- Email: [email protected]
A highly specific isolation of rat sinusoidal endothelial cells by the immunomagnetic bead method using SE-1 monoclonal antibody
Journal of Hepatology 36 (2002) 725–733 www.elsevier.com/locate/jhep
A highly specific isolation of rat sinusoidal endothelial cells by the immunomagnetic bead method using SE-1 monoclonal antibody Takuo Tokairin, Yuji Nishikawa, Yuko Doi, Hitoshi Watanabe, Toshiaki Yoshioka, Mu Su, Yasufumi Omori, Katsuhiko Enomoto* Department of Pathology, Akita University School of Medicine, 1-1-1 Hondo, Akita 010-8543, Japan
Background/Aims: To develop a specific isolation method of hepatic sinusoidal endothelial cells (SEC), we applied the immunomagnetic method using a monoclonal antibody (SE-1) that recognizes a membranous antigen expressed only in rat SEC. Methods: Cells were isolated by incubating mixed non-parenchymal cells, which were obtained by collagenase digestion of the liver, with SE-1-conjugated superparamagnetic polystyrene beads. The conventional Percoll method was also performed in parallel to compare with the immunomagnetic method. The isolated cells were cultured on glass coverslips coated with type I collagen in the presence of various growth factors for 6 days. Results: Approximately 98% of the isolated cells were positive for SE-1 and the contamination of Kupffer cells or stellate cells was less than 1%. The purity was significantly better than that obtained by the Percoll method. The cultured cells showed typical SEC features, such as sieve plates and uptake of acetylated low-density lipoprotein. Although the cells continuously underwent apoptotic cell death after 2 days, they started robust cell growth after 3 days and were well maintained during the culture period. Conclusions: Our simple and specific isolation method enables us to culture SEC with high purity and should be useful for the biological analysis of SEC. q 2002 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. Keywords: Sinusoidal endothelial cell; Immunomagnetic bead isolation; SE-1 antibody; Tissue culture; Apoptosis; Cell growth
1. Introduction Sinusoidal endothelial cells (SEC) are known to play an important role in various pathophysiological conditions of the liver. They are characterized morphologically by numerous open fenestrations clustered in groups (sieve plates), lack of distinct basal lamina, and abundant vacuolar and vesicular structures [1]. SEC have been regarded to be distinct from vascular endothelial cells of other tissues in their functions, as well as morphology. Besides providing a barrier of the sinusoid, SEC have crucial roles in the balance and distribution of lipids, cholesterol, and vitamin A, and in
Received 20 June 2001; received in revised form 15 January 2002; accepted 5 February 2002 * Corresponding author. Tel.: 181-18-884-6061; fax: 1 81-18-8362601. E-mail address: [email protected] (K. Enomoto).
the uptake of various substances via their specialized endocytotic mechanisms [1]. However, much remains to be elucidated for understanding the mechanisms of SEC functions. For the study of SEC biology and pathology, it is indispensable to obtain purified SEC suitable for in vitro analysis. Percoll centrifugation [2,3] and centrifugal elutriation [4–6] have been used by most investigators for the isolation of SEC. However, the estimated purity of SEC is rather low (80–95%), since these methods are based on the physical properties of SEC such as specific gravity and size, which might overlap with other types of non-parenchymal cells (NPC) of the liver [6]. Furthermore, to date, no methods are able to separate SEC and other types of vascular endothelial cells in the liver. We previously established a monoclonal antibody designated as SE-1, which specifically recognized an antigen expressed in SEC, but not in other types of cells including vascular endothelial cells in the rat [7]. Although the nature
0168-8278/02/$20.00 q 2002 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. PII: S 0168-827 8(02)00048-X
726
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
and function of the SE-1 antigen have not yet been identified, a receptor-like role was postulated on the basis of its cellular localization in vivo (plasma membranes and pinocytotic vesicles) [7]. The expression of the antigen is first detected in the immature SEC in the liver of 15-day-old fetuses and then increases with structural organization and functional maturation of hepatic sinusoids [8]. The antibody has been widely used as one of the most reliable markers for rat SEC by immunocytochemistry or Western blot analysis [9,10]. Recently, magnetic beads with the ability to bind a variety of antibodies or lectins have been applied to positive and negative selection of cells. Vascular endothelial cells from various sources have been successfully isolated by magnetic beads coated with endothelial-specific antibodies [11–15] or lectins [16–19]. However, these molecules are not specific for SEC and not suitable for SEC isolation. Therefore, the strict SEC-specificity and membranous localization of the SE-1 antigen prompted us to apply SE-1 antibody to the immunomagnetic separation of rat SEC. In this study, we show that rat SEC were able to be purely isolated in one step from a mixture of NPC by immunomagnetic beads coated with the antibody. We also describe the characteristics of isolated SEC in culture with reference to cell growth and apoptosis.
blotting. Protein samples were prepared by homogenization of the cells in a lysis buffer (1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 158 mM sodium chloride in 10 mM Tris–HCl buffer (pH 7.5)) containing protease inhibitors. Aliquots of samples were used for measurement of protein concentration using the Bio-Rad protein assay (Bio-Rad, Hercules, CA). Samples (50 mg protein per lane) were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis using 10% polyacrylamide gels. After electrophoresis, proteins were transferred to polyvinylidene difluoride membranes. After blocking the membrane with Tris-buffered saline with Tween 20 (TBST) containing 5% skim milk, the membrane was incubated with the SE-1 antibody (1:300 dilution), then washed with TBST and incubated with anti-mouse or rabbit IgG coupled to horseradish peroxidase. Detection was performed with enhanced chemiluminescence reagents (Amersham–Pharmacia Biotech, Uppsala, Sweden). The membranes were treated with stripping buffer (100 mM 2-mercaptoethanol and 2% sodium dodecyl sulfate in Tris buffer) and reprobed with anti-actin antibody (Clone AC-40; Sigma, St. Louis, MO) for loading control.
2.3. Isolation of SEC by centrifugation in a two-step Percoll gradient
2. Materials and methods
We compared the immunomagnetic method with the Percoll centrifugation method (isopyknic centrifugation in a two-step Percoll gradient (25 and 50%) plus selective adherence to reduce contaminated Kupffer cells) [3], which has been well established and appears to be one of the most widely used for SEC isolation. NPC obtained from one rat was equally divided into two parts and the above two methods were performed for each part in parallel. For the immunomagnetic method, 6 £ 107 beads labeled with 1.5 mg of SE-1 antibody were used. Cell viability was assessed by trypan blue dye exclusion. Yields of cells were estimated by counting of isolated cells using a hemocytometer. Three separate experiments were performed.
2.1. Isolation of liver non-parenchymal cells
2.4. Culture of isolated SEC
Adult male Fischer 344 rats (150–300 g) were used as the source of cells in this study. Livers were digested by collagenase (Wako, Osaka, Japan) using the two-step collagenase perfusion technique. After removing undigested tissues, cells were resuspended in Hanks’ buffer. Resuspended cells were centrifuged at low speed (70 £ g) for 1 min and the NPC-rich supernatant was recovered. After repeating this procedure three more times to eliminate hepatocytes, the supernatant was centrifuged at high speed (620 £ g) for 4 min to pellet a mixed NPC fraction.
The isolated cells were resuspended in Clonetics EBM-2 medium with EGM-2 supplements (Biowhittaker, Walkersville, MD) containing vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), epidermal growth factor (EGF), insulin-like growth factor-1 (IGF-1), hydrocortisone, ascorbic acid, heparin, and 2% fetal bovine serum. The cells were seeded on type I collagen-coated glass coverslips (25 mm in diameter; Asahi Technoglass, Tokyo, Japan) and incubated in a CO2 incubator.
2.2. Procedure for immunomagnetic separation of SEC and determination of optimal conditions
2.5. Immunocytochemistry
We used superparamagnetic polystyrene beads (Dynabeads, 4.5 mm in diameter, CELLection Pan Mouse IgG kit; Dynal, Oslo, Norway), which are coated with a human anti-mouse IgG antibody via a DNA linker, thereby enabling captured cells to be released from the beads by DNase treatment. First, the beads were incubated with SE-1 monoclonal antibodies for 30 min at room temperature. Then the SE-1 antibody-coated beads were incubated with the mixed NPC fraction, which was resuspended in phosphate-buffered saline containing 0.1% bovine serum albumin, for 15 min at 4 8C. Bead-rosetted cells were collected by a magnet (magnetic particle concentrator, Dynal) and then the cells were released by DNase treatment (at room temperature for 20 min), followed by washing with phosphatebuffered saline containing 0.1% bovine serum albumin. To determine the amounts of antibody and beads necessary for the immunomagnetic cell separation, the mixed NPC from one rat was equally divided into six or seven fractions and incubated with the beads treated with various concentrations of SE-1 antibody or with various numbers of the antibody-coated beads. After bead-rosetted cells were removed by the magnet, each NPC fraction was analyzed for SE-1 antigen by Western
For the assessment of phenotype of the cells isolated by the immunomagnetic method and the Percoll method, immunocytochemistry for SE-1, ED2 (anti-rat aDb2 integrin, a marker for Kupffer cells; Serotec, Oxford, UK) [20] and desmin (a marker for stellate cells; Sanbio, Uden, The Netherlands) was performed on attached cells at 2 h after plating. Cells were fixed in cold acetone and stained for each antibody by the avidin–biotin complex method (LSAB kit; Dako, Glostrup, Denmark), followed by counterstaining with hematoxylin. Positively stained cells and total nucleated cells were counted in ten randomly selected high-power fields. Quantification was performed in three separate experiments. To examine the changes in SE-1 expression of the immunomagnetically isolated cells during culture, cells were fixed in cold acetone after 1, 2, 4 and 6 days and immunostained with SE-1 antibody. Desmin immunostaining was also performed on cultured cells to evaluate the proliferation of contaminated stellate cells. Cells were fixed on 1 through 6 days after plating, stained with desmin antibody, and positively stained cells and total nucleated cells were counted in ten randomly selected high-power fields. Quantification was performed in five separate experiments.
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
2.6. Scanning electron microscopy
3. Results
To examine the presence of fenestrations in the cytoplasm, scanning electron microscopical observation was performed. Cultured cells were fixed with 2.5% glutaraldehyde and postfixed with 1% osmium tetroxide. Samples were critical point dried, sputter-coated with gold, and examined with a JSM-T200 scanning electron microscope (JEOL, Tokyo, Japan).
3.1. Optimization of the immunomagnetic method
727
To test the ability of the isolated cells to uptake acetylated LDL, which is known to be highly specific to endothelial cells, including SEC, cultured cells were incubated with a medium containing 10 mg/ml of 1,1 0 -dioctadecyl-3,3,3 0 ,3 0 -tetramethylindocarbocyanine perchlorate (DiI)-labeled acetylated-LDL (Molecular Probes, Eugene, OR) for 4 h at 37 8C and examined under a fluorescence microscope (Olympus, Tokyo, Japan).
To determine optimal amounts of the antibody and beads for the SEC isolation, Western blot analyses were performed for residual SE-1antigen in the NPC incubated with the beads treated with various concentrations of SE-1 antibody or with various amounts of the antibody-coated beads. The results demonstrated that 1–2 mg of antibody was sufficient for labeling of 4 £ 107 beads (100 ml of original bead suspension supplied by the company) and that 1:6 £ 108 beads (2 mg of antibody/4 £ 107 beads) were enough to deplete almost all SE-1-positive cells from a single adult liver (Fig. 1).
2.8. Assays for cell growth and apoptotic cell death
3.2. Characteristics of immunomagnetically isolated cells
A Hoechst 33258-based assay for DNA was employed for the determination of cell number during culture [21]. DNA synthesis of cultured cells was examined by bromodeoxyuridine (BrdU) labeling. SEC were incubated with 10 mM BrdU for 1 h and then fixed with methanol. After denaturation of DNA by 2 N HCl, incorporated BrdU was detected immunocytochemically by anti-BrdU antibodies (Roche, Mannheim, Germany). Apoptotic cell death during culture was detected by the terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) method (Apoptag; Intergen, Purchase, NY) according to the manufacturer’s protocol.
Incubation with SE-1 antibody-coated magnetic beads with the NPC fraction resulted in the formation of beadrosetted cells (Fig. 2a). After plating they attached to collagen-coated coverslips within 1 h and showed polygonal, net-like cytoplasm (Fig. 2b). Immunocytochemically, almost all of these cells were positive for the SE-1 antigen (Fig. 2c). Isolated SE-1-positive cells displayed typical features of SEC, such as clusters of fenestrations (sieve plates) (Fig. 3a) and uptake of DiI-acetylated LDL (Fig. 4a).
2.7. Uptake of acetylated low-density lipoprotein (acetylated LDL)
Fig. 1. Depletion of the SE-1 antigen from the non-parenchymal cell fractions after the immunomagnetic separation. Protein samples were analyzed for SE-1 antigen and actin (loading control) by Western blotting. (Upper panel) Amount of antibody was changed, while the bead number was constant (1:2 £ 108 beads/liver). (Lower panel) Number of beads was changed, while the amount of antibody was constant (2 mg/4 £ 107 beads).
728
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
Fig. 2. Isolated cells by immunomagnetic beads coated with monoclonal antibody SE-1. (a) Bead-rosetted cells after incubation of a mixed NPC fraction with the immunomagnetic beads. (b) Phase-contrast microscopy of the 6-h-cultured cells. (c) SE-1 immunocytochemistry of the 6-h-cultured cells. Bars: a, 10 mm; b,c, 20 mm.
3.3. Comparison of the immunomagnetic method with the Percoll method We examined cell viability, purity, and yield of the immunomagnetic method and compared them with those
of the method using a two-step Percoll gradient with selective substrate adherence, which is widely applied to isolate rat SEC [3]. The cells released by DNase treatment were apparently intact and viability was constantly more than 90%, similar to that of the Percoll method (Table 1). SE-
Fig. 3. Scanning electron microscopy of the cultured SEC showing fenestrations: (a) 1 day after culture; (b) 4 days after culture. Bar: 1 mm.
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
729
Fig. 4. Uptake of DiI-labeled acetylated LDL by the cultured SEC: (a) 6 h after culture; (b) 6 days after culture. Bar: 20 mm.
1-positive SEC, ED2-positive Kupffer cells, and desminpositive stellate cells were counted on immunocytochemical preparations of the isolated cells after 2 h (just after attachment). As shown in Table 1, the percentage of SE-1-positive cells was significantly higher, while the percentage of ED2positive cells was significantly lower in the immunomagnetic method as compared with the Percoll method. The initial contamination of desmin-positive cells was less than 1% in both methods. The cell yields by the immunomagnetic method were less than those by the Percoll method.
and 5e–h). The cells maintained sieve plates (Fig. 3b) and the capacity of DiI-acetylated LDL uptake (Fig. 4b) for at least 4 days. Although desmin-positive stellate cells were scarcely detected in the initial preparation, colonies of activated stellate cells appeared after 3 or 4 days in some experiments (Fig. 5c,d). The percentages (the mean ^ SEM, n ¼ 5) of desmin-positive cells in the total cells were 0.4 ^ 0.1, 1.6 ^ 0.4, 5.3 ^ 1.1, 6.3 ^ 2.3, 7.2 ^ 2.5, and 8.9 ^ 2.4 at 1, 2, 3, 4, 5, 6 days after culture, respectively.
3.4. Morphological and phenotypical characteristics of the isolated cells in vitro
The number of cells, which was estimated by DNA fluorometric assay, sharply decreased during the first 2 days (Fig. 6). However, the decline in cell number was stopped after 3 days and then the number appeared to increase slightly (Fig. 6). Although the TUNEL method detected only a few apoptotic cells just after attachment, there was an increase in apoptotic cell death after 1 day and the number reached a plateau after 3 days (Fig. 7). BrdU-labeled nuclei, as well as mitotic nuclei, sharply increased after 3 days when a slight increase in cell number was seen (Fig. 8).
Initially, the cultured SEC were polygonal and showed a cobblestone, sheet-like appearance (Fig. 5a,b). Then, the cells gradually became spindle-like and tended to be arranged in groups of varying sizes (Fig. 5c,d). SE-1 was positive in these cells throughout the culture period without apparent changes in its intensity and localization (Figs. 2c Table 1 Comparison of the immunomagnetic separation and the two-step Percoll gradient method a Immunomagnetic separation Percoll method Viability (%) 94.0 ^ 1.3 SE-1-positive cells (%) 97.6 ^ 0.9* ED2-positive cells (%) 0.9 ^ 0.1* Desmin-positive cells (%) 0.4 ^ 0.2 Yield ( £ 10 6 cells/liver) 10.7 ^ 0.5*
93.7 ^ 1.4 92.0 ^ 0.8 2.3 ^ 0.5 0.7 ^ 0.2 24.3 ^ 0.5
a Each value represents the mean ^ SEM for three separate experiments. *Significantly differenct as compared with the Percoll method (P , 0:01, Student’s t-test).
3.5. Changes in cell number, apoptotic cell death, and cell growth during culture
4. Discussion To study SEC functions in normal and pathological settings, it is important to prepare highly purified and viable SEC. In the present study, we have shown that viable SEC of 97.6% purity can be obtained by the immunomagnetic method using the monoclonal SE-1 antibody. SEC have been isolated by various methods including elutriation after pronase digestion [4–6] or Percoll centrifugation
730
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
Fig. 5. Time course of changes in morphology and SE-1 antigen expression in cultured sinusoidal endothelial cells isolated by the immunomagnetic beads. (a–d) Phase-contrast microscopy. (e–h) SE-1 immunocytochemistry. (a,e) One day; (b,f) 2 days; (c,g) 4 days; (d,h) 6 days. Bars: phase-contrast, 100 mm; immunocytochemistry, 20 mm.
[2,3]. SEC purity by these methods has been reported to be 80–95%. Thus, the enrichment of SEC acquired by the immunomagnetic method described in this paper appears to be the best among the previously reported methods for SEC isolation. Furthermore, the method is simple and repro-
Fig. 6. The changes in cell number during culture of sinusoidal endothelial cells. The number of cells were estimated by DNA fluorometric assay and represented as absolute fluorescence units. Data are shown as the mean ^ SEM of three separate experiments performed in duplicate.
ducible, without the need for cumbersome operation of an elutriator or pronase digestion that might hamper cell functions. Since SEC purity is already sufficiently high in the cell after the magnetic separation, it is not necessary to further enrich SEC in vitro by selective attachment and further culturing as in the case of the Percoll method. The lower cell yield of our method is considered to be mainly due to an incomplete release of cells from the beads after DNase treatment. Various kinds of vascular endothelial cells have been isolated by magnetic beads coated with antibodies (antithrombomodulin antibody (QB End/40) [11,14], anti-platelet-endothelial cell adhesion molecule 1 (PECAM-1) antibody [13,15], and S-Endo1 antibody [12]) or lectins (Ulex europaeus agglutinin-1 (UEA-1) [16–18] and Evonymus europaeus agglutinin (EEA) [19]). However, these antibodies and lectins are not necessarily suitable for the specific magnetic bead isolation of SEC. Since both vascular endothelial cells and SEC express thrombomodulin, its antibody is not able to distinguish these endothelial cells. SEC normally do not express PECAM-1 [22] and do not bind to UEA-1 [23]. It is not yet clear whether S-Endo1 recognizes SEC, while it has been reported to be highly specific to bone marrow endothelial cells [24]. An attempt has been made to isolate liver endothelial cells from Rhesus monkeys by magnetic beads coated with EEA lectin [19]. Since EEA also could bind to vascular endothelial cells in general, the isolated cells may be contaminated by non-SEC vascular
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
731
Fig. 7. Detection of apoptosis of cultured sinusoidal endothelial cells by terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL). (a) TUNEL-positive apoptotic cells (arrows) at 2 days after culture. Bar: 20 mm. (b) Quantitative analysis for time course of the frequency of apoptotic cells. Percent of TUNEL-positive cells were counted under a light microscope in ten random high-power fields (using a £40 objective lens) in each preparation. Data are shown as the mean ^ SEM of three separate experiments.
endothelial cells present in the liver. To the best of our knowledge, this is the first report of immunomagnetic separation of SEC. In our experiments, the contamination of stellate cells,
which was recognized by desmin immunostaining, was small (0.4%) in the initial cell fraction (Table 1). Then, the contaminated stellate cells were activated and proliferated to form colonies after 3 or 4 days, although the percen-
Fig. 8. Proliferation of sinusoidal endothelial cells in vitro. (a) BrdU immunocytochemistry of cultured cells (a mitotic figure is indicated by an arrow). Bar: 20 mm. (b) Time course of BrdU labeling index (B) and mitotic indices (A). BrdU-positive and mitotic cells were counted under a light microscope in three random medium-power fields (using a £20 objective lens) in each preparation. Data are shown as the mean ^ SEM of three separate experiments performed in duplicate.
732
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
tage of stellate cell did not exceed 9% even after 6 days. However, since these contaminated stellate cells might hamper long-term cultures of SEC, it should be important to further reduce the stellate cell contamination. For this purpose, Percoll density gradient centrifugation [25], which can separate stellate cells from SEC and Kupffer cells in the mixed NPC fraction, may be effective prior to the immunomagnetic separation. Immunomagnetically isolated SEC could be maintained in vitro on type I collagen-coated coverslips at least for 6 days in the presence of growth factors, including VEGF and bFGF. VEGF has been shown to be both a growth and survival factor for SEC [26,27]. bFGF has been demonstrated to be a survival factor for vascular endothelial cells [28]. As shown in Fig. 6, the cell number decreased during the first two days, when apoptotic cell death occurred without cell proliferation, followed by the phase in which apoptosis was almost counterbalanced by the concomitant cell proliferation. The time course of changes in cell number and DNA synthesis of SEC cocultured with hepatocytes is similar to that reported recently by Krause et al. [29]. SEC underwent cell proliferation and apoptosis simultaneously during culture. Both vascular endothelial proliferation [30,31] and apoptosis [32,33] are known to be regulated by the mitogen-activated protein (MAP) kinase kinase (MEK)/extracellular signal-regulated kinase (ERK) pathway. Integrin avb3 is necessary for bFGF to induce sustained tyrosine phosphorylation of ERK for promoting angiogenesis [34]. On the other hand, vascular endothelial cell apoptosis is shown to be related with calpain-induced cleavage of the integrin b3 subunit, which is preceded by protein tyrosine dephosphorylation [35]. Furthermore, recent experiments showed that SEC apoptosis induced by cold ischemia in vivo is calpain-dependent [36]. Our preliminary results also indicated that SEC proliferation was prevented by a MEK inhibitor, PD098059. Further analysis of the signaling pathway for SEC proliferation and apoptosis is now in progress. In conclusion, using the monoclonal SE-1 antibody, we have succeeded in immunomagnetic separation of rat SEC, which is highly purified, viable, and ready for any types of culture. Our method may provide a useful tool for the biological analysis of SEC.
Acknowledgements The authors greatly appreciate Drs. H. Masuda, H. Senoo, and H. Kotanagi for critically reading the manuscript. The authors are also grateful to Mr. N. Kojima for advice in performing confocal laser scanning microscopy and to Dr. K. Kawamura and Mr. Y. Sasaki for help in preparing samples for scanning electron microscopy. This study was supported in part by grants from the Ministry of Education, Science, Sports, and Culture of Japan (09670214).
References [1] Wisse E, Braet F, Luo D, Vermijlen D, Eddouks M, Konstandoulaki M, et al. Endothelial cells of the hepatic sinusoids: a review. In: Tanikawa K, Ueno T, et al., editors. Liver diseases and hepatic sinusoidal cells, Tokyo: Springer, 1999. pp. 17–55. [2] Smedsrød B, Pertoft H, Eggertsen G, Sundsto¨ m C. Functional and morphological characterization of cultures of Kupffer cells and liver endothelial cells prepared by means of density separation in Percoll, and selective substrate adherence. Cell Tissue Res 1985;241:639– 649. [3] Braet F, De Zanger R, Sasaoki T, Baekeland M, Janssens P, Smedsrød B, et al. Assessment of a method of isolation, purification, and cultivation of rat liver sinusoidal endothelial cells. Lab Invest 1994;70:944–952. [4] Knook DL, Sleyster ECH. Separation of Kupffer and endothelial cells of the rat liver by centrifugal elutriation. Exp Cell Res 1976;99:444– 449. [5] de Leeuw AM, Barelds RJ, de Zanger R, Knook DL. Primary cultures of endothelial cells of the rat liver. Cell Tissue Res 1982;223:201–215. [6] Irving MG, Roll FJ, Huang S, Bissell DM. Characterization and culture of sinusoidal endothelium from normal rat liver: lipoprotein uptake and collagen phenotype. Gastroenterology 1984;87:1233– 1247. [7] Ohmura T, Enomoto K, Satoh H, Sawada N, Mori M. Establishment of a novel monoclonal antibody, SE-1, which specifically reacts with rat hepatic sinusoidal endothelial cells. J Histochem Cytochem 1993;41:1253–1257. [8] Morita M, Qun W, Watanabe H, Doi Y, Mori M, Enomoto K. Analysis of the sinusoidal endothelial cell organization during the developmental stage of the fetal rat liver using a sinusoidal endothelial cell specific antibody, SE-1. Cell Struct Funct 1998;23:341–348. [9] Ishiguro S, Arii S, Monden K, Adachi Y, Funaki N, Higashitsuji H, Fujita S-I, et al. Identification of the thromboxane A2 receptor in hepatic sinusoidal endothelial cells and its role in endotoxin-induced liver injury in rats. Hepatology 1994;20:1281–1286. [10] Stolz DB, Ross MA, Salem HM, Mars WM, Michalopoulos GK, Enomoto K. Cationic colloidal silica membrane perturbation as a means of examining changes at the sinusoidal surface during liver regeneration. Am J Pathol 1999;155:1487–1498. [11] Drake BL, Loke YW. Isolation of endothelial cells from human first trimester decidua using immunomagnetic beads. Hum Reprod 1991;6:1156–1159. [12] George F, Brisson C, Poncelet P, Laurent JC, Massot O, Arnoux D, et al. Rapid isolation of human endothelial cells from whole blood using S-Endo1 monoclonal antibody coupled to immuno-magnetic beads: demonstration of endothelial injury after angioplasty. Thromb Haemost 1992;67:147–153. [13] Hewett PW, Murray JC. Immunomagnetic purification of human microvessel endothelial cells using Dynabeads coated with monoclonal antibodies to PECAM-1. Eur J Cell Biol 1993;62:451–454. [14] Leach L, Bhasin Y, Clark P, Firth JA. Isolation of endothelial cells from human term placental villi using immunomagnetic beads. Placenta 1994;15:355–364. [15] Hull MA, Hewett PW, Brough JL, Hawkey CJ. Isolation and culture of human gastric endothelial cells. Gastroenterology 1996;111:1230– 1240. [16] Jackson CJ, Garbett PK, Nissen B, Schrieber L. Binding of human endothelium to Ulex europaeus 1-coated Dynabeads: application to the isolation of microvascular endothelium. J Cell Sci 1990:257–262. [17] Abbot SE, Kaul A, Stevens CR, Blake DR. Isolation and culture of synovial microvascular endothelial cells: characterization and assessment of adhesion molecule expression. Arthritis Rheum 1992;35:401– 406. [18] Masek LC, Sweetenham JW. Isolation and culture of endothelial cells from human bone marrow. Br J Haematol 1994;88:855–865.
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733 [19] Gomez DE, Hartzler JL, Corbitt RH, Nason AM, Thorgeirsson UP. Immunomagnetic separation as a final purification step of liver endothelial cells. In Vitro Cell Dev Biol 1993;29A:451–455. [20] Dijkstra CD, Do¨ pp EA, Joling P, Kraal G. The heterogeneity of mononuclear phagocytes in lymphoid organs: distinct macrophage subpopulations in the rat recognized by monoclonal antibodies ED1, ED2 and ED3. Immunology 1985;54:589–599. [21] Rago R, Mitchen J, Wilding G. DNA fluorometric assay in 96-well tissue culture plates using Hoechst 33258 after cell lysis by freezing in distilled water. Anal Biochem 1990;191:31–34. [22] Scoazec JY, Racine L, Couvelard A, Flejou JF, Feldmann G. Endothelial cell heterogeneity in the normal human liver acinus: in situ immunohistochemical demonstration. Liver 1994;14:113–123. [23] Terada T, Nakanuma Y. Expression of ABH blood group antigens, Ulex europaeus agglutinin I, and type IV collagen in the sinusoids of hepatocellular carcinoma. Arch Pathol Lab Med 1991;115:50–55. [24] Schweitzer CM, van der Schoot CE, Drager AM, van der Valk P, Zevenbergen A, Hooibrink B, et al. Isolation and culture of human bone marrow endothelial cells. Exp Hematol 1995;23:41–48. [25] Blomhoff R, Rasmussen M, Nilsson A, Norum KR, Berg T, Blaner WS, et al. Hepatic retinol metabolism. Distribution of retinoids, enzymes, and binding proteins in isolated rat liver cells. J Biol Chem 1985;260:13560–13565. [26] Yamane A, Seetharam L, Yamaguchi S, Gotoh N, Takahashi T, Neufeld G, et al. A new communication system between hepatocytes and sinusoidal endothelial cells in liver through vascular endothelial growth factor and Flt tyrosine kinase receptor family (Flt-1 and KDR/ Flk-1). Oncogene 1994;9:2683–2690. [27] Zhang Y-J, Ikejima K, Honda H, Kitamura T, Takei Y, Sato N. Glycine prevents apoptosis of rat sinusoidal endothelial cells caused
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
733
by deprivation of vascular endothelial growth factor. Hepatology 2000;32:542–546. Araki S, Shimada Y, Kaji K, Hayashi H. Apoptosis of vascular endothelial cells by fibroblast growth factor deprivation. Biochem Biophys Res Commun 1990;168:1194–1200. Krause P, Markus PM, Schwartz P, Unthan-Fechner K, Pestel S, Fandrey J, Probst I. Hepatocyte-supported serum-free culture of rat liver sinusoidal endothelial cells. J Hepatol 2000;32:718–726. Vin˜ als F, Pouysse´ gur J. Confluence of vascular endothelial cells induces cell cycle exit by inhibiting p42/p44 mitogen-activated protein kinase activity. Mol Cell Biol 1999;19:2763–2772. Suzuki E, Nagata D, Yoshizumi M, Kakoki M, Goto A, Omata M, et al. Reentry into the cell cycle of contact-inhibited vascular endothelial cells by a phosphatase inhibitor. J Biol Chem 2000;275:3637–3644. Yang C, Chang J, Gorospe M, Passaniti A. Protein tyrosine phosphatase regulation of endothelial cell apoptosis and differentiation. Cell Growth Differ 1996;7:161–171. Kuzuya M, Satake S, Ramos MA, Kanda S, Koike T, Yoshino K, et al. Induction of apoptotic cell death in vascular endothelial cells cultured in three-dimensional collagen lattice. Exp Cell Res 1999;248:498– 508. Eliceiri BP, Klemke R, Stro¨ mblad S, Cheresh DA. Integrin avb3 requirement for sustained mitogen-activated protein kinase activity during angiogenesis. J Cell Biol 1998;140:1255–1263. Meredith JJ, Mu Z, Saido T, Du X. Cleavage of the cytoplasmic domain of the integrin b3 subunit during endothelial apoptosis. J Biol Chem 1998;273:19525–19531. Sindram D, Kohli V, Madden JF, Clavien PA. Calpain inhibition prevents sinusoidal endothelial cell apoptosis in the cold ischemic rat liver. Transplantation 1999;68:136–140.
A highly specific isolation of rat sinusoidal endothelial cells by the immunomagnetic bead method using SE-1 monoclonal antibody Takuo Tokairin, Yuji Nishikawa, Yuko Doi, Hitoshi Watanabe, Toshiaki Yoshioka, Mu Su, Yasufumi Omori, Katsuhiko Enomoto* Department of Pathology, Akita University School of Medicine, 1-1-1 Hondo, Akita 010-8543, Japan
Background/Aims: To develop a specific isolation method of hepatic sinusoidal endothelial cells (SEC), we applied the immunomagnetic method using a monoclonal antibody (SE-1) that recognizes a membranous antigen expressed only in rat SEC. Methods: Cells were isolated by incubating mixed non-parenchymal cells, which were obtained by collagenase digestion of the liver, with SE-1-conjugated superparamagnetic polystyrene beads. The conventional Percoll method was also performed in parallel to compare with the immunomagnetic method. The isolated cells were cultured on glass coverslips coated with type I collagen in the presence of various growth factors for 6 days. Results: Approximately 98% of the isolated cells were positive for SE-1 and the contamination of Kupffer cells or stellate cells was less than 1%. The purity was significantly better than that obtained by the Percoll method. The cultured cells showed typical SEC features, such as sieve plates and uptake of acetylated low-density lipoprotein. Although the cells continuously underwent apoptotic cell death after 2 days, they started robust cell growth after 3 days and were well maintained during the culture period. Conclusions: Our simple and specific isolation method enables us to culture SEC with high purity and should be useful for the biological analysis of SEC. q 2002 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. Keywords: Sinusoidal endothelial cell; Immunomagnetic bead isolation; SE-1 antibody; Tissue culture; Apoptosis; Cell growth
1. Introduction Sinusoidal endothelial cells (SEC) are known to play an important role in various pathophysiological conditions of the liver. They are characterized morphologically by numerous open fenestrations clustered in groups (sieve plates), lack of distinct basal lamina, and abundant vacuolar and vesicular structures [1]. SEC have been regarded to be distinct from vascular endothelial cells of other tissues in their functions, as well as morphology. Besides providing a barrier of the sinusoid, SEC have crucial roles in the balance and distribution of lipids, cholesterol, and vitamin A, and in
Received 20 June 2001; received in revised form 15 January 2002; accepted 5 February 2002 * Corresponding author. Tel.: 181-18-884-6061; fax: 1 81-18-8362601. E-mail address: [email protected] (K. Enomoto).
the uptake of various substances via their specialized endocytotic mechanisms [1]. However, much remains to be elucidated for understanding the mechanisms of SEC functions. For the study of SEC biology and pathology, it is indispensable to obtain purified SEC suitable for in vitro analysis. Percoll centrifugation [2,3] and centrifugal elutriation [4–6] have been used by most investigators for the isolation of SEC. However, the estimated purity of SEC is rather low (80–95%), since these methods are based on the physical properties of SEC such as specific gravity and size, which might overlap with other types of non-parenchymal cells (NPC) of the liver [6]. Furthermore, to date, no methods are able to separate SEC and other types of vascular endothelial cells in the liver. We previously established a monoclonal antibody designated as SE-1, which specifically recognized an antigen expressed in SEC, but not in other types of cells including vascular endothelial cells in the rat [7]. Although the nature
0168-8278/02/$20.00 q 2002 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. PII: S 0168-827 8(02)00048-X
726
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
and function of the SE-1 antigen have not yet been identified, a receptor-like role was postulated on the basis of its cellular localization in vivo (plasma membranes and pinocytotic vesicles) [7]. The expression of the antigen is first detected in the immature SEC in the liver of 15-day-old fetuses and then increases with structural organization and functional maturation of hepatic sinusoids [8]. The antibody has been widely used as one of the most reliable markers for rat SEC by immunocytochemistry or Western blot analysis [9,10]. Recently, magnetic beads with the ability to bind a variety of antibodies or lectins have been applied to positive and negative selection of cells. Vascular endothelial cells from various sources have been successfully isolated by magnetic beads coated with endothelial-specific antibodies [11–15] or lectins [16–19]. However, these molecules are not specific for SEC and not suitable for SEC isolation. Therefore, the strict SEC-specificity and membranous localization of the SE-1 antigen prompted us to apply SE-1 antibody to the immunomagnetic separation of rat SEC. In this study, we show that rat SEC were able to be purely isolated in one step from a mixture of NPC by immunomagnetic beads coated with the antibody. We also describe the characteristics of isolated SEC in culture with reference to cell growth and apoptosis.
blotting. Protein samples were prepared by homogenization of the cells in a lysis buffer (1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 158 mM sodium chloride in 10 mM Tris–HCl buffer (pH 7.5)) containing protease inhibitors. Aliquots of samples were used for measurement of protein concentration using the Bio-Rad protein assay (Bio-Rad, Hercules, CA). Samples (50 mg protein per lane) were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis using 10% polyacrylamide gels. After electrophoresis, proteins were transferred to polyvinylidene difluoride membranes. After blocking the membrane with Tris-buffered saline with Tween 20 (TBST) containing 5% skim milk, the membrane was incubated with the SE-1 antibody (1:300 dilution), then washed with TBST and incubated with anti-mouse or rabbit IgG coupled to horseradish peroxidase. Detection was performed with enhanced chemiluminescence reagents (Amersham–Pharmacia Biotech, Uppsala, Sweden). The membranes were treated with stripping buffer (100 mM 2-mercaptoethanol and 2% sodium dodecyl sulfate in Tris buffer) and reprobed with anti-actin antibody (Clone AC-40; Sigma, St. Louis, MO) for loading control.
2.3. Isolation of SEC by centrifugation in a two-step Percoll gradient
2. Materials and methods
We compared the immunomagnetic method with the Percoll centrifugation method (isopyknic centrifugation in a two-step Percoll gradient (25 and 50%) plus selective adherence to reduce contaminated Kupffer cells) [3], which has been well established and appears to be one of the most widely used for SEC isolation. NPC obtained from one rat was equally divided into two parts and the above two methods were performed for each part in parallel. For the immunomagnetic method, 6 £ 107 beads labeled with 1.5 mg of SE-1 antibody were used. Cell viability was assessed by trypan blue dye exclusion. Yields of cells were estimated by counting of isolated cells using a hemocytometer. Three separate experiments were performed.
2.1. Isolation of liver non-parenchymal cells
2.4. Culture of isolated SEC
Adult male Fischer 344 rats (150–300 g) were used as the source of cells in this study. Livers were digested by collagenase (Wako, Osaka, Japan) using the two-step collagenase perfusion technique. After removing undigested tissues, cells were resuspended in Hanks’ buffer. Resuspended cells were centrifuged at low speed (70 £ g) for 1 min and the NPC-rich supernatant was recovered. After repeating this procedure three more times to eliminate hepatocytes, the supernatant was centrifuged at high speed (620 £ g) for 4 min to pellet a mixed NPC fraction.
The isolated cells were resuspended in Clonetics EBM-2 medium with EGM-2 supplements (Biowhittaker, Walkersville, MD) containing vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), epidermal growth factor (EGF), insulin-like growth factor-1 (IGF-1), hydrocortisone, ascorbic acid, heparin, and 2% fetal bovine serum. The cells were seeded on type I collagen-coated glass coverslips (25 mm in diameter; Asahi Technoglass, Tokyo, Japan) and incubated in a CO2 incubator.
2.2. Procedure for immunomagnetic separation of SEC and determination of optimal conditions
2.5. Immunocytochemistry
We used superparamagnetic polystyrene beads (Dynabeads, 4.5 mm in diameter, CELLection Pan Mouse IgG kit; Dynal, Oslo, Norway), which are coated with a human anti-mouse IgG antibody via a DNA linker, thereby enabling captured cells to be released from the beads by DNase treatment. First, the beads were incubated with SE-1 monoclonal antibodies for 30 min at room temperature. Then the SE-1 antibody-coated beads were incubated with the mixed NPC fraction, which was resuspended in phosphate-buffered saline containing 0.1% bovine serum albumin, for 15 min at 4 8C. Bead-rosetted cells were collected by a magnet (magnetic particle concentrator, Dynal) and then the cells were released by DNase treatment (at room temperature for 20 min), followed by washing with phosphatebuffered saline containing 0.1% bovine serum albumin. To determine the amounts of antibody and beads necessary for the immunomagnetic cell separation, the mixed NPC from one rat was equally divided into six or seven fractions and incubated with the beads treated with various concentrations of SE-1 antibody or with various numbers of the antibody-coated beads. After bead-rosetted cells were removed by the magnet, each NPC fraction was analyzed for SE-1 antigen by Western
For the assessment of phenotype of the cells isolated by the immunomagnetic method and the Percoll method, immunocytochemistry for SE-1, ED2 (anti-rat aDb2 integrin, a marker for Kupffer cells; Serotec, Oxford, UK) [20] and desmin (a marker for stellate cells; Sanbio, Uden, The Netherlands) was performed on attached cells at 2 h after plating. Cells were fixed in cold acetone and stained for each antibody by the avidin–biotin complex method (LSAB kit; Dako, Glostrup, Denmark), followed by counterstaining with hematoxylin. Positively stained cells and total nucleated cells were counted in ten randomly selected high-power fields. Quantification was performed in three separate experiments. To examine the changes in SE-1 expression of the immunomagnetically isolated cells during culture, cells were fixed in cold acetone after 1, 2, 4 and 6 days and immunostained with SE-1 antibody. Desmin immunostaining was also performed on cultured cells to evaluate the proliferation of contaminated stellate cells. Cells were fixed on 1 through 6 days after plating, stained with desmin antibody, and positively stained cells and total nucleated cells were counted in ten randomly selected high-power fields. Quantification was performed in five separate experiments.
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
2.6. Scanning electron microscopy
3. Results
To examine the presence of fenestrations in the cytoplasm, scanning electron microscopical observation was performed. Cultured cells were fixed with 2.5% glutaraldehyde and postfixed with 1% osmium tetroxide. Samples were critical point dried, sputter-coated with gold, and examined with a JSM-T200 scanning electron microscope (JEOL, Tokyo, Japan).
3.1. Optimization of the immunomagnetic method
727
To test the ability of the isolated cells to uptake acetylated LDL, which is known to be highly specific to endothelial cells, including SEC, cultured cells were incubated with a medium containing 10 mg/ml of 1,1 0 -dioctadecyl-3,3,3 0 ,3 0 -tetramethylindocarbocyanine perchlorate (DiI)-labeled acetylated-LDL (Molecular Probes, Eugene, OR) for 4 h at 37 8C and examined under a fluorescence microscope (Olympus, Tokyo, Japan).
To determine optimal amounts of the antibody and beads for the SEC isolation, Western blot analyses were performed for residual SE-1antigen in the NPC incubated with the beads treated with various concentrations of SE-1 antibody or with various amounts of the antibody-coated beads. The results demonstrated that 1–2 mg of antibody was sufficient for labeling of 4 £ 107 beads (100 ml of original bead suspension supplied by the company) and that 1:6 £ 108 beads (2 mg of antibody/4 £ 107 beads) were enough to deplete almost all SE-1-positive cells from a single adult liver (Fig. 1).
2.8. Assays for cell growth and apoptotic cell death
3.2. Characteristics of immunomagnetically isolated cells
A Hoechst 33258-based assay for DNA was employed for the determination of cell number during culture [21]. DNA synthesis of cultured cells was examined by bromodeoxyuridine (BrdU) labeling. SEC were incubated with 10 mM BrdU for 1 h and then fixed with methanol. After denaturation of DNA by 2 N HCl, incorporated BrdU was detected immunocytochemically by anti-BrdU antibodies (Roche, Mannheim, Germany). Apoptotic cell death during culture was detected by the terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) method (Apoptag; Intergen, Purchase, NY) according to the manufacturer’s protocol.
Incubation with SE-1 antibody-coated magnetic beads with the NPC fraction resulted in the formation of beadrosetted cells (Fig. 2a). After plating they attached to collagen-coated coverslips within 1 h and showed polygonal, net-like cytoplasm (Fig. 2b). Immunocytochemically, almost all of these cells were positive for the SE-1 antigen (Fig. 2c). Isolated SE-1-positive cells displayed typical features of SEC, such as clusters of fenestrations (sieve plates) (Fig. 3a) and uptake of DiI-acetylated LDL (Fig. 4a).
2.7. Uptake of acetylated low-density lipoprotein (acetylated LDL)
Fig. 1. Depletion of the SE-1 antigen from the non-parenchymal cell fractions after the immunomagnetic separation. Protein samples were analyzed for SE-1 antigen and actin (loading control) by Western blotting. (Upper panel) Amount of antibody was changed, while the bead number was constant (1:2 £ 108 beads/liver). (Lower panel) Number of beads was changed, while the amount of antibody was constant (2 mg/4 £ 107 beads).
728
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
Fig. 2. Isolated cells by immunomagnetic beads coated with monoclonal antibody SE-1. (a) Bead-rosetted cells after incubation of a mixed NPC fraction with the immunomagnetic beads. (b) Phase-contrast microscopy of the 6-h-cultured cells. (c) SE-1 immunocytochemistry of the 6-h-cultured cells. Bars: a, 10 mm; b,c, 20 mm.
3.3. Comparison of the immunomagnetic method with the Percoll method We examined cell viability, purity, and yield of the immunomagnetic method and compared them with those
of the method using a two-step Percoll gradient with selective substrate adherence, which is widely applied to isolate rat SEC [3]. The cells released by DNase treatment were apparently intact and viability was constantly more than 90%, similar to that of the Percoll method (Table 1). SE-
Fig. 3. Scanning electron microscopy of the cultured SEC showing fenestrations: (a) 1 day after culture; (b) 4 days after culture. Bar: 1 mm.
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
729
Fig. 4. Uptake of DiI-labeled acetylated LDL by the cultured SEC: (a) 6 h after culture; (b) 6 days after culture. Bar: 20 mm.
1-positive SEC, ED2-positive Kupffer cells, and desminpositive stellate cells were counted on immunocytochemical preparations of the isolated cells after 2 h (just after attachment). As shown in Table 1, the percentage of SE-1-positive cells was significantly higher, while the percentage of ED2positive cells was significantly lower in the immunomagnetic method as compared with the Percoll method. The initial contamination of desmin-positive cells was less than 1% in both methods. The cell yields by the immunomagnetic method were less than those by the Percoll method.
and 5e–h). The cells maintained sieve plates (Fig. 3b) and the capacity of DiI-acetylated LDL uptake (Fig. 4b) for at least 4 days. Although desmin-positive stellate cells were scarcely detected in the initial preparation, colonies of activated stellate cells appeared after 3 or 4 days in some experiments (Fig. 5c,d). The percentages (the mean ^ SEM, n ¼ 5) of desmin-positive cells in the total cells were 0.4 ^ 0.1, 1.6 ^ 0.4, 5.3 ^ 1.1, 6.3 ^ 2.3, 7.2 ^ 2.5, and 8.9 ^ 2.4 at 1, 2, 3, 4, 5, 6 days after culture, respectively.
3.4. Morphological and phenotypical characteristics of the isolated cells in vitro
The number of cells, which was estimated by DNA fluorometric assay, sharply decreased during the first 2 days (Fig. 6). However, the decline in cell number was stopped after 3 days and then the number appeared to increase slightly (Fig. 6). Although the TUNEL method detected only a few apoptotic cells just after attachment, there was an increase in apoptotic cell death after 1 day and the number reached a plateau after 3 days (Fig. 7). BrdU-labeled nuclei, as well as mitotic nuclei, sharply increased after 3 days when a slight increase in cell number was seen (Fig. 8).
Initially, the cultured SEC were polygonal and showed a cobblestone, sheet-like appearance (Fig. 5a,b). Then, the cells gradually became spindle-like and tended to be arranged in groups of varying sizes (Fig. 5c,d). SE-1 was positive in these cells throughout the culture period without apparent changes in its intensity and localization (Figs. 2c Table 1 Comparison of the immunomagnetic separation and the two-step Percoll gradient method a Immunomagnetic separation Percoll method Viability (%) 94.0 ^ 1.3 SE-1-positive cells (%) 97.6 ^ 0.9* ED2-positive cells (%) 0.9 ^ 0.1* Desmin-positive cells (%) 0.4 ^ 0.2 Yield ( £ 10 6 cells/liver) 10.7 ^ 0.5*
93.7 ^ 1.4 92.0 ^ 0.8 2.3 ^ 0.5 0.7 ^ 0.2 24.3 ^ 0.5
a Each value represents the mean ^ SEM for three separate experiments. *Significantly differenct as compared with the Percoll method (P , 0:01, Student’s t-test).
3.5. Changes in cell number, apoptotic cell death, and cell growth during culture
4. Discussion To study SEC functions in normal and pathological settings, it is important to prepare highly purified and viable SEC. In the present study, we have shown that viable SEC of 97.6% purity can be obtained by the immunomagnetic method using the monoclonal SE-1 antibody. SEC have been isolated by various methods including elutriation after pronase digestion [4–6] or Percoll centrifugation
730
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
Fig. 5. Time course of changes in morphology and SE-1 antigen expression in cultured sinusoidal endothelial cells isolated by the immunomagnetic beads. (a–d) Phase-contrast microscopy. (e–h) SE-1 immunocytochemistry. (a,e) One day; (b,f) 2 days; (c,g) 4 days; (d,h) 6 days. Bars: phase-contrast, 100 mm; immunocytochemistry, 20 mm.
[2,3]. SEC purity by these methods has been reported to be 80–95%. Thus, the enrichment of SEC acquired by the immunomagnetic method described in this paper appears to be the best among the previously reported methods for SEC isolation. Furthermore, the method is simple and repro-
Fig. 6. The changes in cell number during culture of sinusoidal endothelial cells. The number of cells were estimated by DNA fluorometric assay and represented as absolute fluorescence units. Data are shown as the mean ^ SEM of three separate experiments performed in duplicate.
ducible, without the need for cumbersome operation of an elutriator or pronase digestion that might hamper cell functions. Since SEC purity is already sufficiently high in the cell after the magnetic separation, it is not necessary to further enrich SEC in vitro by selective attachment and further culturing as in the case of the Percoll method. The lower cell yield of our method is considered to be mainly due to an incomplete release of cells from the beads after DNase treatment. Various kinds of vascular endothelial cells have been isolated by magnetic beads coated with antibodies (antithrombomodulin antibody (QB End/40) [11,14], anti-platelet-endothelial cell adhesion molecule 1 (PECAM-1) antibody [13,15], and S-Endo1 antibody [12]) or lectins (Ulex europaeus agglutinin-1 (UEA-1) [16–18] and Evonymus europaeus agglutinin (EEA) [19]). However, these antibodies and lectins are not necessarily suitable for the specific magnetic bead isolation of SEC. Since both vascular endothelial cells and SEC express thrombomodulin, its antibody is not able to distinguish these endothelial cells. SEC normally do not express PECAM-1 [22] and do not bind to UEA-1 [23]. It is not yet clear whether S-Endo1 recognizes SEC, while it has been reported to be highly specific to bone marrow endothelial cells [24]. An attempt has been made to isolate liver endothelial cells from Rhesus monkeys by magnetic beads coated with EEA lectin [19]. Since EEA also could bind to vascular endothelial cells in general, the isolated cells may be contaminated by non-SEC vascular
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
731
Fig. 7. Detection of apoptosis of cultured sinusoidal endothelial cells by terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL). (a) TUNEL-positive apoptotic cells (arrows) at 2 days after culture. Bar: 20 mm. (b) Quantitative analysis for time course of the frequency of apoptotic cells. Percent of TUNEL-positive cells were counted under a light microscope in ten random high-power fields (using a £40 objective lens) in each preparation. Data are shown as the mean ^ SEM of three separate experiments.
endothelial cells present in the liver. To the best of our knowledge, this is the first report of immunomagnetic separation of SEC. In our experiments, the contamination of stellate cells,
which was recognized by desmin immunostaining, was small (0.4%) in the initial cell fraction (Table 1). Then, the contaminated stellate cells were activated and proliferated to form colonies after 3 or 4 days, although the percen-
Fig. 8. Proliferation of sinusoidal endothelial cells in vitro. (a) BrdU immunocytochemistry of cultured cells (a mitotic figure is indicated by an arrow). Bar: 20 mm. (b) Time course of BrdU labeling index (B) and mitotic indices (A). BrdU-positive and mitotic cells were counted under a light microscope in three random medium-power fields (using a £20 objective lens) in each preparation. Data are shown as the mean ^ SEM of three separate experiments performed in duplicate.
732
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733
tage of stellate cell did not exceed 9% even after 6 days. However, since these contaminated stellate cells might hamper long-term cultures of SEC, it should be important to further reduce the stellate cell contamination. For this purpose, Percoll density gradient centrifugation [25], which can separate stellate cells from SEC and Kupffer cells in the mixed NPC fraction, may be effective prior to the immunomagnetic separation. Immunomagnetically isolated SEC could be maintained in vitro on type I collagen-coated coverslips at least for 6 days in the presence of growth factors, including VEGF and bFGF. VEGF has been shown to be both a growth and survival factor for SEC [26,27]. bFGF has been demonstrated to be a survival factor for vascular endothelial cells [28]. As shown in Fig. 6, the cell number decreased during the first two days, when apoptotic cell death occurred without cell proliferation, followed by the phase in which apoptosis was almost counterbalanced by the concomitant cell proliferation. The time course of changes in cell number and DNA synthesis of SEC cocultured with hepatocytes is similar to that reported recently by Krause et al. [29]. SEC underwent cell proliferation and apoptosis simultaneously during culture. Both vascular endothelial proliferation [30,31] and apoptosis [32,33] are known to be regulated by the mitogen-activated protein (MAP) kinase kinase (MEK)/extracellular signal-regulated kinase (ERK) pathway. Integrin avb3 is necessary for bFGF to induce sustained tyrosine phosphorylation of ERK for promoting angiogenesis [34]. On the other hand, vascular endothelial cell apoptosis is shown to be related with calpain-induced cleavage of the integrin b3 subunit, which is preceded by protein tyrosine dephosphorylation [35]. Furthermore, recent experiments showed that SEC apoptosis induced by cold ischemia in vivo is calpain-dependent [36]. Our preliminary results also indicated that SEC proliferation was prevented by a MEK inhibitor, PD098059. Further analysis of the signaling pathway for SEC proliferation and apoptosis is now in progress. In conclusion, using the monoclonal SE-1 antibody, we have succeeded in immunomagnetic separation of rat SEC, which is highly purified, viable, and ready for any types of culture. Our method may provide a useful tool for the biological analysis of SEC.
Acknowledgements The authors greatly appreciate Drs. H. Masuda, H. Senoo, and H. Kotanagi for critically reading the manuscript. The authors are also grateful to Mr. N. Kojima for advice in performing confocal laser scanning microscopy and to Dr. K. Kawamura and Mr. Y. Sasaki for help in preparing samples for scanning electron microscopy. This study was supported in part by grants from the Ministry of Education, Science, Sports, and Culture of Japan (09670214).
References [1] Wisse E, Braet F, Luo D, Vermijlen D, Eddouks M, Konstandoulaki M, et al. Endothelial cells of the hepatic sinusoids: a review. In: Tanikawa K, Ueno T, et al., editors. Liver diseases and hepatic sinusoidal cells, Tokyo: Springer, 1999. pp. 17–55. [2] Smedsrød B, Pertoft H, Eggertsen G, Sundsto¨ m C. Functional and morphological characterization of cultures of Kupffer cells and liver endothelial cells prepared by means of density separation in Percoll, and selective substrate adherence. Cell Tissue Res 1985;241:639– 649. [3] Braet F, De Zanger R, Sasaoki T, Baekeland M, Janssens P, Smedsrød B, et al. Assessment of a method of isolation, purification, and cultivation of rat liver sinusoidal endothelial cells. Lab Invest 1994;70:944–952. [4] Knook DL, Sleyster ECH. Separation of Kupffer and endothelial cells of the rat liver by centrifugal elutriation. Exp Cell Res 1976;99:444– 449. [5] de Leeuw AM, Barelds RJ, de Zanger R, Knook DL. Primary cultures of endothelial cells of the rat liver. Cell Tissue Res 1982;223:201–215. [6] Irving MG, Roll FJ, Huang S, Bissell DM. Characterization and culture of sinusoidal endothelium from normal rat liver: lipoprotein uptake and collagen phenotype. Gastroenterology 1984;87:1233– 1247. [7] Ohmura T, Enomoto K, Satoh H, Sawada N, Mori M. Establishment of a novel monoclonal antibody, SE-1, which specifically reacts with rat hepatic sinusoidal endothelial cells. J Histochem Cytochem 1993;41:1253–1257. [8] Morita M, Qun W, Watanabe H, Doi Y, Mori M, Enomoto K. Analysis of the sinusoidal endothelial cell organization during the developmental stage of the fetal rat liver using a sinusoidal endothelial cell specific antibody, SE-1. Cell Struct Funct 1998;23:341–348. [9] Ishiguro S, Arii S, Monden K, Adachi Y, Funaki N, Higashitsuji H, Fujita S-I, et al. Identification of the thromboxane A2 receptor in hepatic sinusoidal endothelial cells and its role in endotoxin-induced liver injury in rats. Hepatology 1994;20:1281–1286. [10] Stolz DB, Ross MA, Salem HM, Mars WM, Michalopoulos GK, Enomoto K. Cationic colloidal silica membrane perturbation as a means of examining changes at the sinusoidal surface during liver regeneration. Am J Pathol 1999;155:1487–1498. [11] Drake BL, Loke YW. Isolation of endothelial cells from human first trimester decidua using immunomagnetic beads. Hum Reprod 1991;6:1156–1159. [12] George F, Brisson C, Poncelet P, Laurent JC, Massot O, Arnoux D, et al. Rapid isolation of human endothelial cells from whole blood using S-Endo1 monoclonal antibody coupled to immuno-magnetic beads: demonstration of endothelial injury after angioplasty. Thromb Haemost 1992;67:147–153. [13] Hewett PW, Murray JC. Immunomagnetic purification of human microvessel endothelial cells using Dynabeads coated with monoclonal antibodies to PECAM-1. Eur J Cell Biol 1993;62:451–454. [14] Leach L, Bhasin Y, Clark P, Firth JA. Isolation of endothelial cells from human term placental villi using immunomagnetic beads. Placenta 1994;15:355–364. [15] Hull MA, Hewett PW, Brough JL, Hawkey CJ. Isolation and culture of human gastric endothelial cells. Gastroenterology 1996;111:1230– 1240. [16] Jackson CJ, Garbett PK, Nissen B, Schrieber L. Binding of human endothelium to Ulex europaeus 1-coated Dynabeads: application to the isolation of microvascular endothelium. J Cell Sci 1990:257–262. [17] Abbot SE, Kaul A, Stevens CR, Blake DR. Isolation and culture of synovial microvascular endothelial cells: characterization and assessment of adhesion molecule expression. Arthritis Rheum 1992;35:401– 406. [18] Masek LC, Sweetenham JW. Isolation and culture of endothelial cells from human bone marrow. Br J Haematol 1994;88:855–865.
T. Tokairin et al. / Journal of Hepatology 36 (2002) 725–733 [19] Gomez DE, Hartzler JL, Corbitt RH, Nason AM, Thorgeirsson UP. Immunomagnetic separation as a final purification step of liver endothelial cells. In Vitro Cell Dev Biol 1993;29A:451–455. [20] Dijkstra CD, Do¨ pp EA, Joling P, Kraal G. The heterogeneity of mononuclear phagocytes in lymphoid organs: distinct macrophage subpopulations in the rat recognized by monoclonal antibodies ED1, ED2 and ED3. Immunology 1985;54:589–599. [21] Rago R, Mitchen J, Wilding G. DNA fluorometric assay in 96-well tissue culture plates using Hoechst 33258 after cell lysis by freezing in distilled water. Anal Biochem 1990;191:31–34. [22] Scoazec JY, Racine L, Couvelard A, Flejou JF, Feldmann G. Endothelial cell heterogeneity in the normal human liver acinus: in situ immunohistochemical demonstration. Liver 1994;14:113–123. [23] Terada T, Nakanuma Y. Expression of ABH blood group antigens, Ulex europaeus agglutinin I, and type IV collagen in the sinusoids of hepatocellular carcinoma. Arch Pathol Lab Med 1991;115:50–55. [24] Schweitzer CM, van der Schoot CE, Drager AM, van der Valk P, Zevenbergen A, Hooibrink B, et al. Isolation and culture of human bone marrow endothelial cells. Exp Hematol 1995;23:41–48. [25] Blomhoff R, Rasmussen M, Nilsson A, Norum KR, Berg T, Blaner WS, et al. Hepatic retinol metabolism. Distribution of retinoids, enzymes, and binding proteins in isolated rat liver cells. J Biol Chem 1985;260:13560–13565. [26] Yamane A, Seetharam L, Yamaguchi S, Gotoh N, Takahashi T, Neufeld G, et al. A new communication system between hepatocytes and sinusoidal endothelial cells in liver through vascular endothelial growth factor and Flt tyrosine kinase receptor family (Flt-1 and KDR/ Flk-1). Oncogene 1994;9:2683–2690. [27] Zhang Y-J, Ikejima K, Honda H, Kitamura T, Takei Y, Sato N. Glycine prevents apoptosis of rat sinusoidal endothelial cells caused
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
733
by deprivation of vascular endothelial growth factor. Hepatology 2000;32:542–546. Araki S, Shimada Y, Kaji K, Hayashi H. Apoptosis of vascular endothelial cells by fibroblast growth factor deprivation. Biochem Biophys Res Commun 1990;168:1194–1200. Krause P, Markus PM, Schwartz P, Unthan-Fechner K, Pestel S, Fandrey J, Probst I. Hepatocyte-supported serum-free culture of rat liver sinusoidal endothelial cells. J Hepatol 2000;32:718–726. Vin˜ als F, Pouysse´ gur J. Confluence of vascular endothelial cells induces cell cycle exit by inhibiting p42/p44 mitogen-activated protein kinase activity. Mol Cell Biol 1999;19:2763–2772. Suzuki E, Nagata D, Yoshizumi M, Kakoki M, Goto A, Omata M, et al. Reentry into the cell cycle of contact-inhibited vascular endothelial cells by a phosphatase inhibitor. J Biol Chem 2000;275:3637–3644. Yang C, Chang J, Gorospe M, Passaniti A. Protein tyrosine phosphatase regulation of endothelial cell apoptosis and differentiation. Cell Growth Differ 1996;7:161–171. Kuzuya M, Satake S, Ramos MA, Kanda S, Koike T, Yoshino K, et al. Induction of apoptotic cell death in vascular endothelial cells cultured in three-dimensional collagen lattice. Exp Cell Res 1999;248:498– 508. Eliceiri BP, Klemke R, Stro¨ mblad S, Cheresh DA. Integrin avb3 requirement for sustained mitogen-activated protein kinase activity during angiogenesis. J Cell Biol 1998;140:1255–1263. Meredith JJ, Mu Z, Saido T, Du X. Cleavage of the cytoplasmic domain of the integrin b3 subunit during endothelial apoptosis. J Biol Chem 1998;273:19525–19531. Sindram D, Kohli V, Madden JF, Clavien PA. Calpain inhibition prevents sinusoidal endothelial cell apoptosis in the cold ischemic rat liver. Transplantation 1999;68:136–140.