen accumulation in the rat liver can be identified using the Periodic acid-Schiff (PAS) reaction (Corrin, 1968).
Taking advantage of the membrane-bound nature of phospholipid and lipid cytoplasmic accumulations is an approach investigated in this study. Thus, immunohistochemical staining for LAMP-2 and adipophilin proteins were evaluated in liver tissue from non-fasted control rats (glycogen accumulation; Corrin, 1968), fasted control rats (hepatic lipidosis; Delzenne, 1997), and rats treated with a CAD (hepatic phospholipidosis) to determine if they can be used to distinguish the different lipid-containing vacuoles.
Methods
Animals and Tissues
Female Sprague–Dawley (Crl:CD(SD)IGS BR) rats, which were approximately 6–8 weeks old and weighed 125–225 grams at study initiation, were obtained from Charles River Laboratories. All in-life animal procedures were conducted under an IACUC approved protocol and in an AAALC accredited facility. Standard procedures/conditions for animal care, feeding, and maintenance of room, caging, and environment were applied. For the hepatic lipidosis study, animals were fasted for 18 hours, euthanized by CO2 and exanguinated via cardiac puncture. Nonfasted animals served as controls. Tissues were fixed in 10% buffered formalin, embedded in paraffin, sectioned at 3 microns, and stained with hematoxylin and eosin for histologic evaluation. For the hepatic phospholipidosis samples, archival paraffin-embedded tissue blocks were utilized from a 7-day dose-range finding study of an experimental compound that had cationic amphiphilic properties and caused disseminated phospholipidosis in dosed-rats as observed by light microscopic examination and confirmed via electron microscopy.
Immunohistochemistry (IHC)
Tissue samples were sectioned at 3 microns, placed on charged slides, air-dried, and loaded onto the Discovery XT Autostainer (Ventana Medical Systems). All solutions used for automated immunohistochemistry were from Ventana Medical Systems unless otherwise specified. Slides underwent de-paraffinization, heat-induced epitope retrieval (CC1 solution, mild protocol), and staining online using a heat protocol. Slides were incubated four minutes with Inhibitor D, LAMP-2 slides were incubated with serum free Dako blocker (Dako) for 32 minutes, and slides were incubated for 4 minutes (adipophilin) or 16 minutes (LAMP-2) with Endogenous Biotin Blocking Kit reagent followed by rabbit anti-LAMP-2 antibody (1:100; Zymed) or guinea pig anti-adipophilin antibody (1:1,000; Progen) for 32 minutes. Next, the appropriate biotinylated link was applied (32 minutes, 1:200, Vector Laboratories, rabbit or guinea pig), followed by the DAB Map Detection Kit. Slides were counterstained with hematoxylin for four minutes, blued with Bluing Reagent for four minutes, removed from the autostainer, washed in warm soapy water, dehydrated through graded alcohols, cleared in xylene, and mounted with Permount.
Electron Microscopy
Tissue samples were retrieved from neutral buffered formalin, minced, further fixed in 0.1M cacodylate buffered 2.5% glutaraldehyde +4% formaldehyde, post-fixed in 0.1M cacodylate buffered 1% osmium tetroxide, and routinely processed through epoxy resin and cured. Thin sections were prepared, routinely stained with 2% uranyl acetate (aqueous) and Reynold’s Lead Citrate, and examined on an FEI-Philips CM100 BioTWIN transmission electron microscope.
Results
Microscopic Evaluation of Live
The hepatocyte cytoplasm was expanded by irregular non-membrane bound clear spaces (lacy), which was characteristic of normal glycogen storage in a nonfasted state. Non-fasted rats also had rare lipid vacuoles evident within hepatocytes (Figure 1A). Rats fasted for 18 hours exhibited mild lipid accumulation within hepatocytes, which was characterized by one to multiple variably sized, discrete, round, membrane-bound clear vacuoles within the cytoplasm. In addition, these rats had a depletion of their hepatocyte glycogen stores as evidenced by the loss of the lacy staining pattern observed in the nonfasted rats (Figure 1B). CAD-treated rats had diffuse, moderate to marked vacuolation of hepatocytes that were characterized by multiple irregular to round membrane-bound vacuoles within the cytoplasm (Figure 1C). In addition, multifocal aggregates of foamy macrophages were present within the sinusoidal space of the centrilobular regions. Electron microscopic evaluation of these liver tissues revealed the presence of multiple membrane-bound, variably sized laminated structures consist with phospholipid accumulations, which corresponded to the intracytoplasmic hepatocellular vacuoles observed via light microscopy (Figures 2A and 2B).
Lamp-2 IHC
Livers from non-fasted and fasted control rats contained punctate cytoplasmic staining within hepatocytes which was distributed in a linear fashion emanating from the edge of the nuclear membrane. Kupffer cells also contained punctate cytoplasmic staining which was less prominent (Figures 3C, 4A and 3E). Hepatocytes from rats treated with a CAD had intense immunostaining of the lysosomal membranes surrounding the cytoplasmic phospholipid accumulations (vacuoles) and a lighter diffuse staining of the intervening cytoplasm (Figures 3G and 4B).
Adipophilin IHC
Livers from fasted control rats were characterized by prominent immunostaining of the membranes surrounding the cytoplasmic lipid vacuoles within hepatocytes, which were located primarily along the cell periphery; whereas, only rare lipid vacuoles stained within hepatocytes from nonfasted control rats. In addition, Ito cells exhibited intense membrane staining of their lipid storage vacuoles (Figures 3B and 3F). The lysosomal membranes surrounding the phospholipid accumulations (vacuoles) within the hepatocytes from rats treated with a CAD did not stain nor did the phospholipid accumulations. Since the CAD-treated rats underwent an overnight fast prior to termination, their hepatocytes contained rare cytoplasmic lipid vacuoles that stained. In addition, lipid storage vacuoles within Ito cells also stained (Figure 3H).
Discussion
Taking advantage of the membrane-bound nature of phospholipid and lipid cytoplasmic accumulations, immunohistochemistry for LAMP-2 and adipophilin proteins, respectively, was employed to differentiate these sources of hepatocellular vacuolation.
Livers from nonfasted (glycogen accumulation) and fasted (lipid accumulation) control rats contained punctate cytoplasmic LAMP-2 immunostaining within hepatocytes; whereas, hepatocytes from rats treated with a CAD exhibited intense immunostaining of the lysosomal membranes surrounding the cytoplasmic phospholipid accumulations. The lysosomal membranes surrounding these phospholipid accumulations did not label immunohistochemically for adipophilin. As expected, the hepatocytes from fasted control rats had prominent adipophilin immunostaining of the membranes surrounding the cytoplasmic lipid vacuoles.
We have demonstrated that immunohistochemistry can be used to differentiate cytoplasmic vacuolation due to phospholipidosis and/or other intralysosomal accumulations from lipidosis in the rat liver utilizing antibodies against LAMP-2 and adipophilin, respectively. We believe these staining methods offer considerable advantages over current methods: (1) they can be performed on formalin-fixed, paraffin-embedded tissues, (2) they are more sensitive than histochemistry and (3) they are easier to perform than ultrastructural evaluation. These immunostains also are quite robust and cross-react with several different species.
Additionally, LAMP immunostaining can be exploited to delineate the lysosomal membranes surrounding phospholipid accumulations within other cell types. We stained the lysosomal membranes surrounding phospholipid accumulations within renal proximal convoluted tubules and tissue macrophages from the same CAD-treated rats using LAMP-2 and LAMP-1 immunohistochemistry, respectively (data not shown). Likewise, adipophilin immunohistochemistry has been used in our laboratory to label lipid storage vacuoles in multiple tissues including adrenal gland, heart, skin (including sebaceous glands), mammary gland, and adipose tissue (data not shown).