In advanced stages of hepatic fibrosis, the liver sinusoidal endothelium transforms to vascular endothelium with accompanying expression of factor VIII-related antigen (FVIIIRAg), a phenotypic marker of vascular endothelial cells. Liver fibrosis has been shown to be associated with aging and was found to be prevalent in elderly cadavers. Using immunohistochemistry, we studied FVIIIRAg expression in the livers of elderly cadavers with progressive stages of fibrosis. The vascular endothelium of portal tracts and central veins was stained for FVIIIRAg, providing an internal positive control. The incidence of FVIIIRAg expression was low in the sinusoids of livers that showed minimal fibrosis or perisinusoidal fibrosis but was increased in livers with advanced fibrosis (i.e., septa formation, bridging fibrosis, and cirrhosis). FVIIIRAg positive sinusoidal endothelial cells were distributed in loose aggregates in the periportal, periseptal, and midlobular parenchyma and were found less frequently in the centrilobular area. FVIIIRAg immune deposits appeared patchy and discontinuous along the sinusoidal lining, likely representing focalized transformation of sinusoidal to vascular endothelium. There was a discrete localization of FVIIIRAg immunoreactivity in the foci of severe parenchymal fibrosis.
The liver parenchyma is extensively supplied by sinusoids that are lined by endothelial cells, which contain fenestrae but lack a continuous basal lamina when viewed by electron microscopy [
Factor VIII-related antigen (FVIIIRAg), also known as von Willebrand factor, is a multimeric glycoprotein that binds and stabilizes the coagulation factor VIII as well as mediating platelet adhesion to injured vessels [
Fibrotic changes are often seen in aged livers [
Paraffin embedded liver tissue of embalmed elderly cadavers from our previous study provided the source of specimens in the present investigation [
To these samples, we added 12 additional liver specimens from the 2012 Fall Anatomy Course. The embalming fluid contained formaldehyde (2.2%), methanol (24.9%), and phenol (25%). As previously described [
The primary antibody was rabbit polyclonal FVIIIRAg, purchased from Cell Marque (Rocklin, CA). The secondary antibody was anti-rabbit polymer-horseradish peroxidase (polymer-HRP) from the Dako EnVision + System-HRP/DAB (Dako, Carpinteria, CA, USA). Trilogy was obtained from Cell Marque. It is an EDTA-based solution that combines the steps of deparaffinization, rehydration, and antigen unmasking in immunohistochemical staining of paraffin embedded tissue sections—vide infra.
Paraffin liver sections were cut at a thickness of 5
For each individual liver, one section was used for assessment of liver fibrosis by Sirius stain for collagens and two nonoverlapped sections were used for evaluation of FVIIIRAg immunostaining. These liver sections—approximately 1
The scoring was performed based on the presence/absence of FVIIIRAg staining in the liver lobular parenchyma. For each liver, two nonoverlapped sections were examined. These liver sections provided well-marked liver lobules for the analysis—vide supra. A liver was scored positive for FVIIIRAg expression when the staining was detected—focal, scattered, or diffuse—in any part of the lobular parenchyma. A negative case was scored when no staining was visible. The incidence of FVIIIRAg expression was expressed as the number positive cases per number of livers examined. The difference in the incidence between groups was analyzed by chi-square test, and
To test whether FVIIIRAg can be used as a phenotypic marker for the vascular endothelium in the cadaveric liver, we examined the antigen immunoreactivity in the endothelium of portal tracts and central veins (also called terminal hepatic venules). Figure
Immunostaining of portal vascular endothelium for FVIIIRAg. FVIIIRAg immunoreactivity (brown color) can be seen in the endothelium of the portal venule, V, and hepatic arterioles, A. The arrows mark bile ductules, which are negative for FVIIIRAg.
Immunostaining of vascular endothelium of central veins for FVIIIRAg. (a) Fibrotic central vein. The black lines indicate the wall of the vein (measured >10
FVIIIRAg immunostaining of sinusoidal endothelial cells in the liver lobules. (a) Representative FVIIIRAg immunopositive endothelial cell (arrow) in the sinusoidal lining. The FVIIIRAg immune deposits are visible as fine granules (brown color)—less than 1
Endothelial cells lined the lumen of liver sinusoids. As illustrated in Figure
FVIIIRAg immunostaining of sinusoidal endothelial cells from a cirrhotic liver. (a) The sinusoid, S, is seen in transverse section. The black arrow labels anendothelial cell displaying granular immune deposits of FVIIIRAg (brown color) in the cytoplasm. FVIIIRAg-negative endothelial cell (blue arrow) is also present in the same sinusoidal lining. The wall of this sinusoid is only slightly thickened with fibrous tissue. (b) The wall of this sinusoid, S, sectioned longitudinally, is prominently thickened with fibrous tissue—slightly gray in the hematoxylin counterstained section. The endothelial cells reveal granular FVIIIRAg immunostaining (brown color) in the cytoplasm.
In many specimens, FVIIIRAg stained platelets—measured 2–4
Immunostaining of platelets for FVIIIRAg. (a) This image shows platelets (arrows) scattered in the lumen of sinusoids. (b) The platelets, measured 2–4
We then determined the incidence of FVIIIRAg expression—positive staining—in the liver lobules of the aged cadavers with progressive stages of fibrosis. Expression of FVIIIRAg in sinusoidal endothelium was found in one of the nine livers showing minimal fibrotic change and two of the eight livers with perisinusoidal/pericellular fibrosis. The incidence of livers showing positive FVIIIRAg staining was increased in the livers with advanced fibrosis—septa formation (8/15 livers), bridging fibrosis with linking septa (11/15 livers), and cirrhosis (3/5 livers). As summarized in Table
Incidence of FVIIIRAg expression in sinusoidal endothelium of aged cadaveric livers with progressive stages of fibrosis.
Stages of fibrosis | Incidence |
---|---|
Minimal fibrosis | 1/9 |
Perisinusoidal/pericellular fibrosis | 2/8 |
Septa formation | 8/15a |
Bridging fibrosis | 11/15b |
Cirrhosis | 3/5c |
Incidence was expressed as number of positive cases/number of livers.
Statistics were performed by the chi-square test. A value of
aSepta formation versus minimal fibrosis
bBridging fibrosis versus minimal fibrosis,
cCirrhosis versus minimal fibrosis,
Next, we examined the anatomical sites of FVIIIRAg immunoreaction within the liver lobules. FVIIIRAg stained sinusoidal endothelial cells were commonly observed in the periportal parenchyma nearby enlarged fibrotic portal tracts (Figure
(a) Periportal parenchyma. FVIIIRAg immunoreactive sinusoidal endothelial cells are seen in loose aggregate close by the portal tract, PT. The immune deposits of FVIIIRAg (brown color) are patchy and discontinuous along the sinusoidal lining. Note FVIIIRAg immunostained blood vessels (arrows) in the portal tract. (b) Developing septum and periseptal parenchyma. FVIIIRAg immunoreactive sinusoids (brown color) can be seen nearby the septum, ST, particularly in the parenchyma around the growing front of the septum. The immunostaining reaction appears patchy and discontinuous along the sinusoidal lining. Note immunoreactive blood vessels (arrows) in the septum. (c) Bridging septum and periseptal parenchyma. This shows FVIIIRAg-positive endothelial cells (brown color) in the sinusoids in the vicinity of the septum, ST.
(a) Midlobular parenchyma; (b) centrilobular parenchyma. FVIIIRAg-positive sinusoidal endothelial cells are distributed in loose aggregates in the parenchyma. The immune deposits (brown color) are patchy and discontinuous along the sinusoidal lining. CV = central vein. FVIIIRAg-positive sinusoidal endothelial cells (brown color) are also seen scattered in small groups of a few cells (c) and singly (d) in the lobular parenchyma.
((a) and (b)) These illustrate localized distribution of FVIIIRAg immunoreactive sinusoids in the foci of severe parenchymal fibrosis. The fibrous lesion is extensive—verified by Sirius red stain for collagens in a serial section—extending from the centrilobular to periportal areas. The immune deposits of FVIIIRAg (brown color) are prominent but are patchy along the sinusoidal lining. The arrow labels a scar tissue that appears slightly gray after hematoxylin counterstaining. CV = central vein.
Septa formation and bridging septa are advanced fibrotic changes associated with the stages of septal fibrosis and cirrhosis. The septa are made up of vascularized fibrous tissue containing ductular structures and mesenchymal cells [
FVIIIRAg immunostaining of vasculature in fibrous septa and microscar. These sections were poststained with Sirius red for collagens. (a) Developing septum. The septum discloses a vein (arrow) of moderate length, which is stained positively for FVIIIRAg (brown color). (b) Bridging septum. This rather broad septum contains vessels (arrows) marked by positive FVIIIRAg staining (brown color). (c) Microscar. The scar—measured 20–30
Small fibrous scars, measured less than 100
In this study, we first documented the localization of FVIIIRAg in the vascular endothelium of portal vessels and central veins, which are the systemic vasculature. The findings are consistent with the notion that FVIIIRAg is a specific phenotypic marker of vascular endothelial cells, thereby validating its expression in the lobular sinusoids as representative of phenotypic transformation to vascular endothelium in the aged cadaveric livers. In support of this contention, we have examined three cases of archival, deidentified adult human livers with little or no fibrotic changes and found that FVIIIRAg staining was uniformly present in the endothelium of portal vessels and central veins but was absent from the sinusoidal endothelium of the lobular parenchyma (unpublished observations). These data are in accord with the observations of others in the human liver with nearly normal histology [
We found that the incidence of FVIIIRAg expression was low in the livers with minimal fibrotic change or perisinusoidal/pericellular fibrosis, while the incidence was significantly higher or showed a trend of increase in the livers with advanced stages of septal fibrosis and cirrhosis. While these findings principally corroborate the studies of others reporting transformation of liver sinusoidal endothelium to FVIIIRAg-positive cells in advanced hepatic fibrotic disease of various causes [
It is known that the pathogenesis of vascular endothelium in the liver lobules is accompanied by the formation of a basement membrane beneath the endothelium in the space of Disse. Indeed, we found in our previous study [
Our study revealed that the incidence of FVIIIRAg expression was low in the aged liver of cadavers with minimal fibrotic change or perisinusoidal/pericellular fibrosis. The findings suggest that, at these early stages of liver fibrosis, vessels within the liver lobules remain largely sinusoidal and, thus, may have little impact on the exchange of materials between the sinusoidal blood and liver parenchyma. However, we also found that, even in the advanced stages of liver fibrosis in which the incidence of positive FVIIIRAg staining is higher, the immunostaining of sinusoidal endothelium is not diffusely panlobular but is characteristically patchy and discontinuous along the sinusoidal lining. Accordingly, this staining pattern reflects that the transformation to vascular endothelium is focal, rather than diffuse, in fashion. Thus, in the parenchyma of these fibrotic livers, while many sinusoidal vessels have become capillarized, others remain sinusoidal. Physiologically, in the areas of the parenchyma where sinusoidal capillarization prevails, periportal, periseptal, midlobular, or foci with extensive fibrosis, the exchange of materials across the endothelium between the sinusoidal blood and liver parenchyma could be compromised. This effect may lead to localized dysfunction of hepatocytes. Conversely, the presence of nontransformed sinusoidal endothelium could provide a local pathway for the bidirectional exchange of metabolites, thereby sustaining local hepatic functions. Nonetheless, the lesions could collectively have resulted in a substantial loss of hepatic function in these elderly individuals and may have aggravated their existing health conditions.
It has been reported that the FVIIIRAg immunostaining was enhanced in liver sinusoids of elderly humans, baboons, and rats with normal histology, whereas it was undetectable in the respective younger counterparts [
In conclusion, FVIIIRAg is a reliable immunohistochemical marker for vascular endothelium, providing a tool for detecting the transformations of sinusoidal endothelial cells to vascular endothelial cells and of sinusoids to capillaries within the liver lobules of elderly cadavers. The incidence of vascular endothelium formation is low in the early stages of hepatic fibrosis but becomes prevalent in the advanced stages of fibrosis. The replacement of sinusoidal endothelium by the vascular type in the liver parenchyma appears to be focalized, rather than diffuse, in fashion, and thus its impairment of hepatic functions may be localized. By examining the aged cadaveric livers, our study provides significant scientific information that advances our understanding of the pathogenesis of sinusoidal transformation to vascular endothelium in association with hepatic fibrosis in the aging human.
The authors declare that they have no conflict of interests regarding the publication of this paper.
The authors wish to acknowledge the support by the Research Fund of the Center for Anatomy and Functional Morphology. The authors also thank Dr. Scott L. Friedman, Division of Liver Diseases at Icahn School of Medicine, for providing them with technical support in this project.