| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Pathology, University of Washington, Seattle, Washington
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
PDGF receptor ß (PDGFRß) is expressed by vascular SMCs and other stromal and connective tissue cells.2 It can mediate both chemotaxis and proliferation of cultured cells in response to binding of its ligands, PDGF B-chain2-4 or the recently discovered protease-activated PDGF-D chain.5,6 This suggested the hypothesis that injury results in the release of PDGF from activated platelets and macrophages, and that this PDGF contributes to driving the cell migration and proliferation involved in the response to injury.2 This general hypothesis has been supported by many studies in animal models in which the effects of augmenting or inhibiting the PDGF/PDGF receptor system have been evaluated.7-12 These studies have generally concluded that PDGF is acting primarily as a chemotactic agent rather than as a mitogen. For example, when rats were depleted of platelets, a rich source of PDGF B-chain,13 the neointima that formed after balloon injury was smaller, but there was no detectable difference in SMC replication.14 Studies using exogenous PDGF-BB9 and neutralizing antibodies against PDGF-BB,7 or specific inhibitor of PDGF receptor tyrosine kinase,15 further suggested that the primary role of PDGF-BB was to induce migration of SMCs from the media into the intima, with little or delayed effects on replication.
Embryos that are homozygous for disruption of PDGF B-chain16 or PDGFRß17 die at or before birth, with comparable abnormalities in vascular development that reflect reduced presence of vascular SMCs, or SMC-like cells, in small vessels.18,19 The phenotype of PDGFRß-/- mice supports a role for PDGFRß in vascular development, but the early lethality prevents their use in evaluating the role of PDGFRß system in adult tissue responses to injury. However, chimeric embryos, prepared by fusing wild-type (WT) and PDGFRß-/- pre-implantation mouse embryos, are viable and develop into normal adults.20 During developmental processes that use PDGFRß ligands to drive cell proliferation or recruitment, PDGFRß-/- cells in a chimeric embryo become progressively less abundant as they are "out-competed" by the WT control cells which do express PDGFRß.20 This competitive disadvantage becomes the basis for calculating the magnitude of the role that PDGFRß plays in the development of different cell lineages. Chimera analysis revealed that PDGFRß plays an important role in the development of all muscle lineages, including pericytes,20 mesangial cells,21 and arterial SMCs,20 and it revealed that, counter to expectation, PDGFRß does not play a significant role in the development of fibroblast or endothelial cell populations.
Because the requirement for PDGFRß expression during development is not absolute, every cell lineage and tissue in an adult PDGFRß-/- 
WT chimera includes at least some PDGFRß-/- cells. This makes it possible to evaluate the further roles of PDGFRß in adult disease processes by comparing the responses of adjacent WT versus PDGFRß-/- cells within the tissue. Because the positive and negative cells share the same local environment in the same chimeric individual, animal-to-animal variation and indirect and systemic effects of the gene deletion are minimized. Chimera analysis of adult responses to injury have revealed some surprises. In granulation tissue formed in response to subcutaneous implantation of a surgical sponge, we found that fibroblasts and endothelial cells, which did not depend on PDGFRß expression during development,20
were virtually excluded from accumulation in developing granulation tissue unless they expressed normal levels of PDGFRß.22
In this report, we describe the use of chimeric mice to support the hypothesis that, after arterial injury, PDGFRß plays an important direct (ie, cell-autonomous) role in driving SMC migration from the media to the intima, but not in the response of fibroblasts in the adventitia.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Two lines of mice were used to generate morulae for the creation of aggregation chimeras, as previously described.20
In brief, the wild-type SWR (PDGFRß+/+) line provided control (WT) component morulae. PDGFRß+/- C57Bl/6 mice17
homozygous for a pBR322/globin insert23
provided marked PDGFRß+/+, +/-, and -/- morulae. The pBR322/globin insert is not expressed and serves as a target for non-isotopic in situ hybridization which allowed us to identify cells from the experimental line.20
Morulae from these two lines were fused ex vivo to produce PDGFRß+/+ WT, PDGFRß+/- WT, or PDGFRß-/- WT chimeric embryos, which were implanted into pseudopregnant CD-1 females.
Mouse Model of Vascular Injury
All mice were conventionally housed in the University of Washington vivarium and all procedures were approved by the University of Washington Institutional Animal Care and Use Committee. Adult chimeric mice of both sexes were anesthetized by intraperitoneal injection of 100 mg/kg ketamine and 8 mg/kg xylazine. A ventral midline cervical incision was made, the left common carotid artery was isolated by blunt dissection and ligated with 6-0 silk just proximal to the bifurcation into the internal and external branches, as described by Kumar and Lindner.24 The skin was closed using Michel wound clips. Mice were euthanized 4 weeks post-surgery, and the left and right carotid arteries were harvested, fixed with methyl Carnoy’s (60% methanol, 30% acetic acid, 10% chloroform), embedded in paraffin, and 5-micron sections cut.
Identification of Cells of the Experimental Genotype
The marker for the experimental component of the chimera (which was PDGFRß+/+, PDGFRß+/-, or PDGFRß-/-) is a non-expressed 1000-copy tandem repeat of a mouse ß-globin/pBR322 construct23
which can be detected by non-isotopic in situ hybridization using digoxygenin-labeled probes as previously described.20
Nuclei were counter-stained with methyl green. Our primary interest was to evaluate vascular SMCs and adventitial fibroblasts. To identify and exclude leukocytes from our evaluation, all sections were immunostained for the pan-leukocyte marker CD45, using anti-mouse CD45 (clone 30-F11, PharMingen, San Diego, CA) followed by biotinylated rabbit anti-rat IgG then Vectastain elite ABC peroxidase (Vector Laboratories, Burlingame, CA) and diaminobenzidine. Peroxidase activity was then quenched by 30 minutes in 3% hydrogen peroxide after which the in situ hybridization was begun. We also identified SMCs by direct immunostaining for smooth muscle 
-actin combined with cell shape. This approach identified the same numbers of cells identified by cell shape and non-CD45-positivity, but was more difficult to score for the nuclear genotype marker. We obtained comparable results using the two methods for identifying SMCs.
Data Analysis and Normalization
500 to 1000 cells of each type of interest from sections of left carotid (ligated) and right carotid (uninjured control) were counted as either marked (experimental component) or unmarked (WT component) by the pBR322-targeted in situ hybridization. To control for interindividual differences in percent initial carotid artery chimerism, the values for the media and neointima in the ligated left carotid were compared to the value for the media of the unmanipulated right carotid in the same individual, and results are expressed as the "relative abundance," ie, the ratio of the marked percentage in the ligated left carotid to the marked percentage of the same cell type in the unmanipulated right carotid. Comparisons between groups were evaluated for statistical significance using Student’s t-test.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The most thoroughly studied model of arterial injury in the mouse, whose small size makes mechanical injury relatively difficult to control, is ligation of the left common carotid artery just proximal to the carotid bifurcation.24 This results in formation of a neointima containing SMCs and leukocytes. We identified the leukocytes by immunostaining for CD45 and excluded them from our analysis. The adventitia of the ligated carotid also thickens and there is an increase in cellularity, constituted of increased numbers of fibroblasts and CD45-positive leukocytes. At 4 weeks after ligation, most of the adventitial cells are fibroblasts, with smaller numbers of leukocytes.
We created chimeric mice by in vitro aggregation of early embryos (morula stage) from two component mouse lines. The "experimental component" was marked by the homozygous presence of an expression-independent DNA marker23
and was either PDGFRß+/+, PDGFRß+/-, or PDGFRß-/-. The control component was unmarked and PDGFRß+/+, ie, WT. The resulting chimeric mice are designated: PDGFRß+/+ WT, PDGFRß+/- WT, or PDGFRß-/- WT. We used these chimeras to compare, in the same individual, the behavior of WT and PDGFRß-deficient cells in response to carotid injury. The PDGFRß+/+ WT chimeras served as a control for possible differences in cell behavior that did not result from difference in PDGFRß expression, ie, that resulted from some other difference between the two parental lines. The media of an injured carotid in a chimeric mouse contained comparable numbers of cells in both PDGFRß-/- WT and WT WT chimeras (data not shown).
We used the value of marked/unmarked cells in the right (uninjured) carotid as a reference for the value that would have been found in the left (injured) carotid vessel before the experimental injury. This is justified by previous data, in which we determined that marked/unmarked cells make the same relative contribution to different arteries within an individual mouse.20
It is supported in this report by comparisons between vessels in PDGFRß+/+ WT chimeras: the percentage of medial SMCs (and adventitial fibroblasts) from the experimental line in the ligated carotid is equal to percentage in the contralateral carotid (ratio of 1, Figure 1
), and the percentage of marked to unmarked SMCs in the neointima is equal to the percentage in the contralateral media (ratio of 1, Figure 1
).
|
In rodent models of arterial response to injury, many publications demonstrate that SMCs proliferate in the injured media and migrate into the intima, where they continue to proliferate to form the neointima.1,24,25
What role does PDGFRß play in these processes? If medial SMCs respond to PDGF (via PDGFRß) largely as a mitogen, then the ratio of PDGFRß-/- to WT SMCs in the media should decrease after injury, as the WT SMC proliferate and the PDGFRß-/- SMCs do not. If PDGFRß mediates largely a chemotactic response, then the ratio should increase, as the WT SMCs migrate out of the media into the intima and leave the PDGFRß-/- cells behind. The data in Figure 1
, and illustrated in Figure 2
, (many "globin-marked" SMC nuclei in the media of injured carotid compared to uninjured control carotid and very few globin-marked SMC in the neointima) support the hypothesis that PDGFRß is predominantly mediating a chemotactic response. In PDGFRß-/- WT chimeras there is an approximately 4.5-fold increase (P < 0.01) in the percentage of PDGFRß-/- SMCs remaining in the ligated media as compared to the control media (Figure 1)
.
|
documents that this is the case. In both PDGFRß+/- WT and PDGFRß-/- WT chimeras, the intima is composed of approximately 40% fewer PDGFRß-deficient SMCs then in the uninjured contralateral media. In control PDGFRß+/+ WT chimeras there is no difference between the relative abundance of marked and unmarked cells (representing the two components of the chimera) in the injured media, the contralateral media, or the neointima. This demonstrates that the difference in the PDGFRß-/- WT chimeras was due to absence of PDGFRß expression rather than to some other difference between the two lines used to generate the chimeras. In PDGFRß+/- WT chimeras, in which the experimental component is heterozygous for PDGFRß and expresses half the normal level of PDGFRß, there is an intermediate magnitude of change (1.75-fold increase in the media, P < 0.01). The dose dependence of behavior on PDGFRß expression level was also observed during SMC development20
and fibroblast participation in granulation tissue formation,22
and demonstrates how sensitive SMCs are to relatively small differences in PDGFRß expression level. This supports the possibility that the magnitude of changes in level of ligand and receptor expression that have been observed in various pathophysiologic responses could be of consequence in fine tuning cell responses.26 The Participation of Adventitial Fibroblasts in the Fibrotic Response Is Not Affected by PDGFRß Expression
In addition to the formation of a neointima, carotid injury stimulates changes in the adventitial connective tissue layer around the vessel, which becomes thicker and more cellular. Leukocytes are abundant in the adventitia soon (1 week) after injury, but by 4 weeks, the adventitia is composed largely of an expanded population of fibroblasts and a more abundant extracellular matrix. To determine whether PDGFRß mediates this fibrotic response in the adventitia, we compared the behavior of PDGFRß-deficient fibroblasts to WT fibroblasts. As shown in Figure 1
, the relative numbers of PDGFRß-deficient fibroblasts did not change in response to ligation in either the PDGFRß+/+ WT, PDGFRß+/- WT or PDGFRß-/- WT chimeras. This demonstrates that the increased number and density of adventitial fibroblasts is not dependent on PDGFRß expression.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The formation of a neointima, and the remodeling of the media in response to injury, are known to involve migration, proliferation, and death.7-12,25,27 Multiple growth factors have been implicated driving these processes, but it has been difficult to determine which growth factor is responsible for driving which process. The chimera analysis results presented above show that, during the formation of the neointima, the proportion of PDGFRß-deficient SMCs increases in the media and decreases in the neointima. The simultaneous disappearance of medial and appearance of intimal PDGFRß-expressing SMCs suggests that the PDGF B-chain/PDGFRß system is acting largely as a chemotactic agent, to drive migration of SMC from the media to the neointima.
If PDGFRß is mediating SMC migration from the media to the neointima, why isn’t the neointimal population enriched in PDGFRß+/+ cells to the same extent that the media is depleted (compare Figure 1, A and B
)? We suggest two non-exclusive explanations for this: 1) Technical limitations resulting from failure to distinguish between leukocytes and SMCs. Because PDGFRß-/- SMCs are already under-represented in the media before injury (due to selection for WT SMCs during development), it is more difficult to precisely measure further decreases in the neointima than to measure an increase in the media. In addition, since the neointima contains many leukocytes, and since PDGFRß-/- leukocytes are not selected against during development,20
the occasional mis-identification of a leukocyte with a SMC would have a proportionately large effect on the calculated ratio of PDGFRß-/- to WT SMCs in the neointima compared to the media, where there are few leukocytes. 2) A more interesting explanation is that the vessel wall is not a closed system for SMCs, and that circulating progenitors may contribute to the SMC population in the neointima. We have used the nuclear marking system in hematopoietic chimeras to demonstrate that up to 11% of the endothelial cells in newly formed capillaries are derived from the hematopoietic compartment.28
In that study, we did not directly evaluate SMC origin from circulating hematopoietic cells, but cell culture evidence supports the existence of progenitors with capacity to become either endothelial cells or SMCs,29
and other investigators, using Y chromosome markers, have suggested that SMCs can derive from circulating progenitors under certain pathological situations.30,31
We plan to directly investigate this possibility in future studies.
Does PDGFRß Expression Also Affect SMC Apoptosis and Proliferation?
We attempted to determine this by directly measuring the relative proliferation and apoptotic rates of PDGFRß+/+ versus PDGFRß-/- cells in injured chimeric vessels. In both cases, we were stymied by a technical limitation created by the nature of our nuclear genotype marker. We found that we could not, in practice, reliably distinguish the nuclear genotype marker while simultaneously immunostaining for nuclear BrdU incorporation to evaluate proliferation, or nuclear TUNEL staining to evaluate apoptosis. Nevertheless, the following arguments suggest that PDGF effects on proliferation and apoptosis are unlikely to play major roles in the changes that we observed.
If PDGFRß signaling was stimulating medial SMC proliferation, we would expect the proportion of PDGFRß deficient medial SMCs to decrease rather than increase. This suggests that any action of PDGF to stimulate proliferation of medial PDGFRß+/+ SMCs is overwhelmed by its action as chemoattractant, drawing PDGFRß+/+ SMCs out of the media. This conclusion is consistent with earlier observations that inhibition of PDGF using a neutralizing antibody did not affect SMC proliferation in the injured rat carotid artery7 and that exogenous PDGF only minimally increased intimal and medial SMC proliferation.9 Thus, although many publications over the last 20 years document the ability of PDGF to stimulate proliferation of cultured SMCs and fibroblasts, this does not appear to be the predominant activity of PDGF in vascular response to injury in vivo.
A substantial frequency of SMC apoptosis has been reported in the ligated carotid model24,25 and in other arterial injury models.1,27,32,33 PDGF is generally reported to be anti-apoptotic for cultured SMCs,34-36 and blocking PDGF-mediated signaling in the rat carotid injury model has been reported to increase SMC apoptosis.32 If PDGF played a predominant role in inhibiting SMC apoptosis in the injured media, we would expect to find an increase, rather than the decrease we observed, in medial PDGFRß-deficient SMCs in the injured chimeric vessels. This argues that the anti-apoptotic effect must be small relative to the effect on chemotaxis. However, under conditions in which proliferation is prevented or limited, PDGF has been reported to be pro-apoptotic.36,37 Our results would be consistent with a pro-apoptotic effect of PDGF in this injury model.
PDGFRß Does Not Mediate Responses of Fibroblasts in the Adventitia
In contrast to the importance of PDGFRß expression for SMC migration, we found that PDGFRß expression did not affect the kinetics of adventitial fibroblasts. The proportion of PDGFRß-deficient fibroblasts did not decrease, as we would have expected if either PDGFRß+/+ fibroblasts were drawn into the adventitia and/or proliferated more readily; nor did the proportion of PDGFRß deficient cells increase as we would have expected if PDGFRß+/+ cells were drawn out of the adventitia. This result may seem surprising, since, for cultured fibroblasts, PDGF is active as a direct mitogen and chemotactic agent,2-4 and since chimera analysis of granulation tissue formation revealed that fibroblast participation depended on PDGFRß expression.22 In that study, in which we induced granulation tissue by implanting a subcutaneous sponge, we found that fibroblasts that accumulated within the sponge showed a greater dependence on PDGFRß expression than did fibroblasts that formed the capsule around the sponge.22 This is consistent with PDGF action as a chemoattractant, since fibroblasts that penetrate the sponge matrix need to travel farther than fibroblasts that remain at the periphery. Since fibroblasts are already present within the normal adventitia, there is no need for migration and no opportunity for selection based on the role of PDGFRß expression in driving migration. The adventitial fibroblast proliferation and fibrosis that occurs in this model is thus largely independent of PDGFRß expression and activation.
Advantages and Limitations of Chimera Analysis of Vascular Response to Injury
PDGF can stimulate increased expression of other growth factors and cytokines38
and it affects extracellular matrix synthesis and degradation.8,12,39,40
These secondary effectors could then act locally or systemically, to contribute to the overall consequences of PDGF action in the vessel wall. Because it detects only cell-autonomous effects, chimera analysis makes it possible to distinguish pathophysiologic changes that are direct effects of PDGF on PDGF-responsive cells, from changes that are downstream consequences of the action of secondary effectors. The chimera analysis reported here indicates that PDGFRß mediates a direct, cell-autonomous, effect in driving SMC migration in response to arterial injury. However, chimera analysis does not reveal a role for PDGFRß and PDGF B-chain in regulating the final cellularity/architecture of the artery. The media of PDGFRß-/- WT and PDGFRß+/+ WT chimeric carotids contain comparable numbers of SMCs, both before and after injury, despite the fact that the PDGFRß-/- SMC are unable to respond via PDGFRß. Similarly, hematopoietic chimeras, in which PDGF B-chain is eliminated in platelets and macrophages, still develop relatively normal granulation tissue41
and fibrous neointimal lesions.10
Finally, the size and structure of larger arteries and veins in late term non-chimeric PDGFRß-/- embryos is relatively normal,17
despite the complete absence of PDGFRß expression. Chimera analysis thus confirms the conclusions of studies using inhibitors of the PDGF/PDGF receptor system: PDGF plays a direct role in driving SMC migration in response to injury, and in determining the rate at which the vessel establishes a new architecture, but the final size/cellularity of the vessel must be determined by other factors.
|
Footnotes |
|---|
Supported by the National Institutes of Health (grant HL03174 to D.F.B.-P.).
Accepted for publication May 28, 2003.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
granule proteins for fibroblasts. J Cell Biol 1983, 96:382-385
B is a target of AKT in anti-apoptotic PDGF signalling. Nature 1999, 401:86-90[Medline]
This article has been cited by other articles:
 |
|
 |
|
  J. Andrae, R. Gallini, and C. Betsholtz Role of platelet-derived growth factors in physiology and medicine Genes & Dev., May 15, 2008; 22(10): 1276 - 1312. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Thornton, D. J. Swanson, M. N. Mary, D. Pei, A. C. Martin, S. Pounds, D. Goldowitz, and S. X. Skapek Persistent Hyperplastic Primary Vitreous Due to Somatic Mosaic Deletion of the Arf Tumor Suppressor Invest. Ophthalmol. Vis. Sci., February 1, 2007; 48(2): 491 - 499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Mallawaarachchi, P. L. Weissberg, and R. C. M. Siow Antagonism of platelet-derived growth factor by perivascular gene transfer attenuates adventitial cell migration after vascular injury: new tricks for old dogs? FASEB J, August 1, 2006; 20(10): 1686 - 1688. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Bujo and Y. Saito Modulation of Smooth Muscle Cell Migration by Members of the Low-Density Lipoprotein Receptor Family Arterioscler. Thromb. Vasc. Biol., June 1, 2006; 26(6): 1246 - 1252. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. P. Kim, X. Wang, A. Nakao, S. I. Kim, N. Murase, M. E. Choi, S. W. Ryter, and A. M. K. Choi Caveolin-1 expression by means of p38{beta} mitogen-activated protein kinase mediates the antiproliferative effect of carbon monoxide PNAS, August 9, 2005; 102(32): 11319 - 11324. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Zhuang, A.-C. Ceacareanu, B. Ceacareanu, and A. Hassid Essential role of protein kinase G and decreased cytoplasmic Ca2+ levels in NO-induced inhibition of rat aortic smooth muscle cell motility Am J Physiol Heart Circ Physiol, April 1, 2005; 288(4): H1859 - H1866. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Banai, S.D. Gertz, L. Gavish, M. Chorny, L. S. Perez, G. Lazarovichi, M. Ianculuvich, M. Hoffmann, M. Orlowski, G. Golomb, et al. Tyrphostin AGL-2043 eluting stent reduces neointima formation in porcine coronary arteries Cardiovasc Res, October 1, 2004; 64(1): 165 - 171. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
