Plasminogen Activator Expression Correlates with Genetic Differences in Vascular RemodelingKorshunov V.A.a · Solomatina M.A.b · Plekhanova O.S.b · Parfyonova Y.V.b · Tkachuk V.A.b,c · Berk B.C.a
aCenter for Cardiovascular Research, Aab Institute of Biomedical Sciences and Department of Medicine, University of Rochester, Rochester, N.Y., USA; bMolecular Endocrinology Laboratory, Institute of Experimental Cardiology, Russian Cardiology Research and Production Center, and cMoscow State University, Moscow, Russia Corresponding Author
Intima-media thickening (IMT) of the carotid artery, a form of vascular remodeling, correlates well with coronary artery disease risk in humans. Vascular remodeling in response to blood flow is a complex process that critically involves altered cell matrix interactions. To gain insight into these events, we performed partial carotid ligation (left carotid (LCA) = low flow and right carotid (RCA) = high flow) in 2 inbred mouse strains: C57Bl/6J (C57) and FVB/NJ (FVB). To evaluate the role of the 2 major matrix-degrading systems, plasminogen activators (PAs) and matrix metalloproteinases (MMPs), we compared the expression of u-PA, t-PA, MMP-2 and MMP-9 in ligated carotids of C57 and FVB mice. The extent of remodeling was greater in response to low LCA than high RCA flow. Despite a similar decrease in LCA flow in both strains, maximal IMT volume was greater in FVB (82 ± 7 × 10–6 μm3) than in C57 (38 ± 4 × 10–6 μm3) after ligation. Among PAs and MMPs, increased expression of t-PA and u-PA correlated with increased IMT (p < 0.0005 and p < 0.001, respectively). MMP-2, MMP-9 and tissue inhibitors of metalloproteinase-2 expression also increased, but did not differ between strains. In summary, flow-induced IMT of the carotid is genetically determined and correlates with t-PA and u-PA expression in 2 inbred mouse strains.
Copyright © 2004 S. Karger AG, Basel
Carotid intima-media thickening (IMT) is an important predictive phenotype for human cardiovascular disease [1, 2]. For example, in the Rotterdam Study, the odds ratio for stroke per standard deviation increase in the carotid IMT (0.163 mm) was 1.41 . The European Atherosclerosis Study demonstrated that peripheral vascular disease was significantly associated with an increased carotid IMT . These data provide empirical evidence that IMT is a highly predictive clinical measurement of cardiovascular disease. In addition, there appears to be a significant genetic component in IMT development based on epidemiological studies [4, 5].
Blood vessel diameters normally enlarge with increased flow (‘outward’ remodeling) and ‘shrink’ with decreased flow (‘inward’ remodeling), which is known as flow-dependent vascular remodeling . The primary stimulus is thought to be shear stress, which is ‘normalized’ (∼15 dyn/cm2) by changes in vessel diameter. In human renal, carotid and coronary arteries, vessel IMT and atherosclerotic plaque are frequently associated with outward compensatory remodeling termed the ‘Glagov phenomenon’ [7, 8]. Outward remodeling allows the vessel to accommodate the plaque and maintain physiological shear stress. Blood pressure is another important stimulus that alters vascular structure and stimulates remodeling [9, 10]. In hypertension, vascular remodeling also involves IMT, which normalizes wall stress . Recent studies in human coronary arteries showed that regions exposed to low shear stress were associated with increased IMT and outward remodeling (increased external elastic lamina, EEL) [11, 12].
The mechanisms for vessel IMT and outward remodeling remain unknown, but involve changes in both cells and matrix of the vessel wall . In flow-dependent remodeling, essential roles for endothelial cell-derived mediators including nitric oxide, prostanoids and growth factors have been demonstrated [14, 15, 16]. Previous data showed important roles for plasminogen activators (PAs) [17, 18] and matrix metalloproteinases (MMPs) [19, 20] as mediators of vascular remodeling after injury and during development. A key role for u-PA compared with t-PA was suggested after vascular injury [21, 22]. To elucidate the relative importance of the 2 major matrix-degrading systems, PAs and MMPs, we compared the expression of u-PA, t-PA, MMP-2, MMP-9 and tissue inhibitors of metalloproteinase (TIMP)-2 in ligated carotids of C57Bl/6J (C57) and FVB/NJ (FVB) mice. We chose FVB and C57 strains of mice based on their diverse responses to the absence of flow in the carotid artery . Over a 4-week time course we found changes in all proteolytic enzymes, although only t-PA and u-PA differed among strains. The greater IMT observed in FVB was associated with greater t-PA and u-PA expression, suggesting a significant role for t-PA and u-PA in flow-dependent remodeling.
Male and female C57 and FVB mice (8 weeks old, Jackson Laboratories, Bar Harbor, Me., USA) were used in accordance with the guidelines of the National Institutes of Health for the Care and Use of Laboratory Animals. All procedures of this study were approved by the University of Rochester Animal Care Committee.
Mice were anesthetized with a cocktail of ketamine and xylazine (130 and 9 mg/kg, i.p.). Previously, we developed a model of flow-induced vascular remodeling . Briefly, the branches of the left external and internal carotid arteries were isolated and ligated with 6–0 silk suture, and left carotid blood flow (BF) was reduced to flow through the patent occipital artery only. For each inbred strain, 2 groups of ligated or sham-operated animals were processed for morphological and immunohistochemical studies at 7, 14 and 28 days after surgery. Systolic blood pressure (SBP) and heart rate (HR) were measured by tail-cuff plethysmography (Visitech System, USA), and body weights were measured weekly as well. In a separate experiment (n = 2 in each group), BF was measured in both carotids using an ultrasonic transit-time volume flowmeter (Transonic Systems Inc.). Shear stress was calculated as described elsewhere . Because of technical limitations (lumen radius was averaged from histological evaluations), we calculated only 1 shear stress value for each group.
At the time of termination, all animals were anesthetized and perfused with saline at a pressure of 100 mm Hg through the left ventricle for 4 min, followed by 3 min perfusion with 10% paraformaldehyde in sodium phosphate buffer (pH 7.0) . The left and right common carotid arteries were harvested and embedded in paraffin. A series of cross-sections (4 μm) were made from the bifurcation every 200 μm through 2 mm length of carotid artery as described . Cross-sections were stained with hematoxylin and eosin, and were analyzed using MCID image software (Imaging Research Inc., St. Catherines, Canada). The volumes of vessel compartments were calculated as described previously . Since there were no statistically significant differences in vessel compartment volumes between males and females for both strains during the time course (data not shown), we combined data for both sexes. Remodeling of the carotids (‘inward’ vs. ‘outward’) was defined by comparing the area engulfed by the EEL to the control vessels.
Sections were deparaffinized with xylene, rehydrated and treated with 0.3% methanolic hydrogen peroxide to quench endogenous peroxidase. Antigen retrieval methods and antibody sources are shown in table 1. The sections were incubated in 10% serum from the same species as the secondary biotinylated antibodies and then with antibodies to u-PA (polyclonal), t-PA (polyclonal), MMP-9 (polyclonal), TIMP-2 (polyclonal), Ki-67 (monoclonal), CD45 (monoclonal) or control nonimmune IgGs (at concentrations coinciding with those of each immune IgG) for 30 min for anti-Ki-67 or 1 h for other antibodies in a humidified chamber. After multiple washings in phosphate-buffered saline ( pH 7.4; except for anti-Ki-67 where only TBS, pH 7.4, was used), the sections were incubated with biotinylated anti-mouse, anti-rabbit or anti-goat antibodies (15 μg/ml). Antigens were detected using the ABC method (Vector Laboratories Inc.) and chromogen 3,3′-diaminobenzidine tetrahydrochloride before lightly staining the sections with hematoxylin. MMP-2 and TIMP-2 antigens were detected using mouse monoclonal antibodies and an Animal Research kit (Dako) according to the manufacturer’s instructions.
Table 1. Antigen retrieval methods
The intensities of immunohistochemical staining for u-PA, t-PA, MMPs and TIMP-2 were graded in a semi-quantitative blinded manner by 3 histologists, using a scale from 0 to 4, with 0 indicating no staining (background); 1 indicating weak staining in a few cells and/ or matrix; 2 indicating weak staining in the majority or strong staining in a few cells and/or matrix; 3 indicating consistent positive staining in the majority of cells and/or matrix; 4 indicating intense staining among all cells and/or matrix. The observations were made in the media-intima area of cross-sections (3 sections/mouse) of the middle part of the carotid (2–3 mm from carotid bifurcation) . The densities of Ki-67-positive cells  and CD45-positive cells  were calculated in selected samples (counts × 10–6 μm2) for every point in time as described .
All results are reported as mean ± SEM. All statistical tests were done with Statview for MacIntosh, version 5.0.1. A comparison for 2 groups was performed using Student’s t test. For categorical data in table 4, we performed nonlinear analysis using the Mann-Whitney U test. Differences between more then 2 groups were analyzed by means of a repeated-measures one-way ANOVA and followed by a Fisher’s post-hoc test. The level of p < 0.05 was regarded as significant.
Table 4. Immunohistochemical evaluation of PAs and MMPs in the LCA: time course
Baseline physiological characteristics of C57 and FVB mice are shown in table 2. In general, males were significantly heavier (∼3.0 g) than females, but had similar SBP and HR, so these parameters were combined. Initially, FVB mice were heavier than C57. The initial values for SBP and HR were significantly higher in FVB (∼134 mm Hg, 670 beats/min) compared with C57 (∼129 mm Hg, ∼610 beats/min; table 2). There were no significant differences in SBP among experimental groups of mice over the time course (data not shown). The HR of FVB exhibited a trend to decrease over the time course, but did not change in C57 mice (data not shown). Thus, FVB mice, in general, had higher values for physiological parameters than C57 mice, but changes were equivalent over the time course in both strains for sham-operated and ligated mice.
Table 2. Physiological parameters in C57 and FVB mice
The initial level of BF was the same for both inbred strains (table 3). After ligation, the BF significantly decreased (–90%) in the left carotid artery (LCA) and increased (+70%) in the contralateral right carotid artery (RCA). Both strains responded similarly over time to partial ligation (table 3).
Table 3. Blood flow in the carotid arteries after ligation: time course
Although physiological changes were similar in both inbred strains, ligation resulted in dramatic strain-dependent differences in morphology (fig. 1). The major findings based on analyses of vessel compartment volumes were the following: (1) partial ligation of the LCA caused larger outward remodeling in FVB mice compared with C57 mice, (2) LCA IMT was significantly greater in FVB mice, and (3) increased flow enlarged the RCA lumen similarly in both strains.
Fig. 1. Left carotid compartment volumes of C57 and FVB mice after ligation. Since there were no differences in volumes among sham-operated mice over time, sham-operated mice were combined into 1 group for C57 and FVB strains. Remodeling was significantly greater in FVB mice after ligation. Values are mean ± SEM. * p < 0.05, compared with sham (ANOVA). # p < 0.05, compared with C57 (ANOVA). a Lumen volume. b Media volume. c Intima volume. d Adventitia volume.
There were no significant differences between C57 and FVB in the RCA-remodeling response. In the RCA, the biggest changes were observed in lumen volume, which was significantly increased within 1 week (fig. 2a). There was no intima formation in sham-operated mice or RCAs from ligated mice (data not shown). The RCA media and adventitia volumes changed very little during the time course (fig. 2b, c). Vessel size was defined by the area contained within the EEL (sum of lumen + intima + media). Analysis of the EEL volume showed that RCA underwent outward remodeling (∼105 × 10–6 μm3) in both strains compared with sham-operated mice (∼85 × 10–6 μm3).
Fig. 2. Right carotid compartment volumes of C57 and FVB mice after ligation. a Lumen volume. b Media volume. c Adventitia volume. Open bars are C57 mice, solid bars are FVB mice. Since there were no differences in volumes among shams over time, shams were combined into one group for C57 and FVB strains. Values are mean ± SEM. * p < 0.05 compared with sham (ANOVA). # p < 0.05 compared with C57 (ANOVA).
In the LCA, there were significant differences between C57 and FVB, as well as large changes in lumen and vessel wall volumes compared with sham-operated mice. The LCA lumen in the C57 was significantly decreased at 4 weeks after ligation (fig. 1a). The C57 media volume increased 1.5 ± 0.1-fold compared with sham-operated mice during the time course (fig. 1b). A small intima formed within 1 week in C57 ligated LCAs (fig. 1c). The adventitia volume in C57 was also significantly increased at 1 and 2 weeks after ligation (fig. 1d).
The LCA lumen volume increased in FVB (in contrast to C57) and was maximal (1.5 ± 0.1-fold) at 2 weeks (fig. 1a). Media volume increased in FVB, reaching the maximum at 2 weeks (fig. 1b). A large intima developed in FVB over time, which was biggest at 2 weeks (fig. 1c). The adventitia volume in FVB also significantly increased after ligation (fig. 1d).
We next compared the characteristics of remodeling in C57 and FVB. Initially, shear stress was normal and equal in both strains: C57 (LCA = 25.4 dyn/cm2, RCA = 24.8 dyn/cm2) vs. FVB (LCA = 24.6 dyn/cm2, RCA = 25.0 dyn/cm2). After ligation, shear stress decreased equally in the LCA of C57 and FVB (3 and 4 dyn/cm2, respectively) and increased slightly, but equally, in the RCA (32 and 27 dyn/cm2, respectively).
There were significant differences in vascular remodeling between C57 and FVB in both lumen and vessel wall. The biggest strain-dependent differences were observed in the ligated LCA. In FVB, intima-media volumes were always larger than in C57, with maximum remodeling 2 weeks after ligation (fig. 1b, c). Most impressively, the ratio of intima to media volume was 7-fold bigger in FVB (0.36 ± 0.04) compared with C57 mice (0.05 ± 0.03). The LCA intima-media volume was maximal in FVB (82 ± 7 × 10–6 μm3) compared with C57 (38 ± 4 × 10–6 μm3) 2 weeks after ligation. Despite the increase in IMT, the LCA of FVB exhibited greater outward remodeling as measured by EEL volume. The vessel size measured by EEL was dramatically larger in FVB (180 × 10–6 μm3) vs. C57 (90 × 10–6 μm3) at 2 weeks, while in sham-operated mice it was 80 × 10–6 μm3. In summary, both intima-media and outward vascular remodeling in the LCA were significantly greater in FVB compared with C57 mice.
To evaluate possible mechanisms that regulate vessel remodeling, we plotted outward remodeling (EEL volume) and IMT (intima-media volume) in both strains after ligation (fig. 3). No correlation was observed in sham-operated mice (data not shown). There was a significant correlation of EEL with IMT for both strains. However, the slopes differed significantly between C57 (0.55) and FVB (0.80, p < 0.05; fig. 3). The relatively flat slope for C57 suggests that genetic differences in the vessel wall response caused C57 mice to be less ‘sensitive’ than FVB mice to remodeling stimuli.
Fig. 3. Correlation analysis between IMT and EEL logarithmic volumes of the ligated carotid vessel from C57 and FVB mice. Slopes of regression lines were compared by using the Bonferroni correlation, p < 0.05.
Cell Proliferation and Infiltration. To elucidate mechanisms responsible for remodeling, we analyzed cell proliferation and monocyte/macrophage infiltration. The time course for cell proliferation was similar in both inbred strains: significantly increased in the ligated LCAs at 1 week and rapidly decreased to baseline at 2 and 4 weeks (fig. 4a). At 1 week, proliferation was significantly higher in FVB than in C57 mice (fig. 4a). There was no proliferation in the ligated RCA or carotids from sham-operated animals (data not shown). Analysis of vessels for the presence of monocytes/macrophages by CD45 staining showed reactivity only in the adventitia (data not shown). In the ligated LCA, infiltration was greatest at 1 week and decreased at 2 and 4 weeks (fig. 4b). There were no positive cells within the EEL in RCA of sham-operated or ligated mice. Thus, cell proliferation and monocyte/macrophage infiltration were significantly greater in FVB than C57 mice at 1 week after ligation, but then became similar.
Fig. 4. Proliferation and infiltration cell densities in ligated LCA (cell number/intima-media). There were no Ki-67-positive cells in sham-operated mice. CD45-positive cells were found only in the adventitia of sham-operated mice. Values are mean ± SEM. * p < 0.05, compared with 1 week (ANOVA). # p < 0.05, compared with C57 (ANOVA). a Ki-67-positive cells. b CD45-positive cells.
Plasminogen Activators. PAs are candidate mediators for vascular remodeling, based on knockout studies . PAs were readily detected in the media of sham C57, but not in sham FVB (table 4). Both t-PA and u-PA expressions were highly regulated after ligation in the LCA (table 4). The major strain-dependent differences in ligated vessels were increased t-PA and u-PA expression in FVB compared with C57 (fig. 5, table 4). t-PA immunoreactive peptides were detected primarily in the media and intima of ligated LCAs (fig. 5a–c). The score for t-PA expression by immunohistological staining was 3 times higher in FVB compared with C57 carotids, with maximum expression at 2 weeks after ligation (table 4). Increases in u-PA immunoreactive peptides were also apparent after ligation (fig. 5d–f). u-PA expression increased to the same extent in both strains at 1 week after ligation (table 4), when cell proliferation and inflammatory cell infiltration were enhanced (fig. 4). In contrast to C57, u-PA immunostaining remained elevated in FVB mice over the time course (table 4). There was a significant correlation between the t-PA and u-PA expression scores and IMT at 2 and 4 weeks after ligation (fig. 6a, b). There was also a significantly higher expression score (>2) and larger intima-media volume (>45 × 10–6 μm3) in FVB mice (fig. 6a, b). In summary, the time course for plasminogen activator expression correlated with cell proliferation and vascular remodeling after flow reduction. Importantly, strain-dependent differences in vascular remodeling between FVB and C57 inbred mice also correlated with the magnitude of t-PA and u-PA expression.
Fig. 5. Expression of t-PA and u-PA in the LCA of C57 and FVB mice 2 weeks after ligation. ×40. a Controls (nonimmune IgG for t-PA in the C57). b t-PA in the C57. c t-PA in the FVB. d Controls (nonimmune IgG for u-PA in the C57). e t-PA in the C57. f t-PA in the FVB.
Fig. 6. Correlation analyses between IMT volumes and PAs immunological score in the ligated carotid vessel from C57 and FVB mice. Black diamonds are mean for C57; black triangles are mean for FVB. a Intima-media volume vs. t-PA immunological score at 2 and 4 weeks after ligation. b Intima-media volume vs. t-PA immunological score at 2 and 4 weeks after ligation.
Matrix Metalloproteinases and Tissue Inhibitors of Metalloproteinases. MMPs and TIMPs are another pair of candidate mediators for vascular remodeling . MMP-2 was upregulated as early as 1 week after flow alteration in the ligated LCAs of both mice strains and declined slightly without significant differences between strains (table 4). MMP-9 was not detected in the media of carotid arteries of sham-operated C57 or FVB mice (table 4). MMP-9 expression was elevated in the LCAs of both FVB and C57 ligated mice compared with that of sham-operated mice (table 4), but no significant differences between strains were observed.
The expression profile of TIMP-2 was similar in C57 and FVB mice: increased expression at 1 and 2 weeks and return to baseline at 4 weeks after ligation (table 4). In summary, expression of MMPs and TIMP-2 varied with time (peak at 1–2 weeks), but there were no strain-dependent differences (unlike t-PA and u-PA).
The major finding of this study is that flow-induced vascular remodeling of the mouse carotid, similar to IMT in humans, is genetically dependent as shown in 2 inbred mouse strains. The largest changes in vascular remodeling, cell function and protein expression were found in the low flow LCA. The ligated LCAs compensated for growth of intima and media by increasing the lumen and vessel (EEL) area. In both strains, outward remodeling correlated well with intima-media volume. However, there was less outward remodeling in C57 for a given increase in intima-media volume with decreased lumen volume suggesting ‘limited’ remodeling (fig. 3). We recently reported that vascular remodeling initiated by partial ligation was associated with increased extracellular matrix reorganization . In the present study, we found that PA expression correlated with vascular remodeling after flow reduction with large increases in t-PA and u-PA in the ligated LCA. By immunohistochemistry, increased t-PA and u-PA expression correlated with a greater IMT response in FVB mice (fig. 6). The expression of MMP-2, MMP-9 and TIMP-2 increased with remodeling, but did not exhibit strain-dependent differences. In addition, monocyte/macrophage infiltration and cell proliferation differed between strains, with greater increases in FVB (at 1 week only, fig. 4).
The mechanisms for vascular remodeling are complex and require cell and matrix reorganization. Recently, we developed a model of IMT after partial carotid ligation. In this model, we found that extracellular matrix reorganization and vascular smooth muscle cell (VSMC) proliferation were prominent features . In the present study, we focused on the roles of 2 major matrix-degrading systems: PAs and MMPs. There is much evidence to support the importance of PAs in vascular injury and remodeling. For example, transplant arteriosclerosis was reduced in plasminogen-deficient mice . Increased expression of u-PA and PAI-1 was demonstrated in growing intima during adaptive arteriogenesis . We have shown that u-PA appears to stimulate VSMC proliferation and neointima formation after balloon injury more than t-PA . However, u-PA and t-PA have complex contributions after carotid artery injury and thrombosis . Specifically, u-PA deficiency resulted in a significantly increased intima area due to the large amount of unresolved acellular thrombus, while t-PA knockout mice developed cell-rich multilayered media and neointima. In the present study, we observed increased expression of both t-PA and u-PA after partial ligation (table 4). In contrast to previous studies, we found that both u-PA and t-PA significantly correlated with strain-dependent IMT at 2 and 4 weeks after ligation (fig. 6), which may be due to differences in the stimuli used to induce remodeling. These data suggest that both u-PA and t-PA are important in flow-induced IMT.
In contrast to the strain-dependent differences in u-PA and t-PA expression, the levels of MMP-2, MMP-9 and TIMP-2 did not differ. However, since PAs can activate MMPs, the increase in u-PA and t-PA may lead to enhanced MMP activity in FVB mice. PAs may activate MMPs directly and/or by increasing plasmin generation . In vascular remodeling, after cessation of flow , the maximal level of MMP-9 was higher than MMP-2, with a peak 7 days after ligation. A key role for MMP-9 was suggested by the finding in MMP-9-deficient mice of decreased intima formation, increased lumen diameter and significant accumulation of interstitial collagen after complete ligation . In contrast, in the present study, the time course of the expression of MMP-2 and MMP-9 was similar in both strains (table 4). An important difference may be the presence of thrombosis in the flow cessation model . In our experiments, MMP-2, MMP-9 and TIMP-2 were increased 1 week after ligation (table 4), although not in a strain-dependent manner. In summary, we have shown significant upregulation of MMP-9, MMP-2 and TIMP-2 in flow-induced vascular remodeling (IMT), although expression did not differ between strains.
We previously reported that the mechanism for IMT in C57 mice is intrinsic to the vessel wall and unrelated to shear stress . A similar process occurs in outward remodeling associated with neointima in human renal, coronary and carotid arteries (Glagov phenomenon) [7, 8]. In the present study, EEL and IMT were strongly related in both strains, but the slopes differed between C57 and FVB (fig. 3), suggesting a less sensitive response in C57 mice. Potential mechanisms for the differences in response include migration and proliferation of VSMC and inflammatory cells. Proliferation is essential for vascular remodeling after flow reduction, since the absence of the cell cycle inhibitor p130 enhanced the injury response . We observed a rapid (1 week) proliferation of VSMC in the LCA, which was significantly higher in FVB (fig. 4a). The onset of proliferation was coincident with monocyte infiltration (fig. 4b). Furthermore, FVB mice showed significantly higher levels of infiltration and proliferation 1 week after ligation (fig. 4b). In our experiments, maximal expression of u-PA, MMP-2 and MMP-9 (table 4) was coincident with maximal infiltration and proliferation 1 week after ligation (fig. 4). However, only t-PA and u-PA correlated with maximal outward remodeling with IMT (fig. 6). In summary, mechanisms that contribute to outward vascular remodeling in the presence of IMT likely include monocyte infiltration, extracellular matrix reorganization by changes in PA and MMP activity and VSMC proliferation and migration.
Increased carotid artery IMT is associated with an increased risk of coronary heart disease and cerebrovascular events [30, 31]. The adjusted heritability estimate for carotid IMT was relatively high at 0.32 (p = 0.02) in the presence of type 2 diabetes . Recently, Fox et al.  showed that about 40% of the variability in carotid IMT was dependent on genetic factors. There were also significant genetic differences along 11 inbred mice strains in intima formation induced by complete carotid ligation . Our model is uniquely suited to study mechanisms of IMT and compensatory outward remodeling in the presence of an intima . In addition to the roles for PAs and MMPs shown in this study, future efforts to elucidate these mechanisms will include both candidate gene approaches and QTL analysis.
This study was supported by HL-62826 to B.C.B. and by NHLBI funds for the exchange of scientists under the auspices of the US-Russia Joint Agreement in Cardiopulmonary Research. The authors would like to thank Mary Georger, David Nagel and Sarah McCarty for immunohistological measurements.
Bradford C. Berk, MD, PhD
University of Rochester, Center for Cardiovascular Research
601 Elmwood Ave, Rochester, NY 14642 (USA)
Tel. +1 585 273 1946, Fax +1 585 273 1497
Received: May 25, 2004
Accepted after revision: July 30, 2004
Published online: October 28, 2004
Number of Print Pages : 10
Number of Figures : 6, Number of Tables : 4, Number of References : 31
Journal of Vascular Research (Incorporating International Journal of Microcirculation)
Vol. 41, No. 6, Year 2004 (Cover Date: November-December 2004)
Journal Editor: U. Pohl, Munich; G.A. Meininger, College Station, Tex.
ISSN: 1018–1172 (print), 1423–0135 (Online)
For additional information: http://www.karger.com/jvr