- Top of page
- MATERIALS and METHODS
- LITERATURE CITED
Liver sinusoids are lined by a fenestrated endothelium that lacks a basement membrane. Formation of perisinusoidal basement membranes beneath the endothelium is an integral feature of capillarization of sinusoids that is a significant pathology found in advanced fibrosis. Liver fibrosis is prevalent in elderly cadavers; however, basement membrane formation in these liver samples has yet to be studied. Collagen type IV and laminin are major basement membrane proteins and their codistribution around sinusoids provides an immunohistochemical marker of basement membrane formation. Here, we examined the intralobular sites of perisinusoidal basement membrane formation in elderly cadaveric livers having various stages of fibrosis. Collagen IV and laminin codistributed in basement membranes of portal and septal ductular and vascular structures, providing a positive control. In the parenchyma, collagen IV immunostaining of sinusoids was panlobular in all stages of fibrosis, and the stain was continuous along the sinusoids. In contrast, laminin was not detected in livers, showing minimal fibrotic change. It was rarely seen in perisinusoidal/pericellular fibrosis, but frequently in septa formation, bridging fibrosis, and cirrhosis. The laminin stain was patchy, occurring principally in sinusoids of periportal and periseptal areas, less commonly in mid-lobular and rarely in centrilobular areas. Consecutive sections revealed that laminin codistributed with collagen IV in these sinusoidal locations, thus marking the sites of perisinusoidal basement membrane formation in aged fibrotic livers. This development is presumably related to aging of the liver and exacerbated by liver injury caused by advanced liver fibrosis, possibly resulting in sinusoidal capillarization. Anat Rec, 296:953–964, 2013. © 2013 Wiley Periodicals, Inc.
The hepatic sinusoids are lined by a fenestrated endothelium and lack an electron microscopically continuous basement membrane that is present in most systemic blood vessels (Mak and Lieber, 1984); thus, liver sinusoids are classified as discontinuous capillaries (Simionescu and Simionescu, 1988). In the human liver, collagen type IV is present in the space of Disse (perisinusoidal space) surrounding the sinusoids not associated with laminin (Hahn et al., 1980), an integral basement membrane glycoprotein found in most basement membranes (Timpl et al., 1979). Other extracellular matrix proteins such as collagen types I and III and fibronectin are also present in smaller amounts in the space (Wells, 2007).
Capillarization of sinusoids is a significant pathology found in advanced stages of liver fibrosis. It was first described by Schaffner and Popper (1963) in human alcoholic cirrhosis and then observed in chronic liver disease of various causes, as well as in a variety of experimental models of fibrosis (Horn et al., 1987; Babbs et al., 1990; Jezequel et al., 1990; Martinez-Hernandez and Martinez, 1991; Nakayama et al., 1991; Bhunchet and Fujieda, 1993; Urashima et al., 1993; Dubuisson et al., 1995; Xu et al., 2003). As noted by electron microscopy, formation of a continuous basement membrane beneath the endothelial cells is a main feature of sinusoidal capillarization. This development, concomitant with the loss of endothelial fenestra, transforms the sinusoids into continuous capillaries. Laminin production is enhanced and its deposition is localized immunohistochemically with collagen IV in the perisinusoidal basement membrane (Jezequel et al., 1990; Nakayama et al., 1991; Dubuisson et al., 1995). These changes, along with increased presence of type I and III collagens and other matrix proteins in the space of Disse, impair the normal exchange of materials in particular macromolecules between sinusoids and hepatocytes (Martinez-Hernandez and Martinez, 1991).
Hepatic fibrosis, including advanced stages of fibrosis, is prevalent in elderly cadavers with diverse causes of death (Mak et al., 2012a), but the intralobular anatomical sites of perisinusoidal basement membrane formation in these samples have yet to be studied. As collagen type IV and laminin are major proteins of basement membranes, their codistribution in the sinusoidal lining provides an immunohistochemical marker of perisinusoidal basement membrane formation (Jezequel et al., 1990; Nakayama et al., 1991; Dubuisson et al., 1995). Accordingly, the aims of this study were to assess by immunolight microscopy the distribution of collagen IV and laminin in the liver lobules of elderly cadavers, showing progressive fibrotic changes, and to determine the codistribution of these proteins in the sinusoids that, thus, marks the anatomical sites of perisinusoidal basement membrane formation in hepatic fibrosis.
- Top of page
- MATERIALS and METHODS
- LITERATURE CITED
In this study, we first documented the localization of collagen type IV and laminin in the basement membranes of biliary ductular and vascular structures in the portal tract by immunohistochemistry using antibodies to collagen type IV and laminin. The finding confirmed that collagen IV and laminin are normal constituents of basement membranes, and thus validating their codistribution in the liver sinusoids as an immunohistochemical marker of preisinusoidal basement formation.
Generally, basement membranes are visualized using the electron microscope. However, the cadaveric livers were preserved in embalming solutions and the quality of tissue preservation is not optimal for ultrastructural assessment. Accordingly, we sought to localize perisinusoidal basement membrane formation by determining collagen IV and laminin codistribution using immunolight microscopy, which offers several distinct advantages over electron microscopy. These include: (1) a larger liver section (∼1 × 1 cm2 vs. 1 × 1 mm2 in electron microscopy) with well-oriented liver lobules can be studied, facilitating identification of the anatomical sites of perisinusoidal basement formation within the lobules; (2) the process associated with immuolight microscopy is technically less time consuming; and (3) the cost is less prohibitive.
We determined expression of collagen IV and laminin in the liver parenchyma of elderly cadavers having progressive fibrotic changes. We found a striking difference between the staining patterns of these proteins in the liver lobules. Collagen IV is expressed throughout the liver lobules, whether disclosing minimal fibrotic change or advanced fibrosis, reflecting that the protein is a normal component of the sinusoidal lining, regularly present irrespective of hepatic pathology. The collagen IV immunostain is continuous along the sinusoids, rather than discontinuous. In fact, the panlobular distribution of collagen IV in the elderly cadavers is also a feature occurring in the adult human liver with a near normal histology (unpublished observation), in patients with varying degrees of alcoholic fibrosis, or in experimental hepatic fibrogenesis in rats (Hahn et al., 1980; Jezequel et al., 1990; Nakayama et al., 1991; Dubuisson et al., 1995). It is worth noting that despite the deposition of collagen IV in the sinusoidal lining, a typical basement membrane could not be detected beneath the sinusoidal endothelium by electron microscopy (Hahn et al., 1980).
Compared with collagen IV, laminin is generally not expressed in the liver in the absence of active fibrogenesis, whether of human or experimental animals, except in the portal field (Hahn et al., 1980; Jezequel et al., 1990; Nakayama et al., 1991; Tsutsumi et al., 1993; Dubuissen et al., 1995). Similarly, we found no laminin expression in the lobular parenchyma, showing minimal fibrotic change. There was a case of laminin expression detected in the liver that had perisinusoidal/pericellular fibrosis, but laminin expression became prevalent in advanced fibrosis of septa formation, bridging fibrosis (sometimes designated as precirrhosis), and cirrhosis. The laminin sinusoidal stain is patchy and speckled, which is in striking contrast to the continuous fashion for collagen IV. We also observed that the laminin-immunopositive sinusoids are discretely localized, appearing in the periportal and periseptal parenchyma, along the developing septa, and at the front of expanding fibrous septa, as well as in microscopic fibrous scars. The laminin immunostain in these anatomical locations was found to codistribute with collagen IV, and thus marking the sites of perisinusoidal basement membrane formation within the liver lobules. However, our observations also showed that in the aged cadaveric livers collagen type IV immunoreactivity occurs in many sinusoids that are not accompanied by laminin stain, and thus lacks a codistribution with laminin in a vast region of the liver lobules, particularly in the centrilobular parenchyma. This finding suggests that although many vessels in the liver parenchyma remain sinusoidal, others have become capillarized.
The localization of collagen IV alone in the sinusoidal lining reflects a basement membrane-like structure surrounding the perisinusoidal surface in the liver of aged cadavers. Physiologically, the presence of basement-like material composed of collagen IV in the space of Disse does not appear to present a permeability barrier for metabolic exchange between the sinusoidal circulation and the parenchymal cells. However, the addition of laminin to collagen IV already in the space of Disse will restructure the basement membrane-like material into a continuous basement membrane beneath the endothelium, which likely constitutes a functionally significant filtration barrier for bidirectional exchange of macromolecules between sinusoids and hepatocytes, leading to liver dysfunction in advanced fibrosis (Martinez-Hernandez, 1991) and to disease susceptibility in older people (McLean et al., 2003).
The production of perisinusoidal basement membrane is an integral feature associated with the pathogenesis of capillarization of hepatic sinusoids in advanced fibrogenesis of diverse causes. Therefore, the development of basement membrane might be presumed to result in the formation of capillarized sinusoids in the aged liver. There is no study, to date, specifically looking at the incidence of sinusoidal capillarization in the aging human liver. However, a nearly identical histological lesion, termed pseudocapillarization of sinusoids, has been described in the liver of older people even in the absence of overt fibrosis (McLean et al., 2003; Le Couteur et al., 2008). This pathological process appears to be an age-related change and is indicated by thickening of sinusoidal endothelial cells with reduction in fenestration, sporadic development of basal lamina in the Disse space, and slight perisinusoidal collagenization. Thus, it is possible that the pathogenesis of perisinusoidal basement membrane in this study is also related to aging of the liver and aggravated by advanced fibrosis.