SEARCH

SEARCH BY CITATION

Keywords:

  • anti-endothelial cell antibody;
  • elastin;
  • hepatoportal sclerosis;
  • idiopathic portal hypertension;
  • obliterative portal venopathy

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References

Hepatoportal sclerosis accompanied by dense elastic fibre deposition is generally regarded as the primary lesion in the development of idiopathic portal hypertension (IPH). This study was performed to clarify the mechanism of elastic fibre deposition in the peripheral portal tracts of IPH liver in relation to serum anti-endothelial cell antibodies (AECA). In-vitro experiments were performed using human dermal microvascular endothelial cells (HMVEC) and patients' sera. The presence of serum AECA was assayed by a cell-based enzyme-linked immunosorbent assay (ELISA) using HMVEC. Immunohistochemical analysis of elastin was performed using liver tissue sections of IPH patients. IPH sera contained one or more AECA that could bind to the vascular endothelial cells of the peripheral portal tracts of the liver. When the value of AECA greater than the mean ± 2 standard deviations of healthy controls was regarded as positive, the positive detection rate of either immunoglobulin (Ig)G, IgA or IgM AECA in IPH sera was 30% (10 of 33 cases). IPH sera induced the expression of elastin in HMVEC, which appeared to be associated with the presence of AECA. Apoptosis was also induced in HMVEC by the stimulation with IPH sera. In vivo, elastin expression was observed in the endothelial cells of the peripheral portal tracts of IPH livers in a proportion of cases. The disease pathogenesis of IPH seems to be heterogeneous, and this study elucidated a possible contribution of the induction of elastin expression in the portal vessels to hepatoportal sclerosis of IPH, which might be linked to serum AECA as a causative factor.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References

Idiopathic portal hypertension (IPH) is characterized by a presinusoidal, non-cirrhotic portal hypertension in the absence of a known cause of liver disease [1,2]. The obliterative portal venopathy due to hepatoportal sclerosis is generally regarded as the primary lesion in the development of intrahepatic haemodynamic changes of IPH. Luminal narrowing or obliteration of peripheral portal vein accompanied by dense deposits of elastic fibres is a characteristic histological feature observed frequently in IPH [3].

Many theories have been proposed on the development of IPH, and immunological disorder is one of the plausible aetiological factors [2]. The patients with IPH have been reported frequently in association with immunological disorders such as systemic sclerosis [4]. In patients with systemic sclerosis, a fibrogenic process has been suggested as an aetiological factor in the development of IPH [5,6]. The presence of serum autoantibodies such as anti-nuclear or anti-smooth muscle antibodies, the expression of human leucocyte antigen D-related (HLA-DR) antigen on portal microvessels and the predominance of T helper 1 cells in peripheral and spleen lymphocytes in patients with IPH are other examples indicative of the involvement of immunological disorders in the disease process [7–9].

Recently, we have shown that endothelial to mesenchymal transition (EndMT) of the endothelial cells of portal vein is associated with portal venous stenosis and collagen deposition in the peripheral portal tracts of IPH liver [10]. The endothelial cells are capable of transforming to myofibroblast-like cells via transforming growth factor (TGF)-beta1/Smad activation, and participate in tissue fibrosis by producing extracellular matrix molecules, including collagen. EndMT has also been implicated in cutaneous fibrogenesis in systemic sclerosis [11], suggesting that a similar mechanism may exist in the fibrogenic process of IPH and systemic sclerosis.

We also showed that fibulin-5, an essential protein that links elastic fibres to cells and regulate fibre assembly and organization, was involved phlebosclerosis of major portal vein branches associated with elastic fibre deposition in IPH [12]. However, in our series of studies as well as other previous reports, the mechanism of elastic fibre deposition in the peripheral portal tracts of IPH remains to be determined.

Anti-endothelial cell antibodies (AECA) have been identified as circulating autoantibodies targeting the endothelial cells, and are detectable in a heterogeneous group of autoimmune and/or inflammatory conditions, including vasculitis [13]. In systemic sclerosis, 20–86% of patients exhibit positive test for AECA [14,15]. Sera containing AECA from patients with systemic sclerosis have been demonstrated to induce the expression of fibrillin-1, one of the main components of microfibrils which interacts with fibulin-5 during the elastic fibre assembly, and to induce apoptosis in human dermal endothelial cells [16]. Despite the pathogenic significance of ACEA having been clarified in systemic sclerosis, its significance in IPH is largely unknown.

This study was performed to clarify the mechanism of hepatoportal sclerosis of IPH, focusing particularly on the capacity of IPH sera to induce fibrosis-related molecules and to induce apoptosis of the endothelial cells. In this context, the presence and significance of AECA in IPH sera were also examined.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References

Patients

The diagnosis of IPH was made according to the clinical symptoms, laboratory data, image analysis and/or liver biopsy findings, which was based principally on the guidelines of the Japanese Society for Portal Hypertension (The General Rules for Study of Portal Hypertension, which was written in Japanese, published in 2004). Sera from patients with IPH were selected from frozen (−80°C) samples stored at our institute. As controls, sera from healthy volunteers and patients with chronic viral hepatitis/liver cirrhosis (CVH/LC) associated with viral infection of hepatitis C were used. This human study was performed with the approval of the ethics committee of Kanazawa University Graduate School of Medicine.

Endothelial cell culture

Human dermal microvascular endothelial cells (HMVEC) were purchased from Cell Applications, Inc. (San Diego, CA, USA), and were maintained in a humidified incubator at 37°C in 5% CO2 with endothelial growth medium (CadmecTM growth medium, Cell Applications, Inc.). Incubations were conducted with patient and control sera using a concentration of 10% serum. Experiments were performed with HMVEC at passages 2–3.

Reverse transcription–polymerase chain reaction (RT–PCR) and quantitative real-time PCR

Total RNA (1 µg) was extracted using an RNA extraction kit (RNeasy mini; Qiagen, Tokyo, Japan) from HMVEC. RT–PCR was performed with the use of reverse transcriptase (ReverTra Ace; Toyobo Co., Osaka, Japan) and Taq polymerase (TaKaRa Ex Taq; Takara Bio Inc., Ohtsu, Japan). The sequences of the primers and conditions for PCR used are shown in Table 1. The PCR products were subjected to 2% agarose gel electrophoresis and stained with ethidium bromide.

Table 1.  Sequences of the primers and polymerase chain reaction (PCR) conditions used in this study.
GeneSequences (5′-3)Annealing temperature (°C)PCR cyclesProduct size (bp)
  1. CTGF, connective tissue growth factor; bp, base pairs; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; TGF, transforming growth factor.

TGF-β1GCCAGAGTGGTTATCTTTTG6225285
CTGAAGCAATAGTTGGTGTC
CTGFGGGAATCCTGTCATTTGTGG5540183
CTTTCCTGCCTAACCCCAAT
Fibrillin-1TTCCTTCTAACGCTGCACCT6035200
TGATCCACTGTGTGCCAACT
Fibulin-5CAATTTACAAGGGGGCTTCA6035193
GGCTTGCATTTGGAAGATGT
FibronectinCTTCGAATTATGAGCAGGAC6040196
GCATCATAGTTCTGTGTGGT
S100A4GGTGACAAGTTCAAGCTCAA5540214
CTGGGAAGCCTTCAAAGAAT
COL1A1GTCTTCTGCAACATGGAGAC6240245
CAGTGGTAGGTGATGTTCTG
ElastinGTCGCAGGTGTCCCTAGTGT6540120
GGTCCCCACTCCGTACTTG
GAPDHGAGTCAACGGATTTGGTCGT5530240
TTGATTTTGGAGGGATCTC

Quantitative real-time PCR was performed according to a standard protocol using the SYBR Green PCR Master Mix (Toyobo Co.) and ABI Prism 7700 Sequence Detection System (PE Applied Biosystems, Warrington, UK). Cycling conditions were incubation at 50°C for 2 min, 95°C for 10 min and 40 cycles of 95°C for 15 s and 60°C for 1 min. Fold difference compared with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression was calculated.

Immunocytochemistry

HMVEC were fixed with 4% paraformaldehyde for 15 min, and permeabilized for 15 min with 0·1% Triton X-100. After pretreatment with blocking serum (DakoCytomation, Glostrup, Denmark), the cells were incubated for 1 h at room temperature with primary antibodies against elastin (1 : 50, rabbit polyclonal; Biologo, Kronshagen, Germany). Protein expression was detected using the alkaline phosphatase-labelled polymer (the Histofine system; Nichirei, Tokyo, Japan). Colour development was performed using the Vector red alkaline phosphatase substrate kit (Vector Laboratories, Burlingame, CA, USA), and nuclei were stained with 4′6-diamidino-2-phenylindole (DAPI). The signals were detected under immunofluorescence confocal microscopy.

Enzyme-linked immunosorbent assay

Total proteins were extracted from HMVEC using T-PER protein extraction reagent (Pierce Chemical Co., Rockford, IL, USA), and the concentration of total protein was measured spectrometrically. The levels of elastin were determined using an enzyme-linked immunosorbent assay (ELISA) kit (USCN Life Science, Inc., Wuhan, China), according to the manufacturer's instructions. Briefly, samples were added to a 96-well plate coated with an antibody for elastin, and incubated for 2 h at 37°C. After washing, the plate was incubated with the detection reagent for 1 h at 37°C. Colour development was performed using a substrate solution for 30 min and the absorbance at 450 nm was measured. The values were expressed as the ratio of the concentrations of elastin (ng/ml)/total protein (mg/ml).

Apoptosis assay

The effect of patients' sera on apoptosis of HMVEC was determined using the single-strand DNA (ssDNA) apoptosis ELISA kit (Chemicon Int., Temecula, CA, USA), according to the manufacturer's instructions. This method is based on the selective denaturation of DNA in apoptotic cells by formamide, which is a gentle agent that denatures DNA in apoptotic cells, but not in necrotic cells or in cells with DNA breaks in the absence of apoptosis. In brief, HMVEC were seeded on 96-well plates at a concentration of 1 × 104 cells/well. At a subconfluent state, HMVEC were incubated with 10% sera for 3 days. The cells were fixed with 80% methanol in phosphate-buffered saline (PBS) on the wells and incubated with formamide at 75°C for 20 min. For negative control wells, S1 nuclease (10 U/well) (Takara Bio, Inc.) was added and incubated at 37°C for 30 min for the removal of single-stranded regions in DNA–DNA hybrids. After washing with PBS, the wells were incubated with 2·5% bovine serum albumin at 37°C for 1 h to block non-specific binding sites and then incubated with antibody mixture [primary monoclonal to ssDNA and peroxidase-labelled anti-mouse immunoglobulin (Ig)M; provided in the kit] at room temperature for 30 min. After washing, colour development was performed with 2,2′-azino-bis(3-ethylbenziazoline-6-sulphoic acid) solution, and its absorbance at 410 nm was measured by an ELISA reader.

ELISA for AECA

The presence of AECA in sera was assayed by a cell-based ELISA using HMVEC [17]. HMVEC were seeded on 96-well plates at a concentration of 1 × 104 cells/well. When the cellular growth became confluent, cells were fixed with 4% paraformaldehyde for 15 min and incubated with blocking serum (DakoCytomation) for 60 min at room temperature. The serum samples, diluted in PBS at 1 : 50, were added to the well and were incubated for 2 h at room temperature. After washing, peroxidase-conjugated rabbit anti-human IgG, IgA and IgM immunoglobulins (1 : 1000; DakoCytomation) were added to each well and incubated for a further 1 h at room temperature. After washing, tetramethyl benzidine peroxidase substrate (Bethyl Laboratories, Inc., Montgomery, TX, USA) was added for 15 min, and stop solution (1 m hydrochloric acid) for 5 min. The optical density of each well was read at 450 nm by an ELISA reader. Mean values were calculated from duplicate determinations. Samples were recorded as positive if the value was greater than the mean ± 2 standard deviations (s.d.) of healthy controls.

Liver specimens

A total of 47 liver specimens were used. Twenty-two specimens corresponded to IPH. Both liver wedge biopsy and autopsy materials were included. The IPH livers of autopsy cases were collected as described previously [18]. Histology of the liver confirmed the diagnosis of IPH. As controls, liver specimens obtained from patients with CVH/LC (n = 15), and histologically normal livers (n = 10) were used. The causes of CVH/LC were viral infection of hepatitis C. Normal liver specimens were obtained from patients undergoing a partial hepatectomy for the diseases other than hepatobiliary disorders, such as metastatic colon cancer, and macroscopically and microscopically normal areas were used.

Immunohistochemistry

Liver specimens were fixed with neutral formalin, and 4-µm thick paraffin-embedded tissue sections were prepared. The sections were immersed in 0·3% hydrogen peroxidase in methanol for 20 min at room temperature to block the activity of endogenous peroxidase. After pretreatment with blocking serum (DakoCytomation), the sections were incubated overnight at 4°C with the primary antibodies against elastin (1 : 100; Biologo) and CD34 (1 : 100, mouse monoclonal; Immunotech, Marseilles, France). Sections were then incubated with secondary antibodies conjugated to peroxidase-labelled polymer, using the EnVision+ system (DakoCytomation). Colour development was performed using 3,3′-diaminobenzidine tetrahydrochloride (DAB) and the sections were slightly counterstained with haematoxylin. Negative controls were performed by substitution of the primary antibodies with non-immunized serum, resulting in no signal detection.

Double immunofluorescence staining

Double immunofluorescence staining of elastin and CD34 was performed for the liver sections. Deparaffinized sections were incubated 1 h at room temperature with the anti-CD34 antibody (1: 50; Immunotech). The sections were incubated with secondary antibodies conjugated to alkaline phosphatase-labelled polymer (the Histofine system; Nichirei). Colour development was performed using the Vector red alkaline phosphatase substrate kit (Vector Laboratories). Then, the sections were incubated overnight at 4°C with the primary antibody against elastin (1 : 100; Biologo). AlexaFluor 488 (10 µg/ml; Molecular Probes, Eugene, OR, USA) was used as a secondary antibody. Nuclei were stained with DAPI, and the sections were observed under immunofluorescence confocal microscopy.

Binding capacity of AECA to vascular endothelial liver cells

To examine the binding capacity of AECA to vascular endothelial cells of the liver, immunohistochemical analysis was performed for the normal liver using the patients' sera as the primary antibodies [19]. Tissue sections were cut from a paraffin-embedded block of normal liver, in which the liver originated from one individual who was unrelated to the donor of the sera used in this study. After deparaffinization, antigen retrieval was performed by microwaving in 10 mmol/l citrate buffer pH 6·0. To block the activity of endogenous peroxidase, sections were immersed in 0·3% hydrogen peroxidase in methanol for 20 min at room temperature. After pretreatment with blocking serum (DakoCytomation), the serum samples, diluted in PBS at 1 : 100, were applied on the sections as primary antibodies, and the sections were incubated for 1 h at room temperature. After washing, peroxidase-conjugated rabbit anti-human IgG (1 : 100; DakoCytomation) was reacted for 1 h at room temperature. Colour development was performed using DAB and the sections were slightly counterstained with haematoxylin. Negative controls were performed by substitution of the primary antibodies with PBS, resulting in no signal detection.

Histological assessment

Semiquantitative analysis was performed for the normal liver sections that were immunostained with the sera as the primary antibodies. The number of peripheral portal vein and hepatic artery with positive immunohistochemical signals in the endothelial cells was counted, and the percentage to the total number of portal vein and hepatic artery was calculated for each section. The percentage of the number of positively stained vessels was evaluated as three grades (>10%, 0–10%, 0%).

Statistics

The data were expressed as the mean ± s.d. Statistical significance was determined using the χ2 test, the Mann–Whitney U-test and the Pearson correlation test. A P-value less than 0·05 was accepted as the level of statistical significance.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References

Effects of IPH sera on the expression of fibrosis-related molecules in HMVEC

HMVEC were incubated with 10% sera of IPH, healthy controls and CVH/LC for 24 h, and the alteration of the mRNA expression of fibrosis-related molecules was screened using RT–PCR. As shown in Fig. 1, the expression of several molecules including S100A4 (a marker of EndMT) and COL1A1 appeared to be induced in HMVEC following stimulation with the sera of several cases of IPH and CVH/LC when compared to that of the control sera. Among the fibrosis-related molecules tested, the most remarkable difference was found in the induction of elastin mRNA expression following stimulation with IPH sera, and therefore further analysis was performed in this context.

image

Figure 1. Effects of idiopathic portal hypertension (IPH) sera on the expression of fibrosis-related molecules in human dermal microvascular endothelial cells (HMVEC). HMVEC were incubated with 10% sera of idiopathic portal hypertension (IPH; n = 9), healthy controls (n = 3) and chronic viral hepatitis/liver cirrhosis (CVH/LC; n = 5) for 24 h, and the alteration of the mRNA expression of fibrosis-related molecules was examined using reverse transcription–polymerase chain reaction (RT–PCR). Among the molecules tested, the induction of elastin mRNA expression was most remarkably observed following stimulation with IPH sera. CTGF, connective tissue growth factor; TGF, transforming growth factor.

Download figure to PowerPoint

Induction of elastin expression in HMVEC by IPH sera

The quantitative real-time PCR analysis showed that the sera of IPH and CVH/LC induced elastin mRNA expression in HMVEC, while its induction by control sera was negligible (Fig. 2a). Statistical analysis showed that IPH sera increased elastin mRNA expression significantly in HMVEC when compared to those of controls and CVH/LC sera (Fig. 2b).

image

Figure 2. Induction of elastin mRNA expression in human dermal microvascular endothelial cells (HMVEC) by idiopathic portal hypertension (IPH) sera. Quantitative polymerase chain reaction (PCR) analysis of the elastin mRNA expression was performed for the samples used in the studies of Fig. 1. The sera of idiopathic portal hypertension (IPH; n = 9) and chronic viral hepatitis/liver cirrhosis (CVH/LC; n = 5) induced elastin mRNA expression in HMVEC, while its induction by control sera (n = 3) was negligible (a). Statistical analysis showed that IPH sera significantly increased elastin mRNA expression in HMVEC when compared to those of controls and CVH/LC sera (b). *P < 0·01; **P < 0·05.

Download figure to PowerPoint

The protein expression of elastin was examined using HMVEC following stimulation with the sera for 5 days. Immunocytochemistry indicated that IPH sera induced elastin protein expression in HMVEC, although the extent of the positive signals was fairly weak (Fig. 3a). The analysis using ELISA confirmed that the protein extracts of HMVEC stimulated by IPH sera contained significantly high levels of elastin when compared to those of the control group (Fig. 3b).

image

Figure 3. Induction of elastin protein expression in human dermal microvascular endothelial cells (HMVEC) by idiopathic portal hypertension (IPH) sera. HMVEC were incubated with 10% sera of idiopathic portal hypertension (IPH), healthy controls and chronic viral hepatitis/liver cirrhosis (CVH/LC) for 5 days (n = 4 for each experimental group), and the protein expression of elastin was examined using immunocytochemistry and enzyme-linked immunosorbent assay (ELISA). Immunocytochemistry showed that IPH sera induced elastin protein expression in HMVEC, although the extent of the positive signals was fairly weak (a). At the phase-contrast microscope, morphological alterations from epithelioid- to myofibroblast-like appearance were not observed in HMVEC following stimulation with IPH sera (a). The protein extracts of HMVEC stimulated by IPH sera contained significantly high levels of elastin when compared to those of the control group, which were determined using ELISA (b). **P < 0·05. Original magnification ×400 (a).

Download figure to PowerPoint

At the phase-contrast microscope, morphological alterations from epithelioid- to myofibroblast-like appearance, indicative of the occurrence of EndMT, were not observed in HMVEC following stimulation with IPH sera under the experimental conditions in this study (Fig. 3a).

Immunohistochemical expression of elastin in IPH livers

The expression of elastin in the liver of patients with IPH was surveyed immunohistochemically. Positive immunohistochemical signals of elastin were observed diffusely and intensely in the peripheral portal tracts of IPH livers, while they were weak in the normal livers (Fig. 4a). In the peripheral portal tracts of IPH livers, positive immunohistochemical signals of elastin were observed along with the lumen of the portal vein, and the serial sections immunostained with the anti-CD34 antibody demonstrated that the endothelial cells of the portal vein expressed elastin (Fig. 4b, arrows).

image

Figure 4. Immunohistochemical expression of elastin in idiopathic portal hypertension (IPH) livers. Positive immunohistochemical signals of elastin were observed diffusely and intensely in the peripheral portal tracts of IPH livers, while they were weak in normal livers (a). In the peripheral portal tracts of IPH livers, positive immunohistochemical signals of elastin were observed along with the lumen of the portal vein, and the serial sections immunostained with the anti-CD34 antibody showed that the endothelial cells of the portal vein expressed elastin (b, arrows). Double immunofluorescence staining with the antibodies against elastin (coloured green) and CD34 (red) showed that the endothelial cells of the portal vein, hepatic artery and microvessels in the peripheral portal tracts of IPH livers were double-positive for the antigens (c, arrowheads indicate microvessels). BD, bile duct; HA, hepatic artery; PV, portal vein. Original magnifications: ×200 (a), ×100 (b, left panels); ×1000 (b, right panels), ×400 (c).

Download figure to PowerPoint

Double immunofluorescence staining with the anti-elastin and anti-CD34 antibodies confirmed that the endothelial cells of the portal vein of IPH livers expressed elastin (Fig. 4c). In addition, double-positive signals were observed in the endothelial cells of the hepatic artery and microvessels in the peripheral portal tracts (Fig. 4c; microvessels are indicated by arrowheads). However, the positive signals for elastin in the endothelial cells were not observed in all IPH livers studied; they were detectable in five of 22 cases (23%) of IPH. They were also observed in three of 15 cases (20%) of CVH/LC livers. The normal livers (n = 10) typically lacked positive immunohistochemical signals of elastin in the endothelial cells of the portal tracts.

Induction of apoptosis in HMVEC by IPH sera

HMVEC were treated with 10% sera for 3 days, and were subjected subsequently to the apoptosis ELISA assay. Apoptosis was detected similarly in HMVEC treated with sera of healthy controls and CVH/LC, while IPH sera increased apoptosis of HMVEC significantly (Fig. 5).

image

Figure 5. Induction of apoptosis in human dermal microvascular endothelial cells (HMVEC) by idiopathic portal hypertension (IPH) sera. HMVEC were treated with 10% sera of idiopathic portal hypertension (IPH; n = 33), healthy controls (n = 16) and chronic viral hepatitis/liver cirrhosis (CVH/LC; n = 29) for 3 days, and were subsequently subjected to the apoptosis enzyme-linked immunosorbent assay (ELISA) assay as described in the Materials and methods. Apoptosis was detected similarly in HMVEC treated with sera of healthy controls and CVH/LC, while IPH sera significantly increased apoptosis of HMVEC. *P < 0·01; **P < 0·05.

Download figure to PowerPoint

AECA in IPH sera

The presence of IgG, IgA and IgM AECA in the sera was assayed by a cell-based ELISA using HMVEC. The sera of all experimental groups composed of IPH, healthy controls and CVH/LC contained detectable levels of IgG, IgA and IgM AECA (Fig. 6). Among them, IgG AECA (Fig. 6a) and IgM AECA (Fig. 6b) were elevated significantly in IPH sera compared to those of controls and CVH/LC.

image

Figure 6. Anti-endothelial cell antibodies (AECA) in idiopathic portal hypertension (IPH) sera. The presence of immunoglobulin (Ig)G, IgA and IgM AECA in sera was assayed by a cell-based enzyme-linked immunosorbent assay (ELISA) using human dermal microvascular endothelial cells as described in the Materials and methods. The sera of all experimental groups composed of idiopathic portal hypertension (IPH; n = 33), healthy controls (n = 16) and chronic viral hepatitis/liver cirrhosis (CVH/LC; n = 29) contained detectable levels of IgG (a), IgA (b) and IgM (c) AECA. IgG and IgA AECA were significantly elevated in IPH sera compared to those of controls and CVH/LC. *P < 0·01; **P < 0·05.

Download figure to PowerPoint

When the value of AECA greater than the mean + 2 s.d. of healthy controls was regarded as positive, the positive detection rate of either IgG, IgA or IgM AECA was 30% (10 of 33 cases) in IPH sera and 17% (five of 29 cases) in CVH/LC sera (Table 2). The positive detection rate of each of IgG, IgA and IgM AECA was higher in IPH sera than in CVH/LC sera, but a statistical significance was not observed between the IPH and CVH/LC groups. Two cases of IPH were positive for both IgG and IgM AECA, and one case of IPH was positive for both IgG and IgA AECA. Such double-positive cases were not observed in CVH/LC.

Table 2.  Number of patients positive for serum anti-endothelial cell antibodies (AECA).
 ControlIPHCVH/LC
  1. The value of AECA greater than the mean ± 2 standard deviations of healthy control was regarded as positive. Two cases were positive for both immunoglobulin (Ig)G and IgM AECA, and one case was positive for IgG and IgA AECA. CVH/LC, chronic viral hepatitis/liver cirrhosis; IPH, idiopathic portal hypertension.

n163329
IgG AECA0 (0%)6 (18%)2 (7%)
IgA AECA0 (0%)2 (6%)1 (3%)
IgM AECA0 (0%)5 (15%)2 (7%)
Total0 (0%)10 (30%)5 (17%)

Correlation between AECA and elastin expression in HMVEC

As shown in Fig. 2, IPH sera induced elastin mRNA expression in HMVEC that was determined using the quantitative real-time PCR. When the determined value of elastin/GAPDH of the quantitative real-time PCR for the 17 cases studied was plotted against the value of AECA of each serum that was used for the stimulation of HMVEC, a significant linear correlation was observed for IgG AECA (Fig. 7a) and IgA AECA (Fig. 7b). The results were suggestive of the causal correlation between the induction of elastin expression in HMVEC and the presence of serum AECA.

image

Figure 7. Correlation between anti-endothelial cell antibodies (AECA) and elastin expression in human dermal microvascular endothelial cells (HMVEC). Sera from patients with idiopathic portal hypertension (IPH) induced elastin mRNA expression in HMVEC that was determined using quantitative real-time polymerase chain reaction (PCR) (Fig. 2). When the determined value of elastin/glyceraldehyde-3-phosphate dehydrogenase (GAPDH) of the quantitative real-time PCR for the 17 cases [IPH, n = 9; healthy controls, n = 3; chronic viral hepatitis/liver cirrhosis (CVH/LC), n = 5] was plotted against the value of AECA of each serum that was used for the stimulation of HMVEC, a significant linear correlation was observed for IgG (a) and IgA (b) AECA, but not for IgM AECA (c).

Download figure to PowerPoint

Similarly, the determined value of the ssDNA apoptosis ELISA assay (Fig. 5) was plotted against the value of AECA of each serum that was used for the stimulation of HMVEC. However, there was no significant correlation observed between the value of ssDNA apoptosis ELISA assay and the value of each of IgG, IgA and IgM AECA (data not shown).

Binding capacity of AECA to vascular endothelial liver cells

Finally, immunohistochemical analysis was performed for the normal liver tissue using the patients' sera as the primary antibodies. Rabbit anti-human IgG was used as the secondary antibody. This analysis aimed to clarify the binding capacity of serum IgG AECA to vascular endothelial cells of the liver.

Except for one case, control sera did not contain IgG AECA that could bind to the endothelial cells of peripheral portal tracts of the liver (Fig. 8a). By contrast, 15 of 33 cases (45%) of IPH sera yielded positive immunohistochemical signals in the endothelial cells of the peripheral portal tracts (Fig. 8a). In the portal tracts, the positive signals were observed in the endothelial cells of the portal vein, hepatic artery and microvessels.

image

Figure 8. Binding capacity of anti-endothelial cell antibodies (AECA) to vascular endothelial cells of the liver. As described in the Materials and methods, the binding capacity of serum AECA to the endothelial cells of the liver was examined by immunohistochemical analysis for the normal liver tissue using the patients' sera [idiopathic portal hypertension (IPH), n = 33; healthy controls, n = 16; chronic viral hepatitis/liver cirrhosis (CVH/LC), n = 29] as the primary antibodies. IPH sera yielded positive immunohistochemical signals in the endothelial cells of the portal vein, hepatic artery and microvessels of the peripheral portal tracts, while most of the control sera result in no positive immunohistochemical signals of the endothelial cells (a). Semiquantitative analysis of the number of positively stained vessels showed that it was higher in IPH compared to those of the controls and CVH/LC, and a significant difference was observed between IPH and control groups (b). *P < 0·01. BD, bile duct; HA, hepatic artery; PV, portal vein. Original magnifications: ×100 (a, left panels); ×400 (a, right panels).

Download figure to PowerPoint

Semiquantitative analysis of the number of positively stained vessels showed that it was higher in IPH compared to those of the controls and CVH/LC, and a significant difference was observed between IPH and control groups (Fig. 8b). Thus, IPH sera contained IgG AECA that could bind to vascular endothelial cells of the liver, although they were not invariably present in all cases of IPH.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References

This study demonstrated that IPH sera were capable of inducing elastin expression in cultured vascular endothelial cells, and elucidated a possible link between the elastin induction in the endothelial cells and the presence of serum AECA. Elastic fibre deposition in the peripheral portal tracts is one of the most characteristic histological findings of IPH. Immunohistochemical analysis of elastin in this study showed that elastin deposition was observed intensely in the peripheral portal tracts of IPH, and the positive labelling of elastin in vascular endothelial cells of portal tracts was not observed in all but five of 22 cases (22%) of IPH. The aetiopathogenesis of IPH seems to be heterogeneous, and various causative factors may be associated in the disease process [2]. Our data suggest that the induction of elastin expression in the endothelial cells of the peripheral portal tracts is one of the possible mechanisms that relate to hepatoportal sclerosis in IPH.

A previous study showed that sera of patients with systemic sclerosis induced fibrillin-1 expression in human dermal endothelial cells [16]. As shown in Fig. 1, however, the induction of fibrillin-1 mRNA in HMVEC following stimulation with IPH sera was unremarkable in this study, which might be due partially to the difference in the cell lines used for the experiments. Similarly, it is known that the sources of cultured endothelial cells tested, as well as detection methods such as cell-based ELISA and flow cytometry analysis, can influence greatly the detection rate of AECA [20].

The definition of AECA is usually restricted to autoantibodies targeting the antigens expressed on the endothelial cell surface [20]. It has been recognized that the cell-based ELISA method does not necessarily detect AECA directed against membrane antigens, and a possibility of the detection of the some kinds of antibodies against cytoplasmic antigens has been pointed out. In this regard, it is unclear whether AECA detected in this study specifically recognized membrane antigens of the endothelial cells. However, we regarded them as AECA, because cell-based ELISA is the most frequently used method for the detection of AECA. This is also true for the methods used in this study to examine the binding capacity of AECA to vascular endothelial cells of the liver.

AECA are a heterogeneous class of autoantibodies, and due to this heterogeneity identification of their antigens is difficult. Several antigens have been identified belonging to the groups of membrane components, ligand–receptor complexes and the antigens derived from the blood and attached to the cell surface [21,22]. It has been pointed out that the titre of AECA may vary in different disease stages and disease activity, and the target antigens of AECA may be different among the various pathological conditions in even one disease [20]. AECA is also detectable in healthy individuals. These observations might relate to the reason why IPH was not seen in all cases associated with AECA in this study. To address this issue, further studies for the identification of target antigens using the methods such as two-dimensional immunoblotting, together with mass spectrometry, are necessary [23]. In addition, the mechanical correlation between AECA and the elastin induction in endothelial cells is to be studied.

In has been reported that AECA may be pathogenic by inducing endothelial cell apoptosis [16,24,25]. Direct and dose-dependent induction of endothelial cell apoptosis by AECA from patients with systemic sclerosis has been observed [24]. Indeed, IPH sera induced apoptosis in HMVEC significantly in this study, although the extent of apoptotic response in HMVEC was not correlated with AECA. Apoptotic endothelial cells have be been shown to secrete connective tissue growth factor (CTGF) and promote fibrosis [26], and the levels of CTGF in sera and liver tissue are known to be up-regulated in patients with IPH [27,28], suggesting that the endothelial cell apoptosis may be involved in the disease pathogenesis of IPH.

AECA may also be associated with endothelial cell damage by up-regulating the expression of adhesion molecules and triggering an inflammatory process via complement and antibody-dependent cellular cytotoxicity. Sera containing IgG AECA from patients with systemic lupus erythematosus is shown to induce the expression of adhesion molecules such as vascular cell adhesion molecule-1 (VCAM-1) in cultured endothelial cells [29]. Interestingly, the serum level of soluble VCAM-1 is elevated in IPH patients, and VCAM-1 is expressed in vascular endothelial cells of the IPH liver [30]. These findings are supportive of the contribution of AECA in the disease pathogenesis of IPH.

Recently, EndMT has been implicated in the process of tissue fibrogenesis in several organs [31]. Our previous studies demonstrated that EndMT via TGF-beta1/Smad activation was associated with portal venous stenosis in IPH, in which the alteration of elastin expression in the endothelial cells had not been examined [10]. This study showed that IPH sera induced elastin expression in HMVEC, but the morphological alteration of HMVEC into myofibroblast-like appearance, indicative of the occurrence of EndMT, was not observed under the experimental conditions in this study. However, the expression of S100A4 mRNA, a marker of EndMT, appeared to be increased slightly in HMVEC following stimulation with IPH sera, suggesting that a partial acquisition of myofibroblastic features in HMVEC.

In summary, this study demonstrated that IPH sera were capable of inducing elastin expression and apoptosis in cultured vascular endothelial cells. IPH sera contained one or more AECA that could bind to vascular endothelial cells of the peripheral portal tracts of the liver, and a possible link between AECA and the induction of elastin expression in the endothelial cells was suggested. In vivo, elastin expression was observed in the portal vessels of IPH livers in a proportion of cases. The disease pathogenesis of IPH seems to be heterogeneous, and a possible contribution of the induction of elastin expression in the portal vessels through the effects of AECA to hepatoportal sclerosis of IPH was elucidated in this study.

Disclosure

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References

None of the authors has any conflicts of interest to report.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References

Supported by the Japanese Study Group of Intrahepatic Hemodynamics Alterations (Chairman; Professor Fuminori Moriyasu, Tokyo Medical School, Tokyo, Japan).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Disclosure
  8. Acknowledgements
  9. References