Abstract

Background. Implantation of a haemodialysis arteriovenous graft is often followed by the development of neointimal hyperplasia (NH) at the venous anastomosis. The nature of the proliferating cells in these lesions is not well understood. A better understanding of the cells contributing to NH is important to the development of preventive strategies.

Methods. Carotid-jugular PTFE grafts were placed in 21 pigs and characterized at various time points following implantation. Venous anastomotic tissues were harvested at 1, 7, 14, 21, 28 or 49 post-operative days for histology and immunohistochemistry.

Results. Van Gieson staining of the tissues showed that NH was apparent as early as day 7 and progressed over time. Even by day 1, there were cells expressing the proliferation marker Ki-67 in the venous adventitia, but not the media, at the anastomosis. Double immunohistochemical staining showed that these cells were positive for α-smooth muscle actin (α-SMA), but negative for smooth muscle myosin heavy chain (SM MHC), suggesting that the proliferating cells were myofibroblasts rather than smooth muscle cells. By day 7, proliferating cells were abundant in the adventitia and began to appear in the media, surrounded by extracellular matrix visualized using Trichrome staining. By day 49, α-SMA-positive, SM MHC-negative cells were predominant in the NH, and Ki-67 staining had largely vanished.

Conclusions. These results are consistent with the hypothesis that adventitial fibroblasts are transformed into myofibroblasts and begin to proliferate within hours after graft placement. Migration of these cells towards the vessel lumen with subsequent proliferation appears to be a major contributor to NH formation. The pivotal role of the adventitial fibroblasts in the pathogenesis of NH provides a compelling rationale for therapies that target the transformation, proliferation and migration of these cells to prevent arteriovenous graft stenosis.

Introduction

While a well-functioning native fistula is preferred, maturation of fistulas into a functional conduit is sometimes a limiting factor. The arteriovenous (AV) graft is still commonly used as vascular access for haemodialysis patients, especially in the US. Grafts are, however, prone to failure caused by stenosis and occlusion, occurring at a rate of 50% at 1 year and 75% at 2 years after implantation [ 1 ]. Neointimal hyperplasia (NH) is by far the predominant cause of the stenosis and is particularly pronounced at the graft-venous anastomosis. At present, the pathogenesis of NH is not well understood and there are no effective strategies to prevent these lesions.

The proliferation and migration of medial smooth muscle cells (SMCs) have traditionally been considered to be the primary events in the pathogenesis of NH [ 2–4 ]. This notion has been supported by the observation that venous NH in human haemodialysis AV grafts are populated by cells that express α-smooth muscle actin (α-SMA), a marker that is usually employed to identify SMCs [ 5 ]. A small number of studies, however, have shown that vascular injury can result in significant remodelling of the adventitia, accompanied by the differentiation of adventitial fibroblasts into myofibroblasts, which acquire smooth muscle-like properties such as the expression of α-SMA [ 6 , 7 ]. Recent data from a porcine carotid artery injury model and autologous vein bypass graft model in the rat also suggest that the adventitia is a cellular source for the NH [ 8 , 9 ]. Very little is known about the cellular sources that contribute to the NH associated with haemodialysis AV synthetic grafts and their longitudinal dynamics during the course of the NH development, which were the focus of the present study.

Subjects and methods

Animal model

The porcine model of synthetic AV graft stenosis employed in our laboratory has been previously described [ 10 ]. Three-month-old Yorkshire cross domestic swine weighing approximately 30 kg were used in this study. All animal procedures and care were performed in accordance with the ‘Principles of Laboratory Animal Care’ and the ‘Guidelines for the Care and Use of Laboratory Animals’ (NIH Publication No. 85-23, revised 2001) and were approved by the Institutional Animal Care and Use Committee of the University of Utah and the Veterans Affairs Salt Lake City Healthcare System.

Animals were anaesthetized with a combination of xylazine (4 mg/kg), telazol (4 mg/kg) and ketamine (4 mg/kg). The trachea was then intubated and general anaesthesia was maintained using 1–3% isoflurane. Under sterile conditions, expanded polytetrafluoroethylene (PTFE) grafts (spiral re-enforced, 7-cm length, 6-mm internal diameter; Bard Peripheral Vascular Inc., Tempe, AZ) were placed between the common carotid artery and ipsilateral external jugular vein bilaterally. A total of 100–150 units/kg sodium heparin was given during the surgery. Starting 6 days before the graft surgery, aspirin EC (Pharmaceutical Formulations, Edison, NJ) at 81 mg/day was administrated orally. Clopidogrel (Bristol-Myers Squibb, New York, NY) at 225 mg was administered 1 day before surgery and continued at 75 mg/day post-operatively. Both these anti-platelet agents were continued until the time of euthanasia. Enrofloxacin (Baytril, Bayer, Pittsburgh, PA) was administrated subcutaneously at 5 mg/kg/day starting on the day of surgery and was continued for 3 days. Graft patency was monitored weekly by Doppler ultrasound (SonoSite, Bothell, WA) using an L38/10-5 MHz transducer (TITAN, SonoSite).

Study design and graft explantation

In order to follow the longitudinal cellular changes during NH development, the animals were euthanized at various time points: day 1 ( n = 2), day 7 ( n = 3), day 14 ( n = 2), day 21 ( n = 7), day 28 ( n = 5) or day 49 ( n = 2) after graft implantation. All grafts were patent at the time of euthanasia. The animals were anaesthetized as described previously for graft placement surgery. After the administration of 200 units/kg of heparin, the carotid artery was cannulated, through which the graft and adjacent vessels were perfused with saline for 3 min. Following a lethal intravenous injection of pentobarbital (100 mg/kg), the grafts were perfused with 10% zinc formalin (Fisher Scientific, Pittsburgh, PA) at a physiological pressure of 100 mm Hg. After 2 min, the proximal and distal segments of the artery and the vein around the anastomoses were ligated, allowing fixation of the vessels for an additional 5 min. Subsequently, the entire graft, together with attached artery and vein segments, was excised en bloc and immersed in 10% zinc formalin for at least 24 h.

Tissue preparation and morphometric analysis

The fixed graft and jugular vein segments were cut cross-sectionally into 5-mm blocks and embedded in paraffin. Five-micron thick sections were prepared from the block obtained at the centre of the anastomosis ( Figure 1 ). Elastin von Gieson (EvG) stain (Richard Allan Scientific, Kalamazoo, MI) was used to stain the histological slides for morphometric analysis. Quantification of surface areas of interest was performed using computer-assisted planimetry of images captured with a colour digital camera (Nikon, DS-L1) attached to a light microscope. Under the highest magnification that allowed visualization of the entire vein section in a single field, the area of the neointima, media and graft were manually traced. Two parts of the neointimal area were distinguished, the shoulder region that encompassed the neointima covering the graft and the cushioning region that encompassed the neointima located at the venous portion of the anastomosis ( Figure 1 ).

Fig. 1.

Location of histological sections from the graft-venous anastomosis for morphometric and immunohistochemical analysis (as presented in Figures 2–6). Panel A . Characterization of the gross anatomy of the graft-venous anastomosis. V is the location of the cross-section at the center of the anastomosis presented in panel B. Panel B . Histological cross section of the venous anastomosis from the location marked as V in panel A. Neointimal hyperplasia (NH) is found in the lumen of the graft (shoulder region) and the venous wall below the graft (cushioning region).

Fig. 1.

Location of histological sections from the graft-venous anastomosis for morphometric and immunohistochemical analysis (as presented in Figures 2–6). Panel A . Characterization of the gross anatomy of the graft-venous anastomosis. V is the location of the cross-section at the center of the anastomosis presented in panel B. Panel B . Histological cross section of the venous anastomosis from the location marked as V in panel A. Neointimal hyperplasia (NH) is found in the lumen of the graft (shoulder region) and the venous wall below the graft (cushioning region).

An NH index was defined as the neointimal area within the native venous wall (cushion region) and graft (shoulder region) divided by the sum of the medial area and graft area. In essence, this index was designed to normalize the observed NH relative to the normal vascular wall and the fixed size of the graft wall (nominally 0.6 mm in thickness), thus correcting for any distortion of the image produced by deviation from perfect perpendicular sectioning of the tissues. Masson's trichrome stain was used to highlight extracellular matrices. Normal carotid arteries and jugular veins from pigs without graft implantation ( n = 2) were obtained and prepared similarly for comparison.

Immunohistochemical analysis

Serial sections from the centre of the venous anastomosis obtained from each graft explanted at 1, 7, 14, 21, 28 or 49 post-operative days were stained with the following antibodies obtained commercially: anti-α-SMA, anti-smoothelin, anti-smooth muscle myosin heavy chain (SM MHC), anti-von Willebrand factor, and anti-Ki-67. Details regarding these antibodies are shown in Table 1 .

Table 1.

Antibodies used for immunohistochemical staining

Antibody Manufacturer Clone Dilution Cell specificity 
Ki-67 Dako MIB-1 1:100 Proliferating cell 
α-smooth muscle actin Cell marque 1A4 1:100 Smooth muscle cell, myofibroblast, myoepithelial cell 
Smoothelin Chemicon Polyclonal 1:60 Fully differentiated smooth muscle cell 
Smooth muscle myosin heavy chain Chemicon SM-M5 1:200 Fully differentiated smooth muscle cell 
von Willebrand factor Dako Polyclonal 1:600 Endothelial cell 
Antibody Manufacturer Clone Dilution Cell specificity 
Ki-67 Dako MIB-1 1:100 Proliferating cell 
α-smooth muscle actin Cell marque 1A4 1:100 Smooth muscle cell, myofibroblast, myoepithelial cell 
Smoothelin Chemicon Polyclonal 1:60 Fully differentiated smooth muscle cell 
Smooth muscle myosin heavy chain Chemicon SM-M5 1:200 Fully differentiated smooth muscle cell 
von Willebrand factor Dako Polyclonal 1:600 Endothelial cell 

A standard automated biotin-free Universal HRP-Polymer technique (BioCare Nemesis 7200 automated immunostainer, Concord, CA) was used for the immunostaining. In brief, following deparaffinization and hydration, the slides were washed and subjected to heat-induced epitope retrieval. The slides were then incubated sequentially with the primary antibody for 30 min, a murine secondary antibody for 10 min and with the HRP-Polymer for 10 min. All incubations were performed at room temperature with appropriate washes between each step. The slides were then developed with a diaminobenzidine/hydrogen peroxide mixture for 5 min, counterstained with haematoxylin, dehydrated with graduated alcohol and xylene and mounted using a xylene-based medium. Using this immunohistochemical technique, structures with the targeted epitope appeared rust-brown in colour. Negative controls were performed in parallel in each run by substituting the primary antibody with a non-specific rabbit polyclonal or murine monoclonal antibody, as appropriate. In addition, tissues from the gut, lymph node or tonsil were used as positive controls.

For double-label immunohistochemistry, Ki-67 staining was first performed as described previously. Then the slides were washed and exposed to the second primary antibody, which was either anti-α-SMA (1:100 dilution) or anti-SM MHC (1:200 dilution) for 30 min. An alkaline phosphatase-labelled polymer was then applied for 30 min and developed with Vulcan Fast Red for 15 min, which produced a red appearance in positive-staining cells.

Proliferation index

For each time point, Ki-67-stained sections from each venous anastomosis were used to assess the proliferation index. Ki-67-positive cells and Ki-67-negative cells were manually counted in four different randomly chosen fields within the native vessel (cushioning region) under light microscopy at 400× magnification in each layer (adventitia, media and neointima). A mean value was calculated from the four measurements in each layer. The proliferation index was calculated as the mean number of Ki-67-positive cells divided by the mean total number of cells and expressed as a percentage.

Histological scoring

The intensity of immunohistochemical staining for the cellular markers (α-SMA, SM MHC and smoothelin) was scored by three observers using a semi-quantitative scale from 0 (no expression) to 5 (intense expression).

Statistical analysis

Numerical data are presented as mean ± SD. Stat View (version 4.57) was used for all statistical analyses. To ascertain the significance of differences in the proliferation index over time, single-factor ANOVA was performed; each time point was tested against all other time points using Student's t -test with confidence limits modified by the method of Bonferroni. Linear regression was used to assess the change in NH index over time. A P -value <0.05 was considered to be statistically significant.

Results

Progression of neointimal hyperplasia at venous anastomosis

Histological sections obtained from the venous anastomosis of animals sacrificed at various time points from 1 day to 49 days after graft implantation revealed the presence of NH as early as day 7 ( Figure 2 ). The NH index, representing the normalized NH surface area, increased linearly ( r = 0.78; P < 0.0001) with time over this period ( Figure 3 B).

Fig. 2.

Morphological changes in the venous wall and the development of neointimal hyperplasia (NH) over time. The animals were sacrificed at the various time points as indicated above each panel and the histological sections obtained from the explanted anastomosis between the graft (G) and the vein were stained using the Elastin van Gieson stain. The images were obtained at 30 times magnification. These panels show that NH was already observed by day 7 and increased with time. No NH was observed at day 1 (not shown)

Fig. 2.

Morphological changes in the venous wall and the development of neointimal hyperplasia (NH) over time. The animals were sacrificed at the various time points as indicated above each panel and the histological sections obtained from the explanted anastomosis between the graft (G) and the vein were stained using the Elastin van Gieson stain. The images were obtained at 30 times magnification. These panels show that NH was already observed by day 7 and increased with time. No NH was observed at day 1 (not shown)

Fig. 3.

(A) Proliferation index (PI) and (B) neointimal hyperplasia (NH) index at the venous anastomosis over time. Tissues around the venous anastomosis were obtained from pigs sacrificed at various time points post-operatively (n 2, 3, 2, 7, 5, 2 at day 1, 7, 14, 21, 28 or 49, respectively). The surface areas of the neointima in the native vessel (cushion region) and inside the graft (shoulder region), the media and the graft were individually quantified using computer-assisted planimetry on Elastin von Gieson-stained slides. (A) The PI was calculated as the number of proliferating Ki-67-positive cells found in the native vessel wall in a given microscopic field divided by the total number of cells in that same field at 400x magnification. The PI in the adventitia (broken line) was high in the first 7 days and decreased thereafter (P < 0.0001, day 14, 21, 28 and 49 vs the respective value at day 1), while the PI in the neointima (solid line) became apparent only after day 14 (P < 0.0001, day 21, 28 and 49 vs the respective value at day 14). Each data point represents the mean PI SD. (B) The NH index was calculated as the neointimal area divided by the sum of the medial area and graft area. Each bar represents the mean NH index SD. NH index increased linearly with time (r = 0.78; P < 0.0001).

Fig. 3.

(A) Proliferation index (PI) and (B) neointimal hyperplasia (NH) index at the venous anastomosis over time. Tissues around the venous anastomosis were obtained from pigs sacrificed at various time points post-operatively (n 2, 3, 2, 7, 5, 2 at day 1, 7, 14, 21, 28 or 49, respectively). The surface areas of the neointima in the native vessel (cushion region) and inside the graft (shoulder region), the media and the graft were individually quantified using computer-assisted planimetry on Elastin von Gieson-stained slides. (A) The PI was calculated as the number of proliferating Ki-67-positive cells found in the native vessel wall in a given microscopic field divided by the total number of cells in that same field at 400x magnification. The PI in the adventitia (broken line) was high in the first 7 days and decreased thereafter (P < 0.0001, day 14, 21, 28 and 49 vs the respective value at day 1), while the PI in the neointima (solid line) became apparent only after day 14 (P < 0.0001, day 21, 28 and 49 vs the respective value at day 14). Each data point represents the mean PI SD. (B) The NH index was calculated as the neointimal area divided by the sum of the medial area and graft area. Each bar represents the mean NH index SD. NH index increased linearly with time (r = 0.78; P < 0.0001).

Cellular proliferation

Proliferating cells at the venous anastomosis were identified by their expression of the Ki-67 antigen. Minimal reactivity to anti-Ki-67 antibody was observed in control jugular veins obtained from normal pigs without graft implantation (data not shown). In the pigs with graft implantation, however, the proliferation index in the adventitia of the native vessel wall at the venous anastomosis was 96.1% ± 2.2% at day 1, indicating that practically all cells in that layer were stimulated to proliferate within hours after graft implantation ( Figure 3 A). The proliferation index in the adventitia remained high on day 7, but began to significantly decline thereafter ( Figure 3 A). On day 7, Ki-67-positive cells began to appear in the media ( Figure 3 A), accompanied by prominent extracellular matrix deposition seen using the Trichrome stain ( Figure 4 : panel 4c). The proliferation index in the media remained 13–15% between day 7 to day 28 ( Figure 3 A). In contrast, Ki-67-positive cells were undetectable in the NH until after day 14, after which the proliferation index increased progressively until day 28 ( Figure 3 A).

Identification of cell types ( Table 2 )

The presence of α-SMA is typically used as a marker for SMCs, but its expression can be seen in other cell types [ 11 ]. α-SMA is also expressed by myofibroblasts during wound healing and by endothelial cells during vascular remodelling [ 12 ]. To clarify the cellular sources of NH, we employed antibodies directed against SM MHC and smoothelin, which are markers reported to be specific for differentiated, mature SMCs [ 12 ].

Table 2.

Semiquantitative scoring of cell marker expression at the venous anastomosis (scale 0 to 5)

 α-smooth muscle actin Smooth muscle myosin heavy chain Smoothelin 
Venous adventitia    
    Normal vein 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 
    1 day 1.7 ± 0.6 1.2 ± 0.8 2.3 ± 1.0 
    7 day 4.3 ± 0.6 1.2 ± 0.8 2.7 ± 1.0 
    28 day 4.3 ± 0.6 1.0 ± 0.5 1.2 ± 0.8 
    49 day 0.8 ± 0.3 1.3 ± 0.3 2.0 ± 0.9 
Venous media    
    Normal vein 5.0 ± 0.0 5.0 ± 0.0 5.0 ± 0.0 
    1 day 4.8 ± 0.3 5.0 ± 0.0 5.0 ± 0.0 
    7 day 4.7 ± 0.6 4.8 ± 0.3 4.7 ± 0.6 
    28 day 3.5 ± 0.9 4.2 ± 0.3 3.5 ± 0.9 
    49 day 4.3 ± 0.3 4.8 ± 0.3 3.5 ± 0.9 
Neointima, cushion region (vein)    
    28 day 5.0 ± 0.0 4.5 ± 0.5 3.5 ± 0.5 
    49 day 5.0 ± 0.0 5.0 ± 0.0 3.2 ± 0.3 
Neointima, shoulder region (graft)    
    28 day 5.0 ± 0.0 5.0 ± 0.0 3.7 ± 0.8 
    49 day 4.5 ± 0.5 3.5 ± 0.5 2.8 ± 0.8 
 α-smooth muscle actin Smooth muscle myosin heavy chain Smoothelin 
Venous adventitia    
    Normal vein 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 
    1 day 1.7 ± 0.6 1.2 ± 0.8 2.3 ± 1.0 
    7 day 4.3 ± 0.6 1.2 ± 0.8 2.7 ± 1.0 
    28 day 4.3 ± 0.6 1.0 ± 0.5 1.2 ± 0.8 
    49 day 0.8 ± 0.3 1.3 ± 0.3 2.0 ± 0.9 
Venous media    
    Normal vein 5.0 ± 0.0 5.0 ± 0.0 5.0 ± 0.0 
    1 day 4.8 ± 0.3 5.0 ± 0.0 5.0 ± 0.0 
    7 day 4.7 ± 0.6 4.8 ± 0.3 4.7 ± 0.6 
    28 day 3.5 ± 0.9 4.2 ± 0.3 3.5 ± 0.9 
    49 day 4.3 ± 0.3 4.8 ± 0.3 3.5 ± 0.9 
Neointima, cushion region (vein)    
    28 day 5.0 ± 0.0 4.5 ± 0.5 3.5 ± 0.5 
    49 day 5.0 ± 0.0 5.0 ± 0.0 3.2 ± 0.3 
Neointima, shoulder region (graft)    
    28 day 5.0 ± 0.0 5.0 ± 0.0 3.7 ± 0.8 
    49 day 4.5 ± 0.5 3.5 ± 0.5 2.8 ± 0.8 

The media of external jugular veins obtained from normal pigs without graft placement stained strongly positive for α-SMA, SM MHC and smoothelin. The staining was well organized in distinct layers ( Figure 4 : panels 1a, 2a and 3a). On day 7 after graft placement, there was an abundance of α-SMA-positive cells in the adventitia (4.3 ± 0.6; Figure 4 : panel 1c); however staining for SM MHC (1.2 ± 0.8; Figure 4 : panel 2c) and smoothelin (2.7 ± 1.0; Figure 4 : panel 3c) was faint, suggesting that the cells were myofibroblasts transformed from fibroblasts instead of pre-existing differentiated SMCs. Double-immunostaining confirmed that the cells in the adventitia expressing the Ki-67 antigen were positive for α-SMA but negative for SM MHC ( Figure 5 ), indicating that the proliferating cells in the adventitia were indeed myofibroblasts.

Fig. 4.

Expression of α-smooth muscle actin (α-SMA), smooth muscle myosin heavy chain (SM MHC) and smoothelin and extracellular matrix (ECM) at the venous anastomoses and normal vein. The control external jugular vein shows well-organized layers of smooth muscle cells expressing α-SMA (panel 1a), SM MHC (panel 2a) and smoothelin (panel 3a) appearing rust-brown. On day 7 after graft placement, the media at the venous anastomosis stained as intensely as the control for α-SMA (panel 1c), SM MHC (panel 2c) and smoothelin (panel 3c). In contrast, the adventitia also stained positive for α-SMA, but in a disorganized pattern (panel 1c) and it stained much less intensely for SM MHC (panel 2c) and smoothelin (panel 3c). By day 49, α-SMA expression was prominent and more uniform in the neointimal hyperplasia (NH in panel 1g), while staining for SM MHC (panel 2g) and smoothelin (panel 3g) was patchy, suggesting that most of the cells were derived from adventitial fibroblasts instead of smooth muscle cells. The control external jugular vein shows the most ECM (appearing blue by Trichrome staining) in the adventitia (panel 4a). By day 7, more cells (appearing pink) accumulated in the adventitia, accompanied by prominent ECM deposition in the media and inner layers of the adventitia (panel 4c). By day 49, ECM was prominent in the NH (panel 4g). The magnification was 100× in panels a–c and 50× in panel d–g.

Fig. 4.

Expression of α-smooth muscle actin (α-SMA), smooth muscle myosin heavy chain (SM MHC) and smoothelin and extracellular matrix (ECM) at the venous anastomoses and normal vein. The control external jugular vein shows well-organized layers of smooth muscle cells expressing α-SMA (panel 1a), SM MHC (panel 2a) and smoothelin (panel 3a) appearing rust-brown. On day 7 after graft placement, the media at the venous anastomosis stained as intensely as the control for α-SMA (panel 1c), SM MHC (panel 2c) and smoothelin (panel 3c). In contrast, the adventitia also stained positive for α-SMA, but in a disorganized pattern (panel 1c) and it stained much less intensely for SM MHC (panel 2c) and smoothelin (panel 3c). By day 49, α-SMA expression was prominent and more uniform in the neointimal hyperplasia (NH in panel 1g), while staining for SM MHC (panel 2g) and smoothelin (panel 3g) was patchy, suggesting that most of the cells were derived from adventitial fibroblasts instead of smooth muscle cells. The control external jugular vein shows the most ECM (appearing blue by Trichrome staining) in the adventitia (panel 4a). By day 7, more cells (appearing pink) accumulated in the adventitia, accompanied by prominent ECM deposition in the media and inner layers of the adventitia (panel 4c). By day 49, ECM was prominent in the NH (panel 4g). The magnification was 100× in panels a–c and 50× in panel d–g.

Fig. 5.

Representative photomicrographs of double-staining of adventitia at the venous anastomosis explanted on day 7. The Ki-67-positive (rust-brown) cells were also positive for α-SMA (appearing red in panel A with further magnification in panel B ), but negative for SM MHC (appearing red if positive in panel C with further magnification in panel D). These data confirmed that these proliferating cells were transformed myofibroblasts and not smooth muscle cells. The magnification was 100× in panels A and C and 400× in panels B and D .

Fig. 5.

Representative photomicrographs of double-staining of adventitia at the venous anastomosis explanted on day 7. The Ki-67-positive (rust-brown) cells were also positive for α-SMA (appearing red in panel A with further magnification in panel B ), but negative for SM MHC (appearing red if positive in panel C with further magnification in panel D). These data confirmed that these proliferating cells were transformed myofibroblasts and not smooth muscle cells. The magnification was 100× in panels A and C and 400× in panels B and D .

From day 28 to day 49, α-SMA-positive cells predominated in the neointima. Concomitant expression of SM MHC and smoothelin in serial slides from the same tissue bloc was seen in certain areas of the NH at these time points, indicating these cells were likely mature SMC ( Figure 4 : panels 1f–3f; 1g–3g). All other areas of the neointima, however, showed significant staining with α-SMA but little staining of smoothelin or SM MHC, indicating the presence of myofibroblasts ( Figure 4 : panels 1f–3f; 1g–3g). The extracellular matrix stained by Trichrome in the neointima increased over time, and by day 28 became one of the major components of the neointima ( Figure 4 : panels 4a–4g). Taken together, these data suggest that extracellular matrix, myofibroblasts and differentiated SMC are the major components of NH. The intensities of the three markers (α-SMA, SM MHC and smoothelin) were scored using a semi-quantitative scale from 0 to 5. The scoring results are shown in Table 2 .

Anti-von Willebrand factor was used to detect endothelial cells. At none of the time points studied was there evidence of disruption of the endothelial monolayer in the native vascular wall ( Figure 6 C and D), although there was disruption of the endothelium around the suture line on days 1 and 7 (data not shown). Neovascularization with the expression of von Willebrand factor was, however, apparent in the adventitia and neointima after 7 post-operative days ( Figure 6 E–H). Ki-67 expression within the endothelial cells lining microvessels in the adventitia and neointima was observed occasionally (data not shown).

Fig. 6.

Staining for von Willebrand factor at the venous anastomosis ( CH ) and normal vein ( A and B ). Positive staining appears rust-brown. Panels C and D demonstrate the preservation of endothelial integrity in venous wall around the anastomosis even immediately after surgical trauma. Note the positive staining for neovasculature (black arrows) in the adventitia at day 7 (panels E and F ) and in the region of neointimal hyperplasia (NH) at day 49 (panels G and H ).

Fig. 6.

Staining for von Willebrand factor at the venous anastomosis ( CH ) and normal vein ( A and B ). Positive staining appears rust-brown. Panels C and D demonstrate the preservation of endothelial integrity in venous wall around the anastomosis even immediately after surgical trauma. Note the positive staining for neovasculature (black arrows) in the adventitia at day 7 (panels E and F ) and in the region of neointimal hyperplasia (NH) at day 49 (panels G and H ).

Discussion

Understanding the identity and origin of cells that populate the neointima is important from both a pathogenesis and therapeutic standpoint. Historically, it has been thought that SMCs from the media are the predominant cells involved in the development of NH [ 3 , 13 ]. The potential role of the adventitia in the pathogenesis of NH and vascular remodelling has received attention only recently [ 9 , 14–16 ].

In the present study, we provide evidence for the involvement of adventitial cells in the pathogenesis of NH associated with haemodialysis PTFE AV grafts in a porcine model and a detailed description of the spatial and temporal changes of the proliferating cells during this process. The results showed that cell proliferation activity started in the adventitia of the venous anastomosis, which occurred as early as day 1 after graft implantation ( Figure 3 ). Although NH was readily seen by 1 week, there was no demonstrable proliferative activity in the neointima on and before 14 days. On day 7, proliferating cells began to appear in the media but there are not many cells proliferating in the media ( Figure 3 A). These observations indicate that cell migration from the adventitia towards the lumen is the major contributor to NH early in the course in the AV graft model. In the ensuing 2 weeks, however, the proliferation index in the neointima increased linearly and correlated well with the increase in the NH index, while the proliferation index in the adventitia diminished markedly. These observations indicate that cell proliferation in the neointima is the critical determinant of NH progression only in the later phases of NH development.

In addition to the anatomical origin of the proliferating and migrating cells, we have also determined the identity of these cells by immunohistochemical techniques. The predominant cell type in the tunica adventitia is normally the fibroblast which does not express α-SMA [ 16 ], consistent with what was observed in our control jugular veins ( Figure 4 ). Abundant α-SMA-positive cells, however, were observed in the adventitia at the venous anastomosis of the graft on day 7 after implantation in our model, suggesting that fibroblasts were transformed into α-SMA-expressing myofibroblasts soon after graft implantation. α-SMA has been the traditionally accepted SMC-specific marker [ 11 ], but the recent literature suggests that myofibroblasts also express α-SMA under pathological conditions [ 12 ]. Similarities between myofibroblasts and SMCs in their ultrastructural characteristics and the expression of cytoskeletal protein markers such as α-SMA have made distinguishing between these cell types particularly problematic. The expression of SM MHC is now considered the most discriminating marker for SMC identification, with smoothelin being another marker that is relatively specific for SMC [ 12 ]. Double-staining experiments showed that the adventitial cells expressing the Ki-67 antigen soon after graft placement co-localized with α-SMA, but not with SM MHC or smoothelin, indicating that these proliferating cells were indeed myofibroblasts, and not SMCs.

Taken together, these results support the following scenario for NH development following PTFE AV graft placement. Within hours after AV graft placement, adventitial fibroblasts are activated and transform into myofibroblasts through mechanisms which have yet to be elucidated. These adventitial myofibroblasts then proliferate, migrate towards the vessel lumen, whereupon they deposit extracellular matrix and serve as the primary contributors to the formation of NH. In the neointima, there was positive staining for smoothelin and SM MHC in a patchy distribution beyond 4 weeks, although α-SMA staining was globally more uniform and stronger. These data suggest that, while the myofibroblast was the predominant cell type in the NH, differentiated contractile SMC was also a constituent of the lesion. Whether these SMC in the NH were derived from the SMC innate to the vessel media or further transformed from myofibroblasts could not be determined from these data.

A limitation of the present study is that the evidence for adventitial myofibroblast migration to the intima is indirect. In other models of NH development, there was more direct evidence of cell migration from the adventitia to the intima. Using the bromodeoxyuridine (BrdU) pulse-labelling technique in a porcine model of coronary artery balloon injury, Shi et al . [ 17 ] demonstrated that the adventitia contained numerous BrdU-labelled cells at 2–3 days. At 7–8 days, some labelled cells acquired the characteristics of myofibroblasts expressing α-SMA, translocated to the gap between dissected medial layers and contributed to the formation of neointima. By 18–35 days after injury, labelled cells were abundant in the neointima. In a porcine model of autologous saphenous vein interposition graft to the common carotid artery, perivascular fibroblasts were also shown to contribute to NH development using the BrdU pulse-labelling technique [ 18 ].

While these studies demonstrated that the proliferating cells in fact migrated from the adventitia to the intima, there were several limitations to these studies. First, the labelling of cells was limited to a single time point (12 and 24 h after injury), which could miss the cells that entered the cell cycle at a later time. Second, proliferating cells rapidly lose the BrdU label, which makes it difficult to identify the source of proliferating cells in the later stages of NH development. Recently, using the same technique in a porcine iliac artery to ipsilateral iliac vein PTFE graft model, Misra et al . [ 19 ] reported that BrdU immunostaining was present at the adventitia-media junction by day 7, and was seen primarily in the intima by day 26 after graft implantation. This study provided the first evidence of adventitial cell involvement in venous stenosis formation in the AV graft model. The proliferating cell type was, however, not identified in that study and the proliferating cells were only examined long (26 days) after graft placement. Our present results expand upon their findings in an AV graft model with more detailed cell characterization, as well as an earlier (day 1) and a longer (49 days) analysis after graft placement.

Cell transfection techniques have also been used to trace the migration of cells during NH development in other model systems. In a balloon-injured rat carotid artery model, primary adventitial fibroblasts were transfected ex vivo with the β-galactosidase-expressing (LacZ) retrovirus and introduced into the carotid artery immediately after injury [ 9 ]. These cells were later found to be present along the entire thickness of the vessel wall from the adventitia to the neointima. In a different rat model of NH, the external surface of the autologous epigastric vein was transfected with a β-galactosidase-expressing adenovirus before placement as an interposition graft to the femoral artery [ 8 ]. The two ends of the vein graft were ligated in order to prevent viral transfection of the luminal cells. The transfected cells were found to migrate through the wall of the venous graft to the neointima. This technique provides direct evidence for the migration of adventitial cells into the neointima after injury, but cannot identify the precise cell type. There is no report using this technique in the AV graft model because of the difficulty in transfecting only the outer layer of the native vessel wall without explanting the vein or contaminating the luminal cells. Although each of the techniques and models for characterizing the cell types involved in NH development has limitations, taken together with the results of the present study, there is a compelling body of evidence indicating that adventitial fibroblasts that transform to myofibroblasts are likely to be the primary source of cells that eventually give rise to NH following AV graft placement, as well as in other NH models.

The participation of adventitial fibroblasts in the pathogenesis of NH in haemodialysis AV grafts described in the present study has potentially important therapeutic implications. The systemic administration of anti-proliferative agents for the prevention of NH is hampered by the associated side effects related to the inhibition of proliferation and cytotoxicity at non-target sites. Because the activation of adventitial fibroblasts appears to be a critical pathogenic event, the local perivascular delivery of anti-proliferative agents in close proximity to the adventitia should be an effective strategy to prevent NH associated with haemodialysis AV grafts with minimal systemic side effects. The perivascular sustained delivery of small-molecule drugs using polymeric platforms is currently being developed for this specific purpose [ 10 , 20–23 ].

Acknowledgements

This work was supported by the National Heart, Lung and Blood Institute (RO1HL67646), Medical and Research Services of the Department of Veterans Affairs, Dialysis Research Foundation and the National Kidney Foundation of Utah and Idaho. PTFE grafts were kind gifts of Bard Peripheral Vascular, Inc.

Conflict of interest statement . None declared.

References

1
US Renal Data System
2004 USRDS Annual Report
2004
Minneapolois
United States Renal Data System
2
Austin
GE
Ratliff
NB
Hollman
J
, et al.  . 
Intimal proliferation of smooth muscle cells as an explanation for recurrent coronary artery stenosis after percutaneous transluminal coronary angioplasty
J Am Coll Cardiol
 , 
1985
, vol. 
6
 (pg. 
369
-
375
)
3
Schwartz
SM
deBlois
D
O’Brien
ER
The intima. Soil for atherosclerosis and restenosis
Circ Res
 , 
1995
, vol. 
77
 (pg. 
445
-
465
)
4
Roy-Chaudhury
P
Kelly
BS
Miller
MA
, et al.  . 
Venous neointimal hyperplasia in polytetrafluoroethylene dialysis grafts
Kidney Int
 , 
2001
, vol. 
59
 (pg. 
2325
-
2334
)
5
Kelly
BS
Heffelfinger
SC
Whiting
JF
, et al.  . 
Aggressive venous neointimal hyperplasia in a pig model of arteriovenous graft stenosis
Kidney Int
 , 
2002
, vol. 
62
 (pg. 
2272
-
2280
)
6
Shi
Y
Pieniek
M
Fard
A
, et al.  . 
Adventitial remodeling after coronary arterial injury
Circulation
 , 
1996
, vol. 
93
 (pg. 
340
-
348
)
7
Scott
NA
Cipolla
GD
Ross
CE
, et al.  . 
Identification of a potential role for the adventitia in vascular lesion formation after balloon overstretch injury of porcine coronary arteries
Circulation
 , 
1996
, vol. 
93
 (pg. 
2178
-
2187
)
8
Tomas
JJ
Stark
VE
Kim
JL
, et al.  . 
Beta-galactosidase-tagged adventitial myofibroblasts tracked to the neointima in healing rat vein grafts
J Vasc Res
 , 
2003
, vol. 
40
 (pg. 
266
-
275
)
9
Li
G
Chen
SJ
Oparil
S
, et al.  . 
Direct in vivo evidence demonstrating neointimal migration of adventitial fibroblasts after balloon injury of rat carotid arteries
Circulation
 , 
2000
, vol. 
101
 (pg. 
1362
-
1365
)
10
Kuji
T
Masaki
T
Goteti
K
, et al.  . 
Efficacy of local dipyridamole therapy in a porcine model of arteriovenous graft stenosis
Kidney Int
 , 
2006
, vol. 
69
 (pg. 
2179
-
2185
)
11
Owens
GK
Regulation of differentiation of vascular smooth muscle cells
Physiol Rev
 , 
1995
, vol. 
75
 (pg. 
487
-
517
)
12
Owens
GK
Kumar
MS
Wamhoff
BR
Molecular regulation of vascular smooth muscle cell differentiation in development and disease
Physiol Rev
 , 
2004
, vol. 
84
 (pg. 
767
-
801
)
13
Davies
MG
Hagen
PO
Pathobiology of intimal hyperplasia
Br J Surg
 , 
1994
, vol. 
81
 (pg. 
1254
-
1269
)
14
Wilcox
JN
Cipolla
GD
Martin
FH
, et al.  . 
Contribution of adventitial myofibroblasts to vascular remodeling and lesion formation after experimental angioplasty in pig coronary arteries
Ann N Y Acad Sci
 , 
1997
, vol. 
811
 (pg. 
437
-
447
)
15
Zalewski
A
Shi
Y
Vascular myofibroblasts. Lessons from coronary repair and remodeling
Arterioscler Thromb Vasc Biol
 , 
1997
, vol. 
17
 (pg. 
417
-
422
)
16
Sartore
S
Chiavegato
A
Faggin
E
, et al.  . 
Contribution of adventitial fibroblasts to neointima formation and vascular remodeling: from innocent bystander to active participant
Circ Res
 , 
2001
, vol. 
89
 (pg. 
1111
-
1121
)
17
Shi
Y
O’Brien
JE
Fard
A
, et al.  . 
Adventitial myofibroblasts contribute to neointimal formation in injured porcine coronary arteries
Circulation
 , 
1996
, vol. 
94
 (pg. 
1655
-
1664
)
18
Shi
Y
O’Brien
JE
Jr
Mannion
JD
, et al.  . 
Remodeling of autologous saphenous vein grafts. The role of perivascular myofibroblasts
Circulation
 , 
1997
, vol. 
95
 (pg. 
2684
-
2693
)
19
Misra
S
Doherty
MG
Woodrum
D
, et al.  . 
Adventitial remodeling with increased matrix metalloproteinase-2 activity in a porcine arteriovenous polytetrafluoroethylene grafts
Kidney Int
 , 
2005
, vol. 
68
 (pg. 
2890
-
2900
)
20
Masaki
T
Rathi
R
Zentner
G
, et al.  . 
Inhibition of neointimal hyperplasia in vascular grafts by sustained perivascular delivery of paclitaxel
Kidney Int
 , 
2004
, vol. 
66
 (pg. 
2061
-
2069
)
21
Goteti
K
Masaki
T
Kuji
T
, et al.  . 
Perivascular tissue pharmacokinetics of dipyridamole
Pharm Res
 , 
2006
, vol. 
23
 (pg. 
718
-
728
)
22
Zhu
W
Masaki
T
Bae
YH
, et al.  . 
Development of a sustained-release system for perivascular delivery of dipyridamole
J Biomed Mater Res B Appl Biomater
 , 
2006
, vol. 
77
 (pg. 
135
-
143
)
23
Melhem
M
Kelly
B
Zhang
J
, et al.  . 
Development of a local perivascular paclitaxel delivery system for hemodialysis vascular access dysfunction: polymer preparation and in vitro activity
Blood Purif
 , 
2006
, vol. 
24
 (pg. 
289
-
298
)

Author notes

*This work was previously presented in free communication form at the American Society of Nephrology Annual Meeting, in San Diego, California on November 18, 2006.

Comments

0 Comments