Beta-Galactosidase-Tagged Adventitial Myofibroblasts Tracked to the Neointima in Healing Rat Vein GraftsTomas J.J.a · Stark V.E.a · Kim J.L.a · Wolff R.A.a · Hullett D.A.a · Warner T.F.b · Hoch J.R.a
Departments of aSurgery and bPathology, University of Wisconsin School of Medicine, Madison, Wisc., USA
Objective: Myofibroblasts are present transiently in normal healing wounds. However, they have been found to persist in the stroma of neoplasms, fibrotic conditions and other pathological settings. In rat vein grafts, we have observed the prolonged presence of myofibroblasts. Our aim was to determine the origin of myofibroblasts in vein grafts. Methods: Epigastric vein to femoral artery grafts were microsurgically placed in male Lewis rats and harvested. Neointimal development, cellular death and proliferation, and cell phenotypes were analyzed using immunohistochemistry and light and electron microscopy. To follow cellular movement in the vessel wall, vein grafts were transfected with replication-defective adenovirus containing the gene encoding β-galactosidase (n = 50), and harvested at 1, 2, 3, 4, 5, 6, 7, 14 and 28 days. Grafts were analyzed after X-gal staining. Results: Myofibroblasts were detected in the outer adventitia at 4 days, in the media at 1 week and in the developing neointima at 2 weeks. Cells tagged using adenoviral β-galactosidase demonstrated adventitia to neointima cell migration. Conclusions: Although there may be other sources of myofibroblasts in this model, the adventitia has been shown to be an origin of myofibroblasts which subsequently migrate through the vessel wall to the neointima during graft remodeling and contribute to neointimal formation.
Copyright © 2003 S. Karger AG, Basel
Intimal hyperplasia (IH), a fibroproliferative thickening of the intima, is the primary cause of vein graft failure after the first postsurgical year [1, 2]. No treatments exist for IH, and its etiology is not fully understood. The process of IH development in vein grafts resembles arterial restenosis, a condition for which pharmacological interventions have not been successful [3, 4]. While the development of a neointima in IH may be viewed as vein remodeling in an attempt to adapt to the new arterial environment, IH shares important features with abnormal wound healing, including protracted kinetics, altered organization and composition of the extracellular matrix and the presence of poorly differentiated cells. In our studies using a rat vein graft model, we have observed these characteristics of aberrant wound healing.
The myofibroblast is one cell type that has been found to persist in abnormal wounds, the stroma of neoplasms, fibrotic conditions and other pathological settings [5, 6, 7]. Myofibroblasts are also present transiently in normal healing wounds . Myofibroblasts are actively synthetic cells, possessing abundant endoplasmic reticulum and Golgi bodies [9, 10].
We have previously described myofibroblastic cells in healing rat vein grafts . We hypothesize that one source of these cells is the transformation of adventitial fibroblasts which are activated by graft implantation and subsequently migrate towards the lumen to populate the developing neointima. Adventitial myofibroblasts have been detected in other vascular injury models [12, 13, 14, 15], a finding which supports their role in vascular remodeling and repair. In the following experiments, using a rat model of vein graft IH, we employed ultrastructural and biochemical techniques to investigate the phenotype and migration of myofibroblasts in healing vein grafts.
Materials and Methods
Male Lewis rats (350–450 g, 9–12 months old) were used in this study. Epigastric vein to common femoral artery interposition grafts were placed using aseptic microsurgical techniques as previously described . Briefly, each animal was anesthetized intraperitoneally with ketamine (90 mg/kg) and xylazine (5 mg/kg), and the epigastric vein and common femoral artery were dissected. An 8-mm segment of relaxed epigastric vein was excised, reversed and interposed into a 3-mm arterial defect. Each anastomosis was completed with 8 interrupted sutures of 10-0 Nylon. After 45 min of total ischemic time, blood flow was restored to the limb, patency was confirmed and the wound was irrigated and closed. The animal was closely observed until recovery. Grafts and contralateral nonoperated control veins were harvested 4 days and 1, 2, 4, 8 and 12 weeks after surgery. For detection of proliferation, 12 animals were injected with bromodeoxyuridine (BrdU; 50 mg/kg) 18 h prior to harvest. Animal care complied with the ‘Principles of Laboratory Animal Care’ (National Society for Medical Research) and the ‘Guide for the Care and Use of Laboratory Animals’ . Procedures were approved by the Research Animal Resources Center at the University of Wisconsin-Madison.
Thirty grafts and contralateral epigastric veins (n = 5 per time point) were harvested and processed for histological and morphometric analyses as previously described . The proximal region of the graft, which displays the earliest IH, was examined. Serial 5-μm sections were cut starting 1 mm distal to the proximal anastomosis. Two sections per graft were stained with hematoxylin and eosin and imaged with a videomicroscope. For each section, the neointimal and medial/adventitial areas and cellularity were calculated for the entire cross-section with three sets of counts performed in a blinded fashion.
Immunohistochemical staining was performed as previously described [18, 19]. Paraffin tissue sections serial to those used for histological analysis were incubated with monoclonal antibodies (10 μg/ml). Primary antibodies were specific for α-smooth muscle actin (asm-1, Boehringer Mannheim), smooth muscle myosin (hSM-V, Sigma), desmin (D33, Dako), BrdU (Bu20A, Dako) and proliferating cell nuclear antigen (PC10, Pharmingen). Staining for BrdU was preceded by proteolytic digestion of the tissue with 1 M HCl at 60°C for 8 min . Incubation with the primary antibodies was followed sequentially by incubations with a biotinylated goat anti-mouse secondary antibody (10 μg/ml, Dako), streptavidin-peroxidase (Dako) and the substrate 3′3′ diaminobenzidine (Vector). Some specimens were counterstained with hematoxylin. To ensure specificity of the antibody, parallel staining using isotype (IgG1 or IgG2A) matched control antibodies was performed. Rat femoral artery, spleen and small intestine were used as control tissues.
Twelve grafts (n = 2 per time point) and 2 nongrafted epigastric veins were harvested for transmission electron microscopy as previously described . Both thick (1-μm) and thin (900-Å) sections were cut and stained. Three sections from each segment were examined using a Hitachi H500 transmission electron microscope at various magnifications.
Ten grafts (n = 2 at 4 days, and 1, 2, 4 and 8 weeks after surgery) and 2 nonoperated epigastric veins were used for immunogold labeling experiments . Tissues were harvested, fixed in Carson-Millonig solution for 2 h, dehydrated in graded ethanol solutions, infiltrated with LR White resin (Polysciences), embedded and sectioned onto nickel grids (90-nm transverse sections) using a Sorvall MT2-B ultramicrotome with a diamond knife. Thick sections (0.3 mm) were also cut with a glass knife and stained with toluidine blue before and after thin sectioning. After preliminary titrations to determine the optimum antibody concentrations, thin sections were blocked with 2% BSA-0.1% Tween 20 and incubated at 4°C overnight in a humidified chamber with either the mouse monoclonal antibody asm-1, which recognizes α-smooth muscle actin (0.5 mg/ml; Boehringer Mannheim), or with mouse IgG2A (0.5 mg/ml) as the negative control. After washing with TBS (pH 8.2), sections were incubated with the colloidal gold-conjugated goat-anti-mouse IgG (10-nm particles, 1:75 dilution; BBI International) for 2 h at room temperature, and then stained with uranyl acetate for 10 min. Positive staining was visualized with a Philips 301 electron microscope at various magnifications.
Fifty animals were utilized for these experiments. Vein grafts were transfected with replication-defective adenovirus containing the SV40 nuclear targeting region and the gene encoding β-galactosidase, a generous gift of Dr. Tausif Alam, University of Wisconsin . To establish migration from the adventitia of the neointima (n = 5 with harvest at 14 days), the epigastric vein was prepared longer than required to repair the femoral artery resection. Each end was then ligated before the graft was submerged into the viral solution (1 × 107 PFU/ml for 15 min at room temperature). After rinsing the viral solution off the graft, the ligated ends were removed from the graft and the remainder of the graft was placed into the arterial defect according to standard laboratory microsurgical procedure. For the time course (n = 5 at 1, 2, 3, 4, 5, 6, 7, 14 and 28 days), an 8-mm segment of relaxed epigastric vein was excised, transfected, reversed and interposed into a 3-mm arterial defect. After harvest, tissue was fixed in 0.5% glutaraldehyde for 30 min, then placed into a solution containing the substrate X-gal (1 mg/ml; Sigma), incubated at 37°C for at least 4 h, snap-frozen and cryostat-sectioned. Controls showed that there was complete penetration of X-gal throughout the wall of the vein graft. The location and number of blue-stained positive cells were determined by digitized image analysis using a cooled CCD SenSys black and white camera (Photometrics).
Results are expressed as mean ± SEM. To determine whether derived morphometric values represented significant differences from nonoperated controls, multiple analyses of variance was performed, followed by Fisher’s protected least significant difference procedure.
To assess IH development, we examined vein graft sections for histological and morphometric changes. As expected, IH was induced in vein grafts (fig. 1a). All vein grafts displayed characteristics of IH: cellular proliferation and extracellular matrix deposition resulting in a thickened neointimal layer, fragmentation and duplication of the internal elastic lamina and marked medial attenuation and melding with the adventitia. Cellularity in the vein wall increased after grafting. Cell numbers in the media/adventitia peaked at 1–2 weeks and then decreased; neointimal cell number increased until 4 weeks and then remained constant (fig. 1b). All of these results agreed with our previous observations in this model [11, 16, 18, 19].
Fig. 1. IH developed in vein grafts after implantation. a Column chart demonstrating increasing neointimal area up to 12 weeks after grafting (p < 0.05 vs. epigastric vein control at 2–12 weeks); medial/adventitial areas were greater than epigastric vein controls from 1 to 12 weeks (p < 0.05), peaking at 2 weeks. Areas were derived by computerized morphometry. Each measurement was made twice and a mean for each graft was derived. The internal elastic lamina was traced to define the border separating the neointima from the media/adventitia. b Numbers of cells at 1–12 weeks are significantly increased as compared to control nongrafted epigastric vein (p = 0.0001 vs. epigastric vein control at 1–12 weeks for media/adventitia, and at 2–12 weeks for neointima). Total hematoxylin-stained nuclei per cross-section were counted and values were derived using computerized image analysis. Error bars represent SEM.
Using light microscopy, we detected the appearance of putative myofibroblasts in the graft wall after surgery. These cells were not seen in the nongrafted epigastric vein (fig. 2, top) or in control femoral artery, but appeared in the outer adventitia of grafted vein by 4 days (fig. 2, center) and at the internal elastic lamina by 1 week. They continued to be seen in the adventitia at all time points. Identical large cells appeared in the neointima by 1–2 weeks after grafting (fig. 2, bottom) and continued to be seen at later time points. In addition, we detected the loss of endothelial cells early after grafting (fig. 2, center). In contrast to the nongrafted epigastric vein (fig. 2, top), the medial layer of all 4-day grafts was acellular and compressed (fig. 2, center). In 4-day grafts, most medial cells were in various stages of degeneration and death, with disorganized microfilaments, nuclear fragmentation and vacuolation.
Fig. 2. Myofibroblast proliferation and concurrent loss of SMCs and endothelial cells were seen in the vessel wall early after graft implantation. Top: photomicrograph of a nongrafted epigastric vein shows a normal media consisting of SMCs. A sparsely populated adventitia and the endothelium are also seen. No cells resembling myofibroblasts were observed in nongrafted epigastric veins by light or electron microscopy. Center: the media of a 4-day graft is acellular, while the number of cells, both fibroblasts and myofibroblasts, in the adventitia has increased. One remaining endothelial cell is seen on the luminal surface. Bottom: a representative photomicrograph shows the thickening neointima of a 2-week graft. This neointima, as well as the adventitia, is composed of disorganized myofibroblastic cells (arrows) and mononuclear leukocytes. The media contains fibrin, marked cell debris and few cells. All HE. Original magnification ×130. L = Lumen; N = neointima; M = media; A = adventitia.
To localize myofibroblast cytoskeletal markers to discrete vein wall layers, we performed immunohistochemical staining. Medial and capillary smooth muscle cells (SMCs) stained for α-actin, desmin and smooth muscle myosin. We detected a loss of α-actin+ cells in the media by 4 days after grafting; immunoreactive α-actin+ cells appeared focally in the adventitia at 1 week. By 2 weeks, there was marked, diffuse α-actin staining in the neointima and continued focal staining in the adventitia. Parallel staining with isotype-matched control antibodies revealed no positive cells.
To confirm the results seen with light microscopy, we performed electron immunocytochemical staining of graft sections using a monoclonal antibody recognizing α-smooth muscle actin; positive cells were detected by a colloidal gold-conjugated secondary antibody. Electron microscopy confirmed that these large cells displayed all the characteristics of myofibroblasts, i.e. irregular indented nuclei, microfilaments (stress fibers) at the cell periphery, abundant endoplasmic reticulum, fibronexus junctions and long, numerous cytoplasmic processes [5, 10]. Gold particles indicating positive staining for α-smooth muscle actin were clearly visualized at the periphery of the cytoplasm (fig. 3). In nonoperated epigastric veins, only the microfilaments of medial SMCs stained strongly for α-actin, while fibroblasts were negative. In a 4-day graft, a few clearly degenerating medial SMCs as well as occasional large adventitial cells were immunopositive; negative fibroblasts continued to be seen. Positive adventitial cells were very abundant at 1 week, and at 2 weeks, positive cells were more equally distributed throughout the entire vein wall, although staining was more focal in the adventitia, and more diffuse in the neointima. Large positive cells continued to be seen at 8 weeks in the neointima.
Fig. 3. Electron photomicrograph of an α-actin+ myofibroblast in the neointima of an 8-week vein graft. Gold particles are detected in discrete regions of microfilaments at the periphery of the cytoplasm (arrows). Note the irregular nucleus (Nu) and the extensive rough endoplasmic reticulum with ribosomes (RER). Inset: gold particles (10 nm) are seen in the cell periphery. Original magnification ×10,800.
Four days after grafting, a large increase in proliferating cells was seen in the adventitia as detected by staining for incorporated BrdU (fig. 4). Light microscopy indicated that foci of myofibroblasts made up the majority of the proliferating cells seen in the adventitia at this time point. Less marked proliferation was seen in the neointima, starting at 4 days and peaking at 2 weeks. Histologic and immunohistochemical analysis revealed proliferating neointimal cells at 4 days to be mainly mononuclear cells. However, by 2 weeks, proliferating myofibroblasts predominated in the neointima. There were no proliferating cells in control epigastric veins. Staining with an anti-proliferating cell nuclear antigen antibody confirmed the BrdU findings (data not shown).
Fig. 4. Proliferating cells were seen mainly in the adventitia. Column chart demonstrating the increase in the percentage of proliferating cells after grafting. A very low level of proliferation is seen in the adventitia of nongrafted epigastric veins, which increases dramatically by 4 days after grafting. Data were derived by counting BrdU+ cells and total cells for each layer (neointima and media/adventitia) of a vein graft cross-section at the proximal region, and calculating the percentage. Three grafts per time point were assessed; error bars represent SEM.
To track the fate of adventitial cells after grafting, we employed adenoviral transfection of a gene encoding β-galactosidase as a tag. By ligating the ends of the graft, we assured that only the exterior of the graft was exposed to virus. Two weeks after graft placement, cells positive for β-galactosidase could be seen in the neointima (fig. 5a, b). No β-galactosidase activity was seen in control grafts not transfected with virus.
Fig. 5. Transfection of tied vein grafts with recombinant adenoviral β-galactosidase. a Using computer-assisted digital analysis, β-galactosidase-positive cells and total cells were counted in each proximal vein graft transverse cryosection. Cell numbers were calculated twice and a per-graft mean was derived. Five grafts were measured, and an overall mean ± SEM was derived. Columns B–D represent transfection efficiency, while columns E and F pertain to distribution and localization of β-galactosidase-positive cells within the vein wall. b Dual staining of a tied 2-week graft shows cells in the neointima positive for nuclear-targeted adenoviral β-galactosidase (blue) and positive for smooth muscle α-actin (brown). After a vein graft from the 2-week time point was soaked in X-gal solution (see Materials and Methods) to visualize β-galactosidase, it was sectioned onto slides. The slide was then subjected to immunohistochemistry using anti-smooth muscle α-actin as the primary antibody. The micrograph is a digitized image that was derived using a Pixera color camera and Metamorph software. Original magnification ×140. N = Neointima; M = media; A = adventitia. Arrows indicate regions of colocalization within the neointima.
A time course experiment was performed to follow the migration of the β-galactosidase-positive cells from the adventitia to the neointima. Implanted grafts were harvested after 1–7, 14 and 28 days (table 1). Detectable amounts of β-galactosidase accumulated after 3 days. There were no β-galactosidase-positive endothelial cells or medial cells at early time points, consistent with endothelial cell loss and medial SMC death; nor were there positive cells in the femoral artery at any time point, which was consistent with transfection of the graft with adenovirus before graft placement. Total and β-galactosidase-positive cells were counted in vein graft cross-sections that had been digitally imaged with a cooled CCD SenSys camera. By 3–4 days after transfection and grafting, β-galactosidase-positive cells were seen in the outer adventitia (fig. 6, top), and by 1 week, there were positive cells in the inner adventitia (the previously acellular media) (fig. 6, center). At 2 weeks, β-galactosidase-positive cells were detected in the neointima (fig. 6, bottom). Overall, transfection efficiency was approximately 22%; the percentage of β-galactosidase-positive cells remained constant from 3 days to 2 weeks (table 1). At 1 and 2 weeks, the number of β-galactosidase-positive cells increased, suggesting division of infected cells; this finding correlated with BrdU proliferation data (fig. 4). Positive cell numbers decreased by 4 weeks. The colocalization of nuclear β-galactosidase and cytoplasmic α-actin at 2 weeks was demonstrated by immunostaining tissue sections with anti-smooth muscle actin (asm-1) antibody (fig. 5b).
Table 1. Transfection of vein grafts with recombinant adenoviral-β-galactosidase
Fig. 6. Cells tagged with replication-defective, nuclear-target adenoviral β-galactosidase were seen in several locations in the vessel wall. Top: in a 4-day graft, positive nuclei are prevalent in other outer adventitia. Center: cells in the previously acellular media are positive in a 1-week graft. Bottom: in a 2-week graft, β-galactosidase-positive cells are detected in the neointima. Micrographs of noncounterstained cryosections are digitized images that were derived using a cooled CCD black and white SenSys camera. Original magnification ×130. Arrows indicate the internal elastic lamina. L = Lumen; N = neointima; M = media; A = adventitia.
The identity and origin of cells that repopulate the vessel wall in both injured arteries and bypass vein grafts are topics of considerable interest, especially in relation to the prevention of graft occlusion caused by IH. Historically, it has been thought that medial SMCs repopulate the neointima following their activation, proliferation and migration in response to various stimuli [1, 23]. In a rat model of vein graft IH, we have shown that there is significant SMC degeneration and death in the media, leading to an acellular medial layer and a markedly increased quantity of fibrin and cellular debris early after grafting . Medial cell death has also been demonstrated in human grafts . Thus, the origin and identity of neointimal cells remain controversial. Recently, the effect of luminal balloon and stent injury on the adventitia has indicated that the adventitia likely contributes cells to the neointima [13, 25, 26]. We have previously reported that, during neointimal development in vein grafts, the predominant mesenchymal cells seen throughout the vessel wall resemble myofibroblasts . In this study of vein graft IH, we investigated graft myofibroblasts to characterize their origin.
Within hours of implantation into the arterial circulation, there is an inflammatory response in the vein graft during which leukocytes infiltrate the vein graft wall, much of the endothelial monolayer is lost and medial SMCs degenerate and die . At this time (4 days to 1 week), numerous large proliferating cells appear in the adventitia, an area normally populated by widely dispersed fibroblasts. Recent analyses of injured arteries have demonstrated that the first major site of cell proliferation after angioplasty is in the adventitia and not the media [13, 14]. The numbers of macrophages and other leukocytes are also increased in the injured adventitia after grafting [16, 19], and it is conceivable that these cells contribute to an increase in proliferation by secreting mitogens such as platelet-derived growth factor  and transforming growth factor-beta 1 . Proliferation in the media/adventitia decreases by 4 weeks, which, together with a decrease in recruited inflammatory cells, leads to diminished overall cellularity.
In our model, the phenotype of the large adventitial cells, as defined by histology and transmission electron microscopy, resembles that of myofibroblasts [9, 10, 11, 29]. These cells, seen at 4 days in the adventitia, have large indented nuclei, discrete regions of organized microfilaments (stress fibers), copious rough endoplasmic reticulum and long, numerous cytoplasmic extensions. These features are intermediate to those of SMCs and fibroblasts, and are the hallmarks of myofibroblasts . The elevated numbers and overall appearance of these cells suggest marked proliferative and synthetic activities within the vessel wall, consistent with healing and remodeling in the vein graft. In normal wounds, myofibroblasts disappear with time, usually by apoptosis . In our vein grafts, myofibroblasts continue to be seen at late time points and display few signs of apoptosis. As the vein graft ages, the incidence of overall apoptosis in the graft is diminished while myofibroblasts remain and the neointima develops; this suggests a state of abnormal healing .
In this report, we have shown the migration of adventitial myofibroblasts towards the lumen to populate the neointima. To study the migration of adventitial cells towards the lumen, we employed adenovirus-mediated gene transfer of a marker gene, β-galactosidase, to tag the adventitial cells. Solely in the adventitia of 4-day grafts were (1) proliferating cells, (2) cells with myofibroblast morphology and (3) cells expressing the adenovirally introduced gene for β-galactosidase. All of these cells are seen at 1 week in the previously acellular media, and by 2–4 weeks in the developing neointima.
Our results using adenovirus to tag adventitial cells at the time of grafting are consistent with the work of Kalra and Miller , who examined saphenous vein grafts in dogs. They found, by BrdU labeling, early proliferating cells in the adventitia and media. These labeled cells were then found to be α-smooth muscle actin positive (characteristic of myofibroblasts) and to have translocated to the neointima at later time points. In a similar experiment utilizing BrdU, Shi et al.  showed that the fibroblasts that eventually migrate to the lumen need not originate in the vein graft. In these experiments, labeled perivascular fibroblasts translocated through the porcine saphenous vein, resulting in colocalization of BrdU and α-smooth muscle actin in cells of the neointima. In addition to the adventitial cells that proliferate after vessel injury, BrdU labeling inherently labels other cells of the vessel wall that happen to be replicating at the time of the BrdU pulse. In this study, by applying the adenovirus to the exterior of the vessel, we directed the label to the adventitia, which implies that any labeled cells that are seen in the neointima originated from the adventitia. This result does not contradict the idea that perivascular tissue is an additional source of myofibroblasts, as Hu et al.  showed, using transgenic mice, that in their vein isograft model, 40% of neointimal cells originate from the host and 60% from the donor vessel .
In an experiment using retrovirus and rat carotid arteries, primary syngeneic adventitial fibroblasts were stably transduced in cell culture with β-galactosidase and then reintroduced into the adventitia immediately after balloon injury . Our results confirm the time course of adventitial, medial and neointimal β-galactosidase expression seen by Li et al.  using these cultured fibroblasts. On the other hand, de Leon et al.  showed the lack of migration of myofibroblasts from the adventitia to the neointima in rat carotid artery and indicated that the intact internal elastic lamina and medial layers of this artery may be a barrier to migration. The epigastric vein in the rat is very thin walled, similar to the human cephalic and thinner than the greater saphenous vein. We seen florid IH in this model, similar to human cephalic vein grafts, which have poorer patency rates than saphenous vein grafts. In this model, there is near complete medial SMC and endothelial cell death after graft implantation, rendering it an excellent model to study the contribution of the adventitia to vessel remodeling and neointimal development.
We have demonstrated the differentiation of adventitial fibroblasts into myofibroblasts and their subsequent migration to form a neointima. We hypothesize that the remodeling of the neointima by adventitially derived myofibroblasts occurs in human vein grafts to a varying degree, perhaps influenced by the thickness of the media; our model represents an extreme end of that continuum. Thus, the more severe IH seen in the cephalic vein compared to the greater saphenous vein may be a function of wall thickness and subsequent greater contribution of the adventitia to vessel remodeling and neointimal development. These results are exciting because they imply that therapeutic strategies targeting adventitial fibroblasts, perhaps at the time of surgery, might limit the development of intimal hyperplasia.
This study was supported by grant No. 96012580 of the American Heart Association and a Merit Grant from the Department of Veterans Affairs.
Dr. John R. Hoch
University of Wisconsin Hospitals and Clinics
G5/325 CSC, 600 Highland Avenue
Madison, WI 53792 (USA)
Tel. +1 608 263 1388, Fax +1 608 263 7652, E-Mail firstname.lastname@example.org
Received: April 12, 2002
Accepted after revision: December 3, 2002
Number of Print Pages : 10
Number of Figures : 6, Number of Tables : 1, Number of References : 35
Journal of Vascular Research (Incorporating International Journal of Microcirculation)
Founded 1964 as Angiologica by M. Comèl and L. Laszt (1964–1973) continued as Blood Vessels by J.A. Bevan (1974–1991)
Official Journal of the European Society for Microcirculation
Vol. 40, No. 3, Year 2003 (Cover Date: May-June 2003)
Journal Editor: U. Pohl, Munich
ISSN: 1018–1172 (print), 1423–0135 (Online)
For additional information: http://www.karger.com/jvr