Human Liver Endothelial Cells,But Not Macrovascular or Microvascular Endothelial Cells,Engraft in the Mouse Liver

Abstract

Liver cell transplantation has had limited clinical success so far, partly due to poor engraftment of hepatocytes. Instead of hepatocytes. other cell types, such as endothelial cells, could be used in ex vivo liver gene therapy. The goal of the present study was to compare the grafting and repopulation capacity of human endothelial cells derived from various tissues. Human endothelial cells were isolated from adult and fetal livers using anti-human CD31 antibody-conjugated magnetic beads. Human macrovascular endothelial cells were obtained from umbilical vein. Human microvascular endothelial cells were isolated from adipose tissue. Cells were characterized using flow cytometry. Liver engraftment and repopulation of endothelial cells was studied after intrasplenic transplantation in monocrotaline-treated immunodeficient mice. Following transplantation, human liver endothelial cells engrafted throughout the mouse liver. With immunoscanning electron microscopy, fenestrae in engrafted human liver endothelial cells were identified, a characteristic feature of liver sinusoidal endothelial cells. In contrast, CD31-negative liver cells, human macrovascular and microvascular endothelial cells were not capable of repopulating mouse liver. Characterization of human liver, macrovascular, and microvascular endothelial cells demonstrated expression of CD31, CD34, and CD146 but not CD45. Our study shows that only human liver endothelial cells, but not macro- and microvascular endothelial cells, have the unique capacity to engraft and repopulate the mouse liver. These results indicate that mature endothelial cells cannot transdifferentiate in vivo and thus do not exhibit phenotypic plasticity. Our results have set a basis for further research to the potential of human liver endothelial cells in liver-directed cell and gene therapy.

Keywords

Human fetal liver;Cell transplantation;Repopulation;Endothelium;Ex vivo gene therapy

Introduction

Whole-liver transplantation remains the only curative treatment for many patients suffering from inherited and acquired liver disorders. Donor livers are scarce, so treatment options such as liver cell transplantation are an attractive alternative. Although clinical trials have shown that transplantation of hepatocytes is feasible, long-term outcome is disappointing (12). This is due to numerous factors such as poor engraftment of donor hepatocytes in the host liver and loss of hepatocytes after transplantation despite the use of immunosuppression (3, 14, 18). The use of ex vivo genetically modified autologous hepatocytes overcomes the problem of donor scarcity and immunosuppression. However, the problem with poor efficiency of hepatocyte engraftment and repopulation still remains (9).

Instead of hepatocytes, other cell types, such as endothelial cells, could potentially be used in ex vivo liver gene therapy. Cell transplantation studies performed in mice have shown that mouse liver sinusoidal endothelial cells can engraft and repopulate the liver endothelium of hemophilia A mice with subsequent correction of this bleeding disorder (13, 20). However, it is unknown whether human liver endothelial cells have the capacity to engraft and repopulate the liver and thus be used in cell transplantation therapy.

Unfortunately, obtaining sufficient liver endothelial cells for transplantation is difficult due to donor scarcity. Therefore, other sources of endothelial cells suitable for liver-directed cell therapy have to be examined if endothelial cell transplantation is going to be a viable alternative treatment option.

Endothelial cells form the inner lining of blood vessels throughout the human body. These cells exhibit phenotypic heterogeneity, which is displayed by different morphology and functions depending on their environmental niche (1, 2). Liver sinusoidal endothelial cells, for example, are specialized endothelial cells, characterized by the presence of fenestrae and the lack of a basement membrane (7, 11, 21). Regeneration studies in rats have shown that bone marrow progenitor cells are capable of replacing liver sinusoidal endothelial cells after injury (16, 29). Thus far, it is unknown whether other types of human endothelial cells, such as microvascular endothelial cells, also exhibit this trait of phenotypic plasticity by functionally repopulating the mouse liver.

In the present study, we therefore compared the grafting and repopulation capacity of human liver endothelial cells with macrovascular and microvascular endothelial cells.

Materials and Methods

Animal Experiments

Animal experiments were performed in accordance with the animal ethical committee guidelines at the Academic Medical Center of Amsterdam. Male and female recombination activating gene 2 interleukin-2 receptor-g knockout (Rag2^-/-yc^-/-) (30) mice ages 3–8 weeks from our own breeding colony were used in all studies and fed ad libitum on standard laboratory chow.

Animals received an intraperitoneal injection of 200 mg/ kg monocrotaline (31) (Sigma-Aldrich, St. Louis, MO, USA) in saline 7 days and 24 h prior to the intrasplenic cell transplantation.

Mice were anesthetized with an intraperitoneal injection of FFM mix [2.5 mg Fluanisone/0.105 mg fentanyl citrate (Hypnorm, Vetapharma, Leeds, UK) and 1.25 mg Midozalam HCl/kg in H₂O (Roche, Indianapolis, IN, USA) 7 ml/kg]. Under deep anesthesia, the spleen was exposed after a subcostal incision at the left flank. The cell suspension containing 1×10⁶ cells in 100 μl PBS (Fresenius Kabi, Bad Homburg, Germany) was injected into the tip of the spleen with a 30-gauge insulin needle (Terumo, Somerset, NJ, USA). Control mice were injected with PBS. Homeostasis was secured by brief pressure for up to 5 min on the injection site. The animals were sutured and received the analgesic Temgesic (30–50 μl, 0.03 mg/ml; Reckitt Benckiser, Slough, UK) subcutaneously following.

The mice were sacrificed 14–28 days after the transplantation procedure by in vivo fixation. Under deep anesthesia, the peritoneal cavity was opened and the animals were perfused intracardially with 20 ml PBS followed by perfusion with 20 ml 2% paraformaldehyde (Merck KGaA, Darmstadt, Germany) in PBS. Prior to formaldehyde perfusion, a portion of the left liver lobe was cut off and snap frozen in liquid nitrogen for DNA analysis. After perfusion, the liver and spleen were removed and fixed in a 4% formaldehyde solution in PBS for 2–4 h at 4°C. The tissues were transferred to a 30% sucrose solution (BDH Prolabo, VWR International, Lutterworth, Leicestershire, UK) and incubated overnight at 4°C. The following morning, the tissues were snap frozen in liquid nitrogen and stored at −80°C.

In order to determine the hepatotoxicity of monocrotaline, four animals received two intraperitoneal injections of 300 mg/kg monocrotaline 7 days apart. The animals were sacrificed 24 h after the second intraperitoneal injection by in vivo fixation under deep anesthesia as described above. Blood was collected before the first monocrotaline intraperitoneal injection by orbital puncture and prior to in vivo fixation by heart puncture.

Serum alanine aminotransferase (ALT) and aspartate transaminase (AST) were measured by routine clinical chemistry. Hyaluronic acid (HA) level was measured using the Kordia Life Sciences ELISA microplate kit (Leiden, the Netherlands).

Cell Isolation

The use of human fetal liver, human adult liver, human umbilical vein endothelial cells, and human subcutaneous adipose tissue was obtained following informed consent and institutional ethical approval.

Human Fetal Liver

Human fetal liver was obtained from elective abortions. Gestational age ranged from 14 to 20 weeks. Fetal livers were digested in 0.05% collagenase IV (Sigma-Aldrich) in Hank's balanced salt solution (BioWhittaker, Cologne, Germany) for 30 min at 37°C. Dissociated cells were pelleted at 1,000 rpm for 5 min and washed once with Hank's balanced salt solution. The cells were seeded out on Primaria plates (BD Falcon, Bedford, MA, USA) and cultured for 2–7 days in Williams' E medium (BioWhittaker) containing 10% heat-inactivated fetal bovine serum (HI-FBS, BioWhittaker), 2 mM glutamine (BioWhittaker), 10 μg ml insulin, 5.5 μg/ml transferrin, 6.7 ng/ml selenium-X (ITS-mix, Life Technologies, Grand Island, NY, USA), 100 U/ml penicillin, 100 μg/ml streptomycin (penicillin/streptomycin mix, BioWhittaker), and 1 mM dexamethasone (Sigma-Aldrich). Medium was refreshed the first day following isolation to remove most of the nonadherent cells.

Human Adult Liver

Mature primary human liver was obtained from nontumor liver tissue from three female patients undergoing liver resection because of adenoma or secondary carcinoma with ages ranging from 36 to 67 years old. The liver tissue was used for hepatocyte isolation as described by Hoekstra et al. (17). The endothelial cells were collected from the supernatant remaining after washing the hepatocytes. The endothelial cells were pelleted at 1,000 rpm for 5 min and cultured on Primaria plates for 2-7 days in endothelial cell growth medium 2 (EGM2)-basal medium plus EGM-2 MV bullet kit (Lonza, Walkersville, MD, USA). Medium was refreshed the first day following isolation to remove most of the non-adherent cells.

Human Umbilical Vein Endothelial Cells

Human umbilical vein endothelial cells (HUVECs) were isolated as previously described (see Acknowledgments) (19). HUVECs were cultured in M199 medium supplemented with 20% (v/v) fetal calf serum (FCS), glutamine, penicillin/streptomycin mix (all Invitrogen), 0.05 mg/ml heparin, and endothelial cell growth supplement (ECGS) 12.5 μg/ml (both Sigma-Aldrich).

Human Subcutaneous Adipose Tissue

White adipose tissue was obtained from female patients undergoing reconstructive breast surgery using the deep inferior epigastric artery perforator (DIEP) flap procedure. Tissues from six different patients with ages ranging from 40 to 61 years old were included in this study. Microvascular endothelial cell isolation from white adipose tissue was performed by an adaptation of the protocol as described by Arts et al. (5). In short, adipose tissue was mechanically minced and incubated in crude collagenase type 1 from Clostridium histolyticum (0.2% w/v, Sigma Aldrich) in a ratio of 1:3 (fat/tissue:collagenase solution) for 30-60 min at 37°C under continuous vigorous shaking. The adipose tissue was centrifuged for 12 min at 1,200 rpm. The cell pellet containing microvascular endothelial cells was resuspended in Hank's balanced salt solution and centrifuged again for 5 min at 1,200 rpm. The cell pellet was seeded out on Primaria plates and cultured in EGM2-basal medium plus EGM-2 MV bullet kit.

Enrichment of Endothelial Cells From Human Fetal and Adult Liver

The human liver endothelial cells were isolated from the human fetal and adult liver cell suspension, after 2-7 days in culture, via magnetic separation using anti-human cluster of differentiation 31 (CD31) antibody-conjugated magnetic beads (Miltenyi Biotec, Auburn, CA, USA) according to the protocol provided by the manufacturer. The human liver endothelial cells were seeded out in Primaria six-well plates in a density of 0.5×10⁶ cells per well overnight in EGM2-basal medium plus EGM-2 MV bulletkit.

Lentiviral Transduction

To facilitate tracking of the transplanted cells in the mouse liver, the cells were transduced with a green fluorescent protein (GFP) containing lentiviral vector driven by a phosphoglycerate kinase 1 (PGK) promoter 24 h prior to transplantation with a multiplicity of infection of 3. Production of the GFP containing lentiviral vector and cell transduction were performed as described earlier (28).

Flow Cytometry Analysis

The purity of the CD31-positive cell fraction from human adult and fetal liver was determined immediately after isolation by flow cytometry. CD31-positive cells (1×10⁵) were resuspended in Williams' E medium (BioWhittaker) containing 10% heat-inactivated fetal bovine serum, 2 mM glutamine, 10 μg ml insulin, 5.5 μg/ml transferrin, 6.7 ng/ ml selenium-X (ITS-mix, Life Technologies), 100 U/ml penicillin, 100 μg/ml streptomycin, and 1 mM dexamethasone. The cells were incubated for 10 min at 4°C with phycoerythrin (PE)-conjugated mouse anti-human CD31 (Miltenyi Biotec) according to the protocol provided by the manufacturer.

For further characterization of the isolated endothelial cells, the CD31-positive cell population from human adult (n = 2) and fetal (n = 3) liver, the macro- and microvascular endothelial cells (n = 5 and n = 3, respectively) were simultaneously costained for CD45, CD34, and CD146 (Miltenyi Biotec). Cells were analyzed with a flow cytometer (FACS Calibur; Becton Dickinson, Franklin Lakes, NJ, USA). Isotype-matched nonspecific antibodies were used as controls.

Transmission Electron Microscopy

To determine the effect of monocrotaline on sinusoidal endothelial cells, liver tissue was prepared for transmission electron microscopy (TEM) as previously described (23). Following in vivo fixation, small blocks of liver were put in 1% glutaraldehyde (Merck) and 4% formaldehyde in 0.1 M sodium cacodylate buffer (pH 7.4; Sigma-Aldrich) for subsequent immersion fixation and stored at 4°C. After fixation the liver was washed in distilled water, osmicated for 60 min in 1% OsO₄ (Electron Microscopy Sciences, Hatfield PA, USA) in water, and washed again in distilled water. For contrast enhancement, the liver blocks were block-stained overnight in 1.5% aqueous uranyl acetate (Merck). Dehydration was done through a series of ethanols, and the tissue was embedded in epon LX-112 (Ladd Research, Williston, VT, USA). Ultrathin sections of 80 nm were stained with uranyl acetate and lead citrate and examined in a FEI Technai T-12 G2 Spirit Biotwin electron. Microscope (Fei Co., Eindhoven, the Netherlands).

Scanning Electron Microscopy

The presence of fenestrae in human liver endothelial cells upon culturing but prior to transplantation was determined by scanning electron microscopy (SEM). Liver endothelial cells were plated out on collagen coated cover slips (collagen from rat tail; Sigma-Aldrich) following anti-human CD31-conjugated magnetic bead isolation as described above and kept in culture overnight in EGM2-basal medium plus EGM-2 MV bullet kit. The following day, the cells were fixed in 2.5% glutaraldehyde in 0.1 M PBS for 60 min and washed three times with 0.1 M PBS. Cells were fixed in 1% OsO₄ in 0.1 M PBS for 60 min, washed three times with PBS then dehydrated through a graded series of ethanol (30–100%) and finally chemically dried with hexamethyl-disilazane (Sigma-Aldrich). The cells were visualized using a JEM-6335F field emission gun scanning electron microscope (JEOL, Peabody, MA, USA).

Immunohistochemistry

Cryosections were made of the liver and spleen by embedding the tissue in Tissue-Tek OCT medium (Bayer, Leverkusen, Germany). Sections of 5–6 μm were cut, affixed to poly-l-lysine-coated glass slides (Sigma-Aldrich), and kept at −20°C before use. Sections were washed in PBS for 15 min and subsequently incubated in specific primary monoclonal antibody against human CD31 (1:100, DakoCytomation, Glostrup, Denmark) or human CD45 (1:100, eBioscience, San Diego, CA, USA) for 1 h at room temperature. After incubation, the sections were washed in PBS/Tween 0.05% (Merck) followed by incubation with Texas red-conjugated goat anti-mouse (1:1,000, Rockland Immunochemicals, Gilbertsville, PA, USA) for another hour at room temperature. The sections were washed for another 15 min and embedded in mounting medium containing DAPI (Vector Laboratories, Peterborough, UK).

Confocal Imaging

Livers were sectioned at 7 μm and placed on Superfrost/ Plus slides (Fisher Scientific, Pittsburgh, PA, USA). Immunostaining was performed essentially as described previously (25). Livers were labeled for 1 h with either rabbit anti-human lymphatic vessel endothelial receptor 1 (LYVE-1; Angiobio, Del Mar, CA, USA) or rabbit anti-mouse LYVE-1 (Abcam, Cambridge, MA, USA), both used 1:100 in PBG [PBS supplemented with 0.5% bovine serum albumin and 0.15% glycine (Sigma-Aldrich)]. Goat anti-rabbit secondary antibodies conjugated to cyanine 3 (Cy3; Jackson ImmunoLabs, West Grove, PA, USA) were added at a 1:1,000 dilution in PBG for 1 h. Tissue was then rinsed three times with PBG, three times with PBS, counterstained with 0.001% Hoechst dye (bisBenzimide; Sigma-Aldrich) in ddH₂O for 30 s. Samples were cover-slipped using gelvatol [23 g poly (vinyl alcohol) 2000, 50 ml glycerol, 0.1% sodium azide to 100 ml PBS; Sigma-Aldrich] and viewed on a Fluoview 1000 confocal microscope (Olympus, Central Valley, PA, USA).

Immunoscanning Electron Microscopy

The presence of fenestrae in the transplanted human endothelial cells in the mouse liver was determined by immuno-SEM. Liver from transplanted cell recipients and untreated control mice was removed after in vivo fixation with 2% paraformaldehyde in PBS. The tissue was incubated in fixative for another 2–4 h and then transferred to PBS. One-millimeter tissue slices were obtained and evaluated under a fluorescence microscope for the presence of GFP-positive human endothelial cells. The positive slices were washed three times with PBS and then three times with PBG, blocked as above, and then stained with rabbit anti-human LYVE-1 (AngioBio 11-032) in PBG for 4 h. Following six washes with PBG, cells were labeled with goat anti-rabbit secondary antibody (1:25 in PBG) conjugated to 15 nm gold particles (Amersham) overnight at 4^oC followed by three washes with PBG and then three washes with PBS. Cells were fixed in 2.5% glutaraldehyde in PBS for 1 h. Cells were washed three times with PBS, then dehydrated through a graded series of ethanol (30-100%), chemically dried with hexamethyldisilazane, then overcoated with vaporized carbon (Cressington 108 Carbon/A, Watford, England). The cells were visualized using a JEM-6335F field emission gun scanning electron microscope (JEOL). Field emission backscattered electron and standard scanning electron digitized images were taken in tandem to identify areas of gold labeling on the cell surfaces.

Quantification

We determined the repopulation success of the transplanted human endothelial cells in the mouse liver using the PCR approach involving the amplification of human repetitive sequences according to Becker et al. (6). DNA was extracted from cryopreserved liver tissue of transplanted and control mice and normal human liver tissue using either the NucleoSpin Tissue DNA isolation kit (Bioke, Leiden, the Netherlands) or the QIAamp DNA FFPE Tissue (Qiagen, Germantown, MD, USA) isolation kit for PFA fixed tissue. DNA (100 ng) was used to amplify the human DNA sequence by quantitative PCR in the Roche LightCycler480 (Indianapolis, IN, USA) with the fast-start SYBRgreen kit (Roche). Primers: forward GGG ATA ATT TCA GCT GAC TAA ACA G and TaqMan reverse AAA CGT CCA CTT GCA GAT TCT AG. The PCR cycles were: 94°C for 10 min, 94°C for 10 s, 60°C for 10 s, and 72°C for 20 s (45×). Chromosomal DNA was isolated from human and mouse livers and dilutions of human DNA in mouse DNA were used to generate standard curves. The quantitative PCR data were analyzed using the LinRegPCR program (http://linregpcr.hfrc.nl/) (27).

Results

Characterization and Analysis of the Endothelial Cells

We isolated endothelial cells from human fetal and adult liver using magnetic separation with anti-human CD31-conjugated magnetic beads. As shown by flow cytometry, CD31-positive cells were efficiently purified from human adult and fetal liver (Fig. 1).

Figure 1.

Purification and morphological characterization of human liver endothelial cells. The purification of endothelial cells was analyzed by flow cytometry using an antibody to cluster of differentiation 31 (CD31). CD31-positive cells were efficiently purified from human adult (top) and fetal liver (bottom). In each panel, the solid peaks represent staining with an isotype control antibody. The right peaks represent CD31 staining and show that the cells were efficiently purified. The CD31-positive cells were isolated from fetal liver as described and fixed on collagen coated cover slips. Scanning electron microscope (SEM) imaging showed that liver endothelial cells were largely defenestrated, but occasional cells expressed fenestrations (arrows, inset).

Human liver endothelial cells were examined for the presence of fenestrae in vitro using SEM. As shown in Figure 1, the cells were largely defenestrated.

The phenotype of isolated endothelial cells from human adult (n = 2) and fetal liver (n = 3) was characterized using flow cytometry. The CD31-positive cell fraction from both adult and fetal liver highly costained for CD146 (85% and 64 ± 14%, respectively). The endothelial cells isolated from fetal liver also highly costained with CD34 (55 ± 9%). In contrast, slightly lower staining was observed for adult liver endothelial cells with CD34 (54%). There was a minimal amount of CD45-positive cells present in the CD31-positive cell suspension from both adult and fetal liver (3% and 7 ± 5%) that costained with CD31 (Fig. 2, upper two rows). This indicates that the majority of CD31-positive cells are not of hematopoietic origin.

Figure 2.

Characterization of endothelial cells from different tissues. Endothelial cells isolated from different tissues were characterized using a panel of different antibodies. Endothelial cells isolated from adult liver and fetal liver are CD31-positive and CD45-negative. The CD31-positive cells from fetal and adult liver are highly CD146 positive. The endothelial cells from fetal liver are also highly CD34 positive. In contrast, endothelial cells isolated from adult liver have a low CD34 expression. All macrovascular (human umbilical cord vein endothelial cells; HUVECs) endothelial cells are CD31 and CD146 positive. There is no costaining with CD45, and they are low positive for CD34. The microvascular (human white adipose tissue) endothelial cells are CD31 positive, and few cells are CD45 positive. There is a high CD34 expression, but CD146 expression is low.

The macrovascular endothelial cells were isolated from human umbilical veins and characterized after passages 0, 2, and 4. All macrovascular endothelial cells (n = 5) expressed CD31. They were positive for CD34 (66 ± 35%), CD146 (98 ± 1%), and CD45-negative (1 ± 1%) (Fig. 2, third row).

Microvascular endothelial cells were isolated from human subcutaneous adipose tissue. Using flow cytometry, we found that the microvascular endothelial cells (n = 3) were positive for CD31 and highly positive for CD34 (81 ± 15%). They were also CD146 positive (9 ± 3%), although the signal was not as strong. They were negative for CD45 (3 ± 1%) (Fig. 2, lower row).

All gates were corrected for background staining with isotype control-matched antibodies (data not shown).

Selective Injury of Mouse Sinusoidal Liver Endothelium Following Monocrotaline Treatment

The serum hyaluronic acid level, a marker for liver endothelial cell injury, was not significantly increased following monocrotaline treatment (Fig. 3A). The serum ALT and AST levels were also not significantly increased following two intraperitoneal injections of 300 mg/kg monocrotaline 24 h after the second injection (Fig. 3B). However, sinusoidal endothelial cell injury could be observed by electron microscopy in monocrotaline-treated mice (Fig. 3C) compared to the control mouse (Fig. 3D). The mouse endothelium was disrupted, but the morphology of the remaining liver was normal.

Figure 3.

Selective injury of liver sinusoidal endothelial cells following monocrotaline treatment. Animals treated with monocrotaline were analyzed for liver injury. Following two intraperitoneal injections of monocrotaline, no increase in serum liver hyaluronic acid was observed 24 h after the second injection (A). Neither was there an increase in the levels of the liver enzymes alanine aminotransferase (ALT) and aspartate transaminase (B). However, electron microscopy shows disruption of the mouse liver endothelium (C) compared to the control mouse (D). Arrows point at the endothelium. Scale bars: 2 and 5 μm, respectively.

Human Liver Endothelial Cells Can Engraft and Repopulate the Liver of Rag2^–/–yc^–/– Mice

We determined whether human liver endothelial cells are suitable for transplantation and restoration of the liver endothelium by intrasplenically transplanting human liver endothelial cells in Rag2^-/-yc^-/- mice that had been treated with monocrotaline (31). To facilitate tracking, endothelial cells were transduced with a GFP-expressing lentiviral vector prior to transplantation. The mice did not show any signs of adverse effects after monocrotaline treatment. Intrasplenic transplantation of 0.8–1.2×10⁶ human fetal endothelial cells resulted in substantial engraftment and repopulation in mouse liver (n = 6) (Fig. 4). Clusters of GFP-expressing cells were seen in the spleen and throughout the mouse liver. There was staining of the repopulated human liver endothelial cells with an anti-human CD31 antibody that does not cross react with mouse CD31. In sham-operated negative control mouse liver and spleen sections, no GFP-expressing cells and no human CD31 staining were observed. There was no costaining of GFP-expressing CD31-positive cells with specific antihuman CD45 (data not shown). Transplanted human liver endothelial cells were not observed in sections of the lungs and kidneys.

Figure 4.

Immunohistochemistry shows engraftment of human liver endothelial cells in mouse liver. Cryosections of the spleen and liver of recombination activating gene 2 interleukin-2 receptor-γ knockout (Rag2^-/-yc^-/-) mice were stained with a specific anti-human CD31 antibody after intrasplenic transplantation with either CD31-negative cells from the human fetal liver, macrovascular (HUVECs) endothelial cells, or microvascular (human white adipose tissue) or liver endothelial cells (CD31 positive) from the human fetal or adult liver. All spleen sections show the presence of the transplanted green fluorescent protein (GFP)-positive cells, indicating the transplantations were successful. In the liver, clusters of GFP-expressing cells that exhibit specific anti-human CD31 red costaining in mouse livers were only found after transplantation with human liver endothelial cells. Nuclei were stained blue with DAPI. Scale bar: 50 μm.

Similarly, engraftment and repopulation was also observed after transplantation of 0.5–1.0×10⁶ endothelial cells from human adult liver (n = 3). Colocalization of CD31 or LYVE-1 was visible in all GFP-expressing cells. Confocal images confirmed that the GFP-expressing cells had LYVE-1 membrane staining (Fig. 5).

Figure 5.

Confocal image showing GFP-positive cells with LYVE-1 membrane staining. Cryosections of the liver of Rag2^-/-yc^-/- mice were stained with a specific anti-human lymphatic vessel endothelial receptor-1 (LYVE-1) antibody after intrasplenic transplantation with endothelial cells from the human adult liver. The white arrows in the enlargement show LYVE-1 membrane staining of a transplanted GFP-positive cell. Nuclei were stained blue with Hoechst's dye. Scale bar: 10 μm (top) and 5 μm (bottom).

As a control group, we transplanted five mice with CD31-negative cells, obtained by depletion of CD31-positive cells from fetal liver using magnetic purification. The spleen contained GFP-expressing cells that did not express CD31, but no engraftment or repopulation of GFP-expressing and/or CD31-positive cells was observed in the livers of transplanted mice (Fig. 4). These experiments show that the only cells capable of liver repopulation are CD31 positive.

We evaluated the ultrastructural morphology of the transplanted human liver endothelial cells using immuno-SEM. The transplanted mouse liver showed LYVE-1-positive human cells in the mouse liver sinusoids at low magnification (Fig. 6A). Fenestrated endothelium showed positive surface backscatter signal from the LYVE-1 antibody, indicating that liver endothelial cells were of human origin (Fig. 6B). This shows that the transplanted human endothelial cells are fully differentiated into fenestrated endothelium.

Figure 6.

Immunoscanning electron microscopy of LYVE-1-positive sinusoids. (A) Lower panels show a set of human LYVE-1-positive cells as determined by strong backscattered signal within the sinusoid (white punctata representing 20 nm gold-conjugated goat anti-rabbit LYVE-1). Top panels show no backscattered signal indicating there is no human LYVE-1-positive cells in this sinusoid. Hepatocytes also do not stain for LYVE-1. (B) High magnification of repopulated mouse liver shows human LYVE-1-positive sinusoidal endothelial cells form fenestrated endothelium. Gold particles are white in both the left panel (backscattered signal merged on top of the scanning EM image) and the right panel (backscattered signal only). Arrows identify some of the fenestrations, indicating the engrafted human cell has the specific characteristic of liver sinusoidal endothelium. All LYVE-1-positive signals originates from the endothelium surface as none of the gold particles are found within the fenestrations, which would suggest signal from the underlying hepatocyte or stellate cell.

Macrovascular and Microvascular Endothelial Cells Did Not Repopulate the Mouse Liver

In order to compare grafting and repopulation capacity of different endothelial cell populations, we transplanted macrovascular and microvascular endothelial cells using the same transplantation protocol as with transplantation of human liver endothelial cells. Transplantation of 1×10⁶ macrovascular (n = 9) or microvascular endothelial cells (n = 9) showed GFP-expressing cells in the spleen, demonstrating successful transplantation. However, in contrast to human liver endothelial cells, macrovascular and microvascular cells did not engraft in the liver. Immunohistochemical anti-human CD31 staining confirmed that the macrovascular and microvascular cells did not repopulate the mouse liver (Fig. 4). Strikingly, in contrast to microvascular endothelial cells, the macrovascular endothelial cells did incorporate in the large vessel walls of the spleen (Fig. 4).

Quantification of Repopulation

We determined the repopulation success of the transplanted human liver endothelial cells by quantifying the amount of human DNA in the repopulated mice livers using quantitative PCR. After transplantation of human liver endothelial cells, we found an average of 0.1 ± 0.3% of total human DNA in repopulated mouse liver (Table 1).

Table 1.

Quantitative PCR Shows That Only Transplantation With Human Liver Endothelial Cells Leads to Detectable Repopulation

Cell Type	Microscopical Detection of Engraftment	Total Percentage of Human DNA in Mouse Liver	Repopulation Percentage of Mouse Endothelium by Human Endothelial Cells
LSEC adult (n = 2)	Yes	0.1 ± 0.1	1%
LSEC fetal (n = 5)	Yes	0.3 ± 0.4	3%
CD31 negative (n = 5)	No	Below threshold	N/A
Macrovascular (n = 4)	Sporadic single cells	Below threshold	N/A
Microvascular (n = 4)	Sporadic single cells	Below threshold	N/A

We compared the repopulation percentage of different endothelial cell types by determining the amount of human DNA in transplanted mouse livers using quantitative PCR. The mice were transplanted with human liver endothelial cells (LSEC, adult or fetal), CD31-negative liver cells, macrovascular or microvascular endothelial cells. N/A, not applicable. Mean values are presented with ±SD.

Discussion

The present study is the first to show that transplanted human fetal and adult liver endothelial cells can effectively repopulate the liver of immunodeficient mice. The human fetal and adult liver endothelial cells were capable of incorporating in the mouse liver endothelium. We also show that the capacity to repopulate the liver endothelial niche is restricted to liver endothelial cells and that macro- and microvascular endothelial cells are incapable of repopulating the liver.

CD31 is a transmembrane glycoprotein also known as PECAM-1 (platelet endothelial cell adhesion molecule 1) and is present on the surface of most endothelial cells (24). In the present study, we isolated endothelial cells from human fetal and adult liver via magnetic separation using anti-human CD31 antibody-conjugated magnetic beads. In addition to CD31, the human liver endothelial cells expressed both endothelial cell surface markers CD34 and CD146. After transplantation, the engrafted human liver endothelial cells still expressed CD31 and formed a fenestrated sinusoidal lining, a specific characteristic of liver sinusoidal endothelial cells, as shown by the immuno-SEM on repopulated mouse liver sections.

There are several studies that indicate that liver endothelial cells in culture lose their fenestrae organized in sieve plates (8, 10). In our study, the liver endothelial cells dedifferentiated and lost their fenestrae in vitro as well. Regardless of their phenotype in vitro, we have shown that, after transplantation, the CD31-positive cells were capable of engrafting and repopulating the mouse liver and furthermore exhibited fenestrae in vivo. This observation indicates that, in the appropriate liver microenvironment, these cells resemble true liver sinusoidal endothelial cells in vivo.

In some of the transplantation experiments, endothelial cells isolated from human fetal liver were used. The human fetal liver is a prominent site of hematopoiesis and contains differentiated blood cells as well as hematopoietic stem cells (15, 26). Potentially, the human fetal liver can also contain endothelial progenitor cells. However, the CD45-positive hematopoietic cells present in the fetal liver were negative for CD31. Also, immunohistochemistry showed that CD31-positive transplanted human liver endothelial cells were negative for CD45. Furthermore, there are also CD45-positive hematopoietic cells present in the CD31-negative cell fraction and that fraction did not give rise to repopulation of the liver endothelium (data not shown). These experiments confirm that the repopulation can be entirely attributed to CD31-positive human liver endothelial cells.

Phenotypic plasticity is a trait most commonly attributed to stem cells. Recent in vitro studies however have suggested that endothelial cells may also exhibit phenotypic plasticity (4, 22). In order to examine whether mature endothelial cells can transdifferentiate in vivo, we compared the grafting and repopulation capacity of macro- and microvascular endothelial cells to human liver endothelial cells. In vitro, all three types of endothelial cells expressed CD31, CD146, and CD34 at different levels. However, in contrast to human liver endothelial cells, both macrovascular and microvascular endothelial cells were incapable of repopulating the mouse liver after intrasplenic transplantation.

After transplantation of human liver endothelial cells, we found a repopulation success of 0.1 ± 0.3% of total DNA. Since human liver endothelial cells constitute 10–20% of total liver cells (2), an average of 1–3% of the mouse endothelium was replaced by human cells. In some cases, up to 10% replacement was detected.

In contrast, the livers of mice transplanted with either macrovascular or microvascular endothelial cells did not contain a detectable amount of human DNA. From the results of our in vivo transplantation experiments, it appears that repopulation of the mouse liver is a specific characteristic of human liver endothelial cells. Apparently, under our experimental conditions, mature endothelial cells do not have the capacity to transdifferentiate and thus do not exhibit phenotypic plasticity. Our results also indicate that macrovascular and microvascular endothelial cells are not suitable sources for cell transplantation for liver disorders.

In summary, human fetal and adult liver endothelial cells can incorporate in the mouse liver after transplantation. After transplantation, the engrafted human liver endothelial cells exhibited fenestrae. Compared to human macro- and microvascular endothelial cells, human liver endothelial cells can be transplanted with a high efficiency and are the only endothelial cell type that can incorporate in the mouse liver niche. These results show that mature endothelial cells do not exhibit phenotypic plasticity and the ability to repopulate the liver endothelium is a specific characteristic of human liver endothelial cells.

Footnotes

Acknowledgments

We would like to thank the staff of the Bloemenhove clinic (Heemstede, the Netherlands) for supplying us with the fetal liver tissue and Dr. Kees Weijer, Arie Voordouw, Stephan M. Pouw, and Dr. Nicolas LeGrand for processing these tissues (AMC, Amsterdam, the Netherlands). We would also like to thank Dr. Thomas van Gulik and Ruurdtje Hoekstra (AMC, Amsterdam, the Netherlands) for supplying us with human adult liver and Dr. Carlie de Vries, Ruud Fontijn, Maarten Janmaat, Dr. Mirella Gouverneur, and Martin Siemerink (AMC, Amsterdam, the Netherlands) for isolating and supplying us with HUVECs. Finally, we owe thanks to Leonie Woerdeman and Joris Hage from the Netherlands Cancer Institute (NKI, Amsterdam, the Netherlands) for supplying us with human subcutaneous adipose tissue. Financial support: This research was supported by a Mozaïek grant of the Netherlands Organisation for Scientific Research (NWO) under project number 017.002.060 to E. El Filali and by NIH CA76541 to Donna Stolz. The authors declare no conflict of interest.

References

Aird

W. C.

Phenotypic heterogeneity of the endothelium: I. Structure, function, and mechanisms. Circ. Res. 100: 158–173; 2007.

Aird

W. C.

Phenotypic heterogeneity of the endothelium: II. Representative vascular beds. Circ. Res. 100: 174–190; 2007.

Allen

K. J.

; Mifsud

N. A.

; Williamson

; Bertolino

; Hardikar

Cell-mediated rejection results in allograft loss after liver cell transplantation. Liver Transpl. 14: 688–694; 2008.

Amatschek

; Kriehuber

; Bauer

; Reininger

; Meraner

; Wolpl

; Schweifer

; Haslinger

; Stingl

; Maurer

Blood and lymphatic endothelial cell-specific differentiation programs are stringently controlled by the tissue environment. Blood 109: 4777–4785; 2007.

Arts

C. H.

; de

G. P.

; Heijnen-Snyder

G. J.

; Blankensteijn

J. D.

; Eikelboom

B. C.

; Slaper-Cortenbach

I. C.

Application of a clinical grade CD34-mediated method for the enrichment of microvascular endothelial cells from fat tissue. Cytotherapy 6: 30–42; 2004.

Becker

; Nitsche

; Neumann

; Aumann

; Junghahn

; Fichtner

Sensitive PCR method for the detection and real-time quantification of human cells in xenotransplantation systems. Br. J. Cancer 87: 1328–1335; 2002.

Braet

; Wisse

Structural and functional aspects of liver sinusoidal endothelial cell fenestrae: A review. Comp. Hepatol. 1: 1; 2002.

Couvelard

; Scoazec

J. Y.

; Dauge

M. C.

; Bringuier

A. F.

; Potet

; Feldmann

Structural and functional differentiation of sinusoidal endothelial cells during liver organogenesis in humans. Blood 87: 4568–4580; 1996.

Dagher

; Nguyen

T. H.

; Groyer-Picard

M. T.

; Lainas

; Mainot

; Guettier

; Pariente

; Franco

; Weber

Efficient hepatocyte engraftment and long-term transgene expression after reversible portal embolization in nonhuman primates. Hepatology 49: 950–959; 2009.

10.

DeLeve

L. D.

; Wang

; McCuskey

M. K.

; McCuskey

R. S.

Rat liver endothelial cells isolated by anti-CD31 immunomagnetic separation lack fenestrae and sieve plates. Am. J. Physiol. Gastrointest. Liver Physiol. 291: G1187–G1189; 2006.

11.

Eriksson

; Fraser

J. R.

; Laurent

T. C.

; Pertoft

; Smedsrod

Endothelial cells are a site of uptake and degradation of hyaluronic acid in the liver. Exp. Cell Res. 144: 223–228; 1983.

12.

Fisher

R. A.

; Strom

S. C.

Human hepatocyte transplantation: Worldwide results. Transplantation 82: 441–449; 2006.

13.

Follenzi

; Benten

; Novikoff

; Faulkner

; Raut

; Gupta

Transplanted endothelial cells repopulate the liver endothelium and correct the phenotype of hemophilia A mice. J. Clin. Invest. 118: 935–945; 2008.

14.

Fox

I. J.

; Chowdhury

J. R.

; Kaufman

S. S.

; Goertzen

T. C.

; Chowdhury

N. R.

; Warkentin

P. I.

; Dorko

; Sauter

B. V.

; Strom

S. C.

Treatment of the Crigler-Najjar syndrome type 1 with hepatocyte transplantation. N. Engl. J. Med. 338: 1422–1426; 1998.

15.

Hann

I. M.

; Bodger

M. P.

; Hoffbrand

A. V.

Development of pluripotent hematopoietic progenitor cells in the human fetus. Blood 62: 118–123; 1983.

16.

Harb

; Xie

; Lutzko

; Guo

; Wang

; Hill

C. K.

; Kanel

G. C.

; DeLeve

L. D.

Bone marrow progenitor cells repair rat hepatic sinusoidal endothelial cells after liver injury. Gastroenterology 137: 704–712; 2009.

17.

Hoekstra

; Deurholt

; ten

B. L.

; Desille

; van Wijk

A. C.

; Clement

; Oude Elferink

R. P.

; van Gulik

T. M.

; Chamuleaut

R. A.

Assessment of in vitro applicability of reversibly immortalized NKNT-3 cells and clonal derivatives. Cell Transplant. 15: 423–433; 2006.

18.

Horne

P. H.

; Zimmerer

J. M.

; Fisher

M. G.

; Lunsford

K. E.

; Nadasdy

; van

R. N.

; Bumgardner

G. L.

Critical role of effector macrophages in mediating CD4-dependent alloimmune injury of transplanted liver parenchymal cells. J. Immunol. 181: 1224–1231; 2008.

19.

Jaffe

E. A.

; Nachman

R. L.

; Becker

C. G.

; Minick

C. R.

Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J. Clin. Invest. 52: 2745–2756; 1973.

20.

Kumaran

; Benten

; Follenzi

; Joseph

; Sarkar

; Gupta

Transplantation of endothelial cells corrects the phenotype in hemophilia A mice. J. Thromb. Haemost. 3: 2022–2031; 2005.

21.

Limmer

; Ohl

; Kurts

; Ljunggren

H. G.

; Reiss

; Groettrup

; Momburg

; Arnold

; Knolle

P. A.

Efficient presentation of exogenous antigen by liver endothelial cells to CD8⁺ T-cells results in antigen-specific T-cell tolerance. Nat. Med. 6: 1348–1354; 2000.

22.

Medici

; Shore

E. M.

; Lounev

V. Y.

; Kaplan

F. S.

; Kalluri

; Olsen

B. R.

Conversion of vascular endothelial cells into multipotent stem-like cells. Nat. Med. 16: 1400–1406; 2010.

23.

Mook

O. R.

; Van

M. J.

; Vreeling-Sindelarova

; Jonges

; Frederiks

W. M.

; Van Noorden

C. J.

Visualization of early events in tumor formation of eGFP-transfected rat colon cancer cells in liver. Hepatology 38: 295–304; 2003.

24.

Pusztaszeri

M. P.

; Seelentag

; Bosman

F. T.

Immunohistochemical expression of endothelial markers CD31, CD34, von Willebrand factor, and Fli-1 in normal human tissues. J. Histochem. Cytochem. 54: 385–395; 2006.

25.

Ross

M. A.

; Sander

C. M.

; Kleeb

T. B.

; Watkins

S. C.

; Stolz

D. B.

Spatiotemporal expression of angiogenesis growth factor receptors during the revascularization of regenerating rat liver. Hepatology 34: 1135–1148; 2001.

26.

Rowley

P. T.

; Ohlsson-Wilhelm

B. M.

; Farley

B. A.

Erythroid colony formation from human fetal liver. Proc. Natl. Acad. Sci. USA 75: 984–988; 1978.

27.

Ruijter

J. M.

; Ramakers

; Hoogaars

W. M.

; Karlen

; Bakker

; van den Hoff

M. J.

; Moorman

A. F.

Amplification efficiency: Linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res. 37: e45; 2009.

28.

Seppen

; Rijnberg

; Cooreman

M. P.

; Oude Elferink

R. P.

Lentiviral vectors for efficient transduction of isolated primary quiescent hepatocytes. J. Hepatol. 36: 459–465; 2002.

29.

Stolz

D. B.

; Ross

M. A.

; Ikeda

; Tomiyama

; Kaizu

; Geller

D. A.

; Murase

Sinusoidal endothelial cell repopulation following ischemia/reperfusion injury in rat liver transplantation. Hepatology 46: 1464–1475; 2007.

30.

Weijer

; Uittenbogaart

C. H.

; Voordouw

; Couwenberg

; Seppen

; Blom

; Vyth-Dreese

F. A.

; Spits

Intrathymic and extrathymic development of human plasmacytoid dendritic cell precursors in vivo. Blood 99: 2752–2759; 2002.

31.

Witek

R. P.

; Fisher

S. H.

; Petersen

B. E.

Monocrotaline, an alternative to retrorsine-based hepatocyte transplantation in rodents. Cell Transplant. 14: 41–47; 2005.

Human Liver Endothelial Cells,But Not Macrovascular or Microvascular Endothelial Cells,Engraft in the Mouse Liver

Abstract

Keywords

Introduction

Materials and Methods

Animal Experiments

Cell Isolation

Human Fetal Liver

Human Adult Liver

Human Umbilical Vein Endothelial Cells

Human Subcutaneous Adipose Tissue

Enrichment of Endothelial Cells From Human Fetal and Adult Liver

Lentiviral Transduction

Flow Cytometry Analysis

Transmission Electron Microscopy

Scanning Electron Microscopy

Immunohistochemistry

Confocal Imaging

Immunoscanning Electron Microscopy

Quantification

Results

Characterization and Analysis of the Endothelial Cells

Selective Injury of Mouse Sinusoidal Liver Endothelium Following Monocrotaline Treatment

Human Liver Endothelial Cells Can Engraft and Repopulate the Liver of Rag2–/–yc–/– Mice

Macrovascular and Microvascular Endothelial Cells Did Not Repopulate the Mouse Liver

Quantification of Repopulation

Discussion

Footnotes

Acknowledgments

References

Human Liver Endothelial Cells Can Engraft and Repopulate the Liver of Rag2^–/–yc^–/– Mice