Vascular endothelial growth factor increases fenestral permeability in hepatic sinusoidal endothelial cells

Liver International 2003: 23: 467–475 Printed in Denmark. All rights reserved

Copyright r Blackwell Munksgaard 2003

Basic Studies

Vascular endothelial growth factor increases fenestral permeability in hepatic sinusoidal endothelial cells

Yokomori H, Oda M, Yoshimura K, Nagai T, Ogi M, Nomura M, Ishii H. Vascular endothelial growth factor increases fenestral permeability in hepatic sinusoidal endothelial cells. Liver International 2003: 23: 467–475. r Blackwell Munksgaard, 2003 Abstract: Vascular endothelial growth factor (VEGF) is an important regulator of vasculogenesis and vascular permeability. Hepatic sinusoidal endothelial cells (SECs) possess sieve-like pores that form an anastomosing labyrinth structure by the deeply invaginated plasma membrane. Caveolin is the principal structural protein in caveolae. In this study, we examined the role of VEGF on the fenestration and permeability of SECs and the relation with caveolin-1. SECs isolated from rat livers by collagenase infusion method were cultured for 24 h with (10 or 100 ng/ml) or without VEGF. The cells were then examined by transmission and scanning electron microscopy (EM). The expression of caveolin was investigated by confocal immunofluorescence, immunogold EM, and Western blot. Endocytosis and intracellular traffic was studied using horseradish peroxidase (HRP) reaction as a marker of fluid phase transport in SECs. Both transmission and scanning EM showed an increased number of sinusoidal endothelial fenestrae (SEF) in SECs cultured with VEGF. By confocal immunofluorescence, SECs cultured with VEGF displayed prominent caveolin-1-positive aggregates in the cytoplasm, especially surrounding the nucleus region. Immunogold EM depicted increased caveolin-1 reactivity on vesicles and vacuoles of VEGF-treated SECs compared with VEGF-nontreated cells. However, there was no change in the level of caveolin-1 protein expression on Western blot. After HRP injection, an increase of electron-dense tracer filled the SEF in cells treated with VEGF. Our results suggested that VEGF induced fenestration in SECs, accompanied by an increased number of caveolae-like vesicles. Increased caveolin-1 might be associated with vesicle formation but not with fenestration. Increased fenestration may augment hepatic sinusoidal permeability and transendothelial transport.

Liver sinusoids are unique capillaries lined by endothelium expressing open fenestrae without a diaphragm and lacking an underlying basal lamina (1). Fenestrae are known to filter fluids, solutes and particles exchanged between the sinusoidal lumen and the space of Disse (2). In addition to fenestrae, other specialized structures found in hepatic sinusoidal endothelial cells (SECs) are pronounced components of the vacuolar apparatus, showing a variety of vacuolar and Abbreviations: VEGF, vascular endothelial growth factor; SEC, sinusoidal endothelial cell; SEF, sinusoidal endothelial fenestrae; VVO, vesiculo-vacuolar organelles.



Hiroaki Yokomori1, Masaya Oda2, Kazunori Yoshimura3, Toshihiro Nagai4, Mariko Ogi5, Masahiko Nomura3 and Hiromasa Ishii2 1 Department of Internal Medicine, Kitasato Medical Center Hospital, Saitama, Japan, 2 Department of Internal Medicine, School of Medicine, Keio University, Tokyo, Japan, 3 Department of Physiology, Saitama Medical School, Saitama, Japan, 4Electron Microscopy Laboratory, School of Medicine, Keio University, Tokyo, Japan, 5Laboratory of Pathology, Kitasato Medical Center Hospital, Saitama, Japan

Key words: immunogold electron microscopy – sinusoidal endothelial cell – sinusoidal endothelial fenestrae – vesiculo-vacuolar organelles Hiroaki Yokomori, MD, Department of Internal Medicine, Kitasato Institute Medical Center Hospital, 121-1 Arai, Kitamotoshi, Saitama 364-8501, Japan Tel: 1811485-93-1212. Fax: 1811485-93-1239. e-mail: [email protected] Received 24 March 2003, accepted 29 July 2003

vesicular structures such as bristle-coated Golgi complex and macropinocytotic vacuoles (3). Vascular endothelial growth factor (VEGF) was initially characterized as a vascular permeability factor secreted by tumor cells (4), and later found to be also an endothelial growth factor (5). The application of VEGF in vivo directly and rapidly induced fenestrae in the continuous endothelium of skeletal muscle and skin (6), and fenestration was observed in neovessels of VEGF-secreting tumors (7). Caveolae are believed to be responsible for transcytosis, a process by which plasma proteins are transported across the capillary endothelium 467

Yokomori et al. (8). Capillary caveolae lack a visible cytoplasmic coat when viewed by electron microscopy (EM), but a 21 kD transmembrane protein caveolin-1 is associated with their cytoplasmic face (9). Electron microscopic studies have also demonstrated caveolin-1 in caveolae and also fused clustered vesicles, also referred to as vesiculo-vacuolar organelles (VVOs) (10). While VVOs of normal venules are relatively impermeable, VEGF increases the leakage of macromolecules from normal venules through extravasation by way of the VVOs (11). By scanning EM, increased porosity of sinusoidal endothelial fenestrae (SEF) was observed in SECs cultured with VEGF (12). In a preliminary study, we localized caveolin-1 on the Golgi complex, vesicles, and SEF of SECs (13). The present study was designed to examine the immunofluorescent and immunoelectron localization of caveolin-1 in the hepatic SECs, aiming to elucidate the condition under which SECs develop fenestration in response to VEGF.

Iwaki, Tokyo, Japan), six-well dishes (Falcon 3046; Falcon, Lincoln Park, NJ), or two-well dishes (Falcon 3001). Serum-free SEC culture medium consisted of RPMI-1640 with 2 mM Lglutamine and 100 mg/ml gentamicin. The plates were incubated in a CO2 gas incubator at 37 1C for 24 h.

Material and methods

Scanning EM

Experimental animal

SECs cultured on a glass coverslip were fixed in 1.2% glutaldehyde buffered with 1 M cacodylate buffer (pH 7.4) for 1 h at 4 1C, followed by postfixation with 1% osmium teroxide in cacodylate buffer (pH 7.4) for 1 h at 4 1C. After dehydration in a grade series of ethanol solutions, the cultured cells were dried in a critical point apparatus (Hitachi HCR-2, Tokyo, Japan) and coated with gold in a Hitachi vacuum coating unit. The cell surfaces were observed under a scanning EM (Hitachi, Tokyo, Japan) at 15-kV acceleration voltage.

Male Wistar strain rats weighing 150–180 g were used in this study. Animals were housed in individual cages and allowed free access to chow and water until the start of the study. All animal procedures in the present experiments were performed in compliance with the ‘Guide for Care and Use of Laboratory Animals’ by the National Academy of Sciences. In this study, we investigated 10 animals. Isolation, purification, and culture of SECs

The method for the isolation of SECs was modified from the method of Braet et al. (14). In brief, the liver of a male Wistar strain rat was perfused with Ca21–Mg21-free Hanks balanced salt solution followed by 0.6% collagenase A (Sigma type 1) solution via a polyethylene catheter inserted into the portal vein trunk. After incubation of the fragmented tissue in the same solution, the resulting cell suspension was centrifuged at 100g for 10 min to remove the parenchymal cells. The supernatant containing a mixture of sinusoidal liver cells was then layered on the top of a two-step Percoll gradient (25–50%) and centrifuged for 20 min at 900g. The intermediate zone located between the two density layers was enriched in SECs. The purity of SECs was further enhanced by selective adherence of Kupffer cells and spreading of the SECs on collagen. SECs were seeded on collagen-coated coverglass (Becton Dickinson, Biocoat 4426; 468

Treatment of cultured cells with VEGF

Six-hour cultures of hepatic endothelial cells plated in two-well dishes (Biocoate Cell Environment; Beckton Dickinson Labware, Bedford, MA) pretreated with collagen type 1 were washed with L-15 culture medium (37 1C). Recombinant human VEGF (R&D Systems Inc., Minneapolis, MN) at 10 or 100 ng/mL or medium (control) was added to the culture medium and the cells were incubated for 24 h. We investigated 20 dishes for each treatment in the present study.

Transmission EM

SECs cultured with or without VEGF in two-well dishes were directly fixed with 1.2% glutaraldehyde buffered with 0.05 M cacodylate buffer (pH 7.4) for 10 min at room temperature, further fixed in fresh 1.2% glutaraldehyde solution for 1 h at 4 1C, followed by postfixation with 1% osmium tetroxide buffered with 0.05 M cacodylate buffer (pH 7.4) and dehydration in a graded series of ethanol solutions. For transmission EM, couplet or triplet hepatocytes at the bottom of a petri dish were selected under a phase-contrast microscope and embedded in Epon. Ultrathin sections were cut with a diamond knife on an LKB ultramicrotome, stained with uranyl acetate and lead citrate, and observed under a transmission electron microscope (JEM-1200 EX) at 80-kV acceleration voltage.

Caveolin-1 and sinusoidal endothelial fenestrae Confocal immunofluorescence

Cultured SECs were fixed with 4% formaldehyde in phosphate buffered saline (PBS) for 1 h at room temperature for 30 min and then permeabilized for 30 min with 0.1% Triton X-100 in PBS containing 1% bovine serum albumin (BSA). The cells were incubated with primary antibody, anticaveolin-1 (polyclonal antibodies; Santa Cruz Bio., Santa Cruz, CA), at 1:50 dilution at 4 1C overnight. The cells were then incubated with tetra-methylrhodamin isomer R1-conjugated swine anti-rabbit IgG (1:100) at room temperature for 1 h. Each incubation was followed by three washes with PBS. Next, a confocal laser scanning microscope (FV 300; Olympus, Tokyo, Japan) equipped with angon/krypton laser capable of dual excitation and detection was used to visualize the distribution of caveolin-1. Immunoelectron microscopic procedures for ultrastructural localization of caveolin-1

The postembedding method was used. According to a modification of Berryman and Rodewald (15), samples of isolated cultured cells were fixed in periodate-lysin-paraformaldehyde (PLP) for 1 h. After washing the cells for 2 h with several changes of cold 0.1 M phosphate buffer containing 3.5% sucrose and 0.5 mM calcium chloride, free aldehyde was quenched in sucrose–phosphate buffer containing 50 mM ammonium chloride and 0.5 mM calcium chloride (pH 7.4) for 1 h at 0 1C. Phosphate ions were removed by rinsing four times with cold 0.1 M maleate buffer (final pH 6.0) containing 3.5% sucrose, pH 6.5 (four times for 15 min each). The cells were poststained for 10 min at 0 1C with 2% uranyl acetate in sucrose–maleate buffer (final pH 6.0). Distilled water (final pH 4.0), 0.05 M maleate buffer (final pH 4.2), and veronal acetate buffer (final pH 4.2) were also used as vehicles to prepare 2% uranyl acetate for poststaining for 5 min at 0 1C. The cells were dehydrated in 50%, 70%, 90%, and 100% acetone sequentially, at
20 1C for 45 min each, and then embedded at
20 1C in LR Gold (London Resin Co., Hants, UK). To block nonspecific binding sites, the grids were placed on drops of 5% normal goat serum (NGS) diluted in TBS. Then the grids were transferred successively on drops of primary antibody (CAV-1, polyclonal, Santa Cruz Bio.), secondary antibody, and 5-nm colloidal goldconjugated anti-rabbit IgG antibody (Cosmo Bio, Tokyo, Japan). After the final washing, the grids were fixed in glutaraldehyde for 5 min, rinsed in water, stained for 15 min in 2% aqueous osmium tetroxide and then examined under a

JEM 1200 EX electron microscope at an accelerating voltage of 80 kV. Western blotting of caveolin-1 in isolated SECs

SECs were lyzed with 200 mmol/l octyl-a-D glucopyranoside in 100 mM Tris-HCl with 1 mmol/l phenylmethylsulfonyl fluoride for 1 h at 4 1C and centrifuged twice at 2000 rpm for 10 min each. The protein concentrations were determined by the Bradford assay. Samples (50 mg/well) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) under a nonreducing condition on 7.5% acrylamide gel, and electrophoretically transferred onto a nylon membrane (Hybond-N; Amersham). Filters were blocked with 1% dry milk and 1% BSA in 0.02% Tween-20 PBS for 2 h, probed with the anti-caveolin-1 antibody at 1/500 dilution for 1 h, rinsed in PBS, and then probed with biotinylated goat anti-rabbit antibody at 1:5000 dilution for 30 min. After rinsing in PBS, color was developed with SuperSignal West Femto Maximum Sensitivity Substrate (Pierce). Horseradish peroxidase (HRP) tracing of fenestrae

HRP was employed as a tracer to elucidate the pathway of macromolecule extravasation. SECs cultured with 100 ng/ml of VEGF or without VEGF were incubated in a medium containing HRP (10 mg/ml) for 2 min at 37 1C, washed, incubated in HRP-free medium for 3 min at 37 1C, and then washed in HRP-free medium (4 1C) for 10 min (16). The SECs were fixed in 1.5% glutaraldehyde buffered with 0.1 mol/l cacodylate buffer (pH 7.4) and 2.5% sucrose buffer (pH 7.4) for 1 h at 4 1C. After washing in 0.1 mol/l Tris-HCl buffer (pH 7.4) for 5 min at room temperature, the cells were incubated in 0.2% 3,3 0 -diaminobenzine (Sigma) in 0.1 mM Tris buffer (pH 7.4) at room temperature for 30 min, followed by further incubation in 0.2% 3,3 0 -diaminobenzine with 0.01% H2O2 in the same buffer. The cells were washed again with Tris-HCl buffer and then with 0.1 mol/l cacodylate buffer (pH 7.4) for 15 min. The cells were fixed in 1% osmium teroxide and 1% potassium ferrocyanide (Sigma) in 0.1 mM cacodylate buffer (pH 7.4) for 1 h, dehydrated in graded ethanol and embedded in Epon. Ultrathin sections were cut with a diamond knife on an LKB ultramicrotome, stained with uranyl acetate and lead citrate, and observed under a transmission electron microscope (JEM-1200 EX) at 80-kV acceleration voltage. 469

Yokomori et al. Statistics

For scanning EM, 30 cells were examined for each experimental variable in 20 randomly selected fields. The SEM was regularly calibrated at a magnification of  5000, with the specimen in the center. For automatic analysis of fenestrae, images of randomly selected fields were saved. The number of pores in digital images containing a minimum of 20 fenestrae were counted using Winroof (Mitani. Co., Tokyo, Japan). Statistical significance of the difference between two groups was assessed with the Mann–Whitney U-test. A P value less than 0.05 was regarded as indicating a significant difference. Data are expressed as mean7SEM.

extended and contained clusters of fenestrae and sieve plates (Fig. 1a). SECs treated with 100 ng/ml VEGF showed an enormous increase of fenestrae. Long and thin cytoplasmic arms free of fenestrae extended from the nucleus and seemed to organize in huge sieve plates (Fig. 1b). The numbers of fenestrae per square micrometer were: control; 2.470.7 fenesrae/mm2, VEGF 10 ng/ml; 3.471.2 (Po0.05, compared with control), VEGF 100 ng/ml; 4.170.7 (Po0.05, compared with control). Treatment with VEGF significantly increased the number of pores compared with untreated control (Fig. 1c).

Transmission EM Results

Scanning EM

We first investigated the degree of fenestration in SECs by scanning EM. Control SECs were well

a

Transmission EM depicted a normal component of fenestrae in control SECs (Fig. 2a). In the presence of 100 ng/ml VEGF, individual fenestrae were observed to be swollen and interconnected with each other (Fig. 2b).

b 4.1±0.7*

nF/µm2

3.4±1.2*

2.4±0.7

0

c

Control

VEGF 10ng/ml

VEGF 100ng/ml

Fig. 1. Scanning electron micrograph of sinusoidal endothelial cells (SECs) in primary culture: (a) Control SECs show thin cytoplasmic extensions containing clustered fenestrae and sieve plates. (b) SECs treated with 100 ng/ml vascular endothelial growth factor (VEGF) for 24 h show an enormous increase in the number of fenestrae. Long and thin cytoplasmic arms free of fenestrae extend from the nucleus and seem to organize the fenestrae in huge sieve plates. Bar denotes 1 mm. (c) Number of fenestrae per square micrometer. Data are expressed as mean7standard error. The numbers of pores are significantly increased in SECs treated with VEGF at 10 and 100 ng/ml (both Po0.05 compared with control).

470

Caveolin-1 and sinusoidal endothelial fenestrae SECs, caveolin-1 immunostaining was, for the most part, finely punctuate and evenly distributed through the cytoplasm, with only a few aggregates (Fig. 3a). In contrast, SECs cultured with VEGF displayed prominent aggregates of caveolin-1-positive structures in the cytoplasm, especially surrounding the nuclear region (Fig. 3b). Immunoelectron microscopy

Postembedding immunogold EM was used to localize caveolin-1 in the fenestral portion at the ultrastructural level. In cultured control SECs, electron-dense particles showing the presence of caveolin-1 were localized on the plasma membrane of a part of the fenestrae in isolated SECs (Fig. 4a). In isolated SECs cultured with 100 ng/ml VEGF, electron-dense particles of caveolin-1 immunoreactivity were found in abundance on vesiculo-vacuolar organelles, and some particles were also observed on the plasma membrane of fenestrae (Fig. 4b). Western blotting

Next, we investigated the protein levels of caveolin-1 in control SECs and SECs cultured with VEGF by Western blot. Immunoblotting revealed no significant difference in the quantity of caveolin-1 protein expression in control SECs and SECs cultured with VEGF (Fig. 5). HRP-containing fenestrae

Fig. 2. Transmission electron micrograph of sinusoidal endothelial cells (SECs) in primary culture. (a) Control SECs: sinusoidal endothelial fenestrae (SEF) are clearly demonstrated in a tangential ultrathin section of the flat sinusoidal endothelium. (b) SECs treated with 100 ng/ml vascular endothelial growth factor for 24 h: an increase in fenestration is evident. f: SEF. Bar denotes 100 nm. Inset shows a low magnification view. N: nucleus, M: mitochondria, f: SEF.

To obtain evidence for a possible structural relationship between caveolae and fenestrae, we performed immunofluorescence and immunoelectron microscopy to determine whether fenestrae expressed caveolin-1. Confocal microscopy

Caveolin-1 expression was examined by the confocal immunofluorescence method. In control

The transcytotic vesicular pathway in SECs was investigated using HRP tracing. In control SECs, 5 min after HRP labeling, some vesicles containing HRP were observed around the fenestrae and associated with the Golgi complex (Fig. 6a). In SECs cultured with 100 ng/ml VEGF, a strong positive reaction was observed on SEF (Fig. 6b). A cloud of HRP-positive products was observed adjacent to the SEF membrane (Fig. 6b). Discussion

Caveolin is most prominently implicated to be expressed within caveolae. In endothelial cells, caveolin may also be either membrane-associated, such as in caveolae-like vesicles and small Golgi localized vesicles, or cytoplasmically distributed as a soluble pool. Ultrastructurally, there are no typical structures of caveolae in SECs. The new finding of the present study that caveolin-1 is expressed in both VVOs and fenestrae of cultured SECs as demonstrated by immunoelectron microscopy using postembedding 471

Yokomori et al.

Fig. 4. Immunoelectron microscopy by the postembedding method focusing on sinusoidal endothelial fenestrae (SEF). (a) Control SECs: electron-dense particles showing the presence of caveolin-1 were located mainly on the plasma membrane of fenestrae in isolated SECs. (b) SECs cultured with 100 ng/ml vascular endothelial growth factor: strong positive reaction products of caveolin-1 are seen in abundance on vesicles and vacuoles, but a few on fenestrae. f: SEF. Bar denotes 100 nm.

Fig. 3. Confocal fluorescence image of caveolin-1 immunostaining in sinusoidal endothelial cells (SECs) grown with or without vascular endothelial growth factor (VEGF). (a) Control SECs: caveolin-1 immunostaining is finely punctate and evenly distributed throughout the cytoplasm, although a few aggregates are present. (b) SECs cultured with VEGF: many cytoplasmic aggregates of caveolin-1-positive structures are observed, the majority of which are localized at the periphery. A part of caveolin-1 staining appears to be in the nuclear region (arrowhead). Bar denotes 10 mm.

472

method adds to the evidence supporting a relationship between SEF and VVOs. Serial ultrathin section sectioning studies have shown that almost all apparent vesicles in capillary endothelial cells are not free vesicles, but invaginations of surface plasma membrane of the capillary endothelium (17). Shah et al. (18) localized caveolin-1 predominantly in the perinucuclear region in cultured SECs. Recently, our

Caveolin-1 and sinusoidal endothelial fenestrae

Fig. 5. Immunoblotting for the detection of caveolin-1 in sinusoidal endothelial cells (SECs) cultured with or without vascular endothelial growth factor (VEGF) for 24 h. Solubilized extracts were subjected to SDS-PAGE (50 mg protein/lane for caveolin-1), blotted, and then detected with antibodies to caveolin-1. Levels of caveolin-1 protein did not change apparently when grown with or without VEGF.

group reported caveolin-1 in sinusoidal cells by immunofluorescence and immunoelectron microscopy (13). Caveolin-1 was localized from SEFs to the Golgi complex in SECs. Another study described a new structure termed ‘fenestralforming center’ reported to play a role in sinusoidal endothelial fenestra formation (19); suggesting that SEF and caveolae are distinct cellular structures, different in function, origin, and architecture. Macromolecules are extravasated across tumor microvessels or normal venules rendered hyperpermeable by VEGF by three pathways: (1) VVOs that are clusters of cytoplasmic vesicles and vacuoles that span endothelial cytoplasm from the lumen to albumen, (2) trans-endothelial cell pores, and (3) fenestrae (20). In addition, the plasma membrane-invaginated, labyrinth-like structures of the SEF interconnected with one another in racemose clusters of caveolae were identified in the vascular endothelial cell in a state of VEGF-induced increased permeability (16). In the present study, we observed a significant and dose-dependent increase in the number of fenestral pores in VEGF-treated SECs by scanning EM, and also ballooning and interconnection of fenestrae by transmission EM, suggesting that VEGF induces increased fenestration in SECs as in the microvascular endothelium. Various findings of caveolin-1 in vascular endothelial cells have been reported. In bovine

Fig. 6. Sinusoidal endothelial cells (SECs) at 5 min after horseradish peroxidase (HRP) labeling detected by cytochemistry. (a) Control cells: some vesicles (arrowheads), Golgi complex, and vesicle-containing HRP are observed. A little HRP reaction is seen in fenestrae. (b) SECs cultured with 100 ng/ml vascular endothelial growth factor are fused with the abluminal plasma membrane and have released a cloud of HRP-positive products adjacent to the SEF membrane. f: SEF. g: Golgi complex. Bar denotes 200 nm.

endothelial cells, caveolae fuse to form large clusters of fused vesicles believed to be comparable to the VVOs (20). Using immunogold postembedding methods, Esser et al. (21) found that clusters of diaphragmed vesicles were very rarely labeled and all fenestrae were negative for caveolin-1, suggesting that caveolin-1 might be specifically absent in fused clustered vesicles and fenestrae. Feng et al. (22) reported that VEGF 473

Yokomori et al. caused translocation of caveolin-1 and Flk-1/ KDR into the nucleus in the vascular endothelium, suggesting that interactions between these proteins have an important role in the VEGF signal transduction process. A study using immunoblot and immunofluorescence microscopy showed that VEGF dramatically down-regulated the expression of caveolin-1 protein in vascular endothelial cells, which appeared after 5 h of treatment and peaked at 24 h, and that angiogenesis inhibitors selectively blocked the ability of VEGF to induce the down-regulation of caveolin (23). In contrast to the above reports, we found an increase in immunoreactive caveolin-1 aggregates in the cytoplasm of VEGF-treated SECs. Immunoelectron microscopy further demonstrated a tendency of increase in immunoreactive caveolin-1 on VVOs, while there was no marked increase on fenestrae. Our finding was supported by that of Vasile et al. (24), who demonstrated by pre-embedding method the stimulation by VEGF. We propose a hypothesis of this phenomenon as follows. In the presence of VEGF, the formation or assembly of VVOs is induced and this early process involves caveolin-1, resulting in large numbers of caveolin-immunoreactive particles on VVOs. The later stage of development from vesicles/vacuoles to fenestrae, however, does not involve additional caveolin. Eventually, by the redistribution of caveolin on a vastly increased number of fenestrae, caveolin labeling may appear to be reduced. On the other hand, we observed some caveolin-1 reactivity in the nuclear region after VEGF treatment. If we extend the observation period, the caveolin-1 on VVOs might translocate into the nucleus and the expression on VVOs might decrease. However, despite the fact that a tendency of increase in immunoreactive caveolin-1 at both caveolar and VVOs were strongly caveolin-positive upon VVOs and in the cytoplasm, we found no increase of caveolin-1 protein expression in Western blot. Vasile reported the same finding in their study (24). A possible explanation for this is that immunocytochemistry only detects membranebound and not soluble cytoplasmic caveolin, while immunoblotting detects the total quantity of membrane-bound and cytoplasmic caveolin. The formation/assembly of VVOs probably takes place by recruiting the cytoplasmic caveolin; consequently, immunocytochemistry may reveal a great increase in membrane-bound caveolin, while there is only a slight increase in total caveolin as detected by immunoblotting. The effect of VEGF in increasing permeability has been suggested to be the formation of VVOs (25). It is conceivable that VEGF elicits a rapid 474

increase in vascular permeability via mobilization of caveolae, whereas the long-term effect requires the formation of fenestare (24). Our functional assay showed that VEGF treatment increased the uptake of HRP within the SEC cytoplasm, especially in the VVOs, suggesting that vesicular uptake and transport are increased by VEGF. These data indicate that the paracellular pathway is altered by VEGF. Finally, a large increase in the number of VVOs in VEGF-treated cells together with immunolocalization data showing that the HRP-labeled vesicles were positive for caveolin-1 provided strong evidence. Feng et al. (26) localized HRP and caveolin-1 within the same cytoplasmic vesicles and suggested that the VEGF-induced permeability increase is mediated by caveolae. However, Krause et al. (27) reported that VEGF did not stimulate the number and size of sieve plate. In our study, VEGF at 10, 100 ng/ml increased the number of pores significantly compared with the control. This effect might be mimicked by combined addition of VEGF and BSA (26). Our results suggested that VEGF induced fenestration in SECs, accompanied by an increased number of caveolae-like vesicles. Increased caveolin-1 might be associated with vesicle formation but not with fenestration. Increased fenestration may augment hepatic sinusoidal permeability and trans-endothelial transport.

References 1. Wisse E. An electron microscopic study of the fenestrated endothelial lining of rat liver sinusoids. J Ultrastruct Res 1970; 31: 125–50. 2. Wisse E, de Zanger R B, Charles K, Van Der Smissen P, McCuskey R S. The liver sieve: considerations concerning the structure and function of endothelial fenestrae, the sinusoidal wall and the space of Disse. Hepatology 1985; 5: 683–92. 3. Blouin A, Bolender R P, Weibel E R. Distribution of organelles and membranes between hepatocytes and nonhepatocytes in rat liver parenchyma. J Cell Biol 1977; 72: 441–55. 4. Keck P J, Hauser S D, Krivi G, Sango K, Warren T, Feder J, Connolly D T. Vascular permeability factor, an endothelial cell mitogen related to PDGF. Science 1984; 246: 1309–12. 5. Senger D R, Galli S J, Dvorak A M, Perruzzi C A, Harvey V S, Dvorak H F. Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science 1983; 219: 983–5. 6. Roberts W G, Palade G E. Increased microvascular extracellular permeability and endothelial fenestration induced by vascular endothelial growth factor. J Cell Sci 1995; 108: 2369–79. 7. Roberts W G, Palade G E. Neovascular induced by vascular endothelial growth factor is fenestrated. Cancer Res 1997; 57: 765–72.

Caveolin-1 and sinusoidal endothelial fenestrae 8. Rothberg K G, Heuse J E, Donzell W C, Ying Y S, Glenney J R, Anderson R G. Caveolin, protein component of caveolae membrane coats. Cell 1992; 68: 673–82. 9. Kurzchalia T V, Dupree P, Parton R G, Kellner R, Virta H, Lehnert M, Lehnert M, Simons K. Vip 21, a 21-kD membrane protein is an integral component of transgolgi network-derived transport vesicles. J Cell Biol 1992; 118: 1003–14. 10. Dvorak A M, Korn S, Morgan E S, Fox P, Nagy J A, Dvorak H F. The vesiclo-vascuolar organelle (VVO): a distinct endothelial cell structure that provides a Tran cellular pathway for macromolecular extravasation. J Leukocyte Biol 1996; 59: 100–15. 11. Feng D, Nagy J A, Hipp J, Dvorak H F, Dvorak A M. Vesiculo-vacuolar organelles and the regulation of venule permeability to macromolecules by vascular permeability to macromolecules by vascular permeability factor, histamine, and serotonin. J Exp Med 1996; 183: 1981–6. 12. Funyu J, Mochida S, Inao M, Matsui A, Fujiwara K. VEGF can act as vascular permeability factor in the hepatic sinusoids trough upregulation of porosity of endothelial cells. Biochem Biophys Commun 2001; 280: 481–5. 13. Ogi M, Yokomori H, Oda M, Yoshimura K, Nomura M, Ohshima S, Akitra M, Toda K, Ishii H. Distribution and localization of caveolin-1 in sinusoidal cells in rat liver. Med Electron Microsc 2003; 36: 33–40. 14. Braet F, De Zanger R, Sasaoki T, Baekeland M, Janssens P, Smedsrod B, Wisse E. Assessment of a method of isolation, purification and cultivation of rat liver sinusoidal endothelial cells. Lab Invest 1994; 70: 944–52. 15. Berryman M K, Rodewalde R D. An enhanced method for post-embedding immunocytochemical staining which preserves cell membranes. J Histochem Cytochem 1990; 38: 159–70. 16. Sakisaka S, Ng O C, Boyer J L. Tubulovesicular transcytotic pathway in isolated rat hepatocyte couplets in culture. Effect of colchicine and taurocholate. Gastroenterology 1988; 95: 793–804. 17. Bundgaard M, Hageman P, Crone C. The three dimensional organization of plasmalemmal vesicular profiles in the endothelium of rat heart capillaries. Microvas Res 1983; 25: 358–68.

18. Shah V, Cao S, Hendrickson H, Yao J, Katuisic Z S. Regulation of hepatic eNOS by caveolin and calmodulin after bile ligation in rat. Am J Physiol 2001; 280: G1209–16. 19. Braet F, Spector I, De Zanger R, Wisse E. A novel structure involved in the formation of liver endothelial fenestrae revealed by using the actin inhibitor misakinolide. Proc Natl Acad Sci USA 1998; 95: 13635–40. 20. Feng D, Nagy J A, Pyne K, Hammel, Dvorak H F, Dvorak A. Pathways of macromolecular extravasation across microvascular endothelium in response to VPF/ VEGF and other vasoactive mediators. Microcirculation 1999; 6: 23–44. 21. Esser S, Wolburg K, Wolburg H, Breier G, Kurrzchalia T, Risau W. Vascular endothelial growth factor induces endothelial fenestrations in vitro. J Cell Biol 1998; 140: 946–59. 22. Feng Y, Venema V J, Venema R C, Tsai N, Caldwell R B. VEGF induces nuclear translocation of Flk-1/KDR, endothelial nitric oxide synthase, and caveolin-1 in vascular endothelial cells. Biochem Biophys Res Commun 1999; 256(1): 192–7. 23. Liu J, Razani B, Tang S, Terman B I, Ware J A. Angiogenesis activators and inhibitors differentially regulate caveolin-1 expression and caveolae formation in vascular endothelial cell. J Biol Chem 1999; 28: 15781–5. 24. Vasile E, Hong Q, Dvorak H F, Dvorak A M. Caveolae and vesiculo-vacuolar organelles in bovine capillary endothelial cells cultured with VPF/VEGF on floating matrigel-collagen gels. J Histochem Cytochem 1999; 47: 159–67. 25. Korn S, Nagy J A, Dvorak H F, Dvorak A M. Pathways of macromolecular tracer transport across venules and small veins: structural basis for the hyperpermeability of tumor blood vessels. Lab Invest 1992; 67: 596–607. 26. Feng Y, Venema V J, Venema R C, Tsai N, Bebzadian M A, Caldwell R B. VEGF-induced permeability increase is mediated by caveolae. Invest Ophthalmol Vis Sci 1999; 40: 157–67. 27. Krause P, Markus P M, Schwartz P, Unthan-Fechner K, Pestel S, Fandrey J, Probst I. Hepatocytesupported serum-free culture of rat liver sinusoidal endothelial cells. J Hepatol 2000; 32: 718–26.

475