© 1999 American Heart Association, Inc.
Original Contributions |
Studies on the Histogenesis of Myxomatous Tissue of Human Coronary Lesions
From the Department of Experimental Pathology, J.H. Holland Laboratory, American Red Cross, Rockville, Md (A.V.T., N.M.A., E.P.S., Y.G., C.C.H.), and the Washington Hospital Center, Washington, DC (M.H.K., M.B.L.).
Correspondence to Christian C. Haudenschild, Department of Experimental Pathology, J.H. Holland Laboratory, American Red Cross, 15601 Crabbs Branch Way, Rockville, MD 20855. E-mail HaudenschildC{at}usa.redcross.org
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract—Myxomatous tissue is a characteristic component of human coronary artery lesions, found more often in restenotic lesions. It represents a bulky accumulation of stellate-shaped cells of unknown histogenesis that are embedded in a loose stroma. We analyzed 64 atherectomy specimens containing substantial amounts of myxomatous tissue by using immunohistochemistry, in situ hybridization, and electron microscopy techniques. Stellate cells represented a heterogeneous population, sharing features of smooth muscle cells (SMCs), macrophages, as well as antigen-presenting dendritic cells. Like quiescent medial SMCs, the stellate cells in all specimens expressed high levels of SM
-actin message and protein and
showed heterogeneity with respect to heavy-chain
myosin, SM22, desmin, and vimentin. Ultrastructurally, stellate cells
resembled SMCs, with some peculiarities that distinguish them from both
differentiated and dedifferentiated SMCs. In contrast to quiescent
SMCs, the stellate cells expressed high levels of acidic fibroblast
growth factor mRNA and protein similar to cells of
monocyte/macrophage lineage. However, stellate cells did not
express the marker of mature macrophages, HAM56, and were
heterogeneous with respect to CD68. Moreover, unlike SMCs,
the stellate cells bore some of the major phenotypic markers of
dendritic cells: they were S100-positive and showed various reactivity
with respect to CD1a and human leukocyte antigen (HLA)-DR. Invasion of
myxomatous tissue with CD45RO-positive T lymphocytes was correlated
with strong expression of CD1a in these specimens. Stellate cells also
expressed a pericyte marker, high-molecular-weight melanoma-associated
antigen. We conclude that stellate cells of myxomatous tissue
represent a specific phenotype of mesenchymal cells
(possibly pericytes), which is activated to express some
markers of antigen-presenting cells. These findings suggest
involvement of the stellate cells in a local immune response.
Key Words: atherosclerosis • angioplasty • smooth muscle cell phenotype • dendritic cells • pericytes
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Proliferation of smooth muscle cells (SMCs) and accumulation of extracellular matrix (ECM) are generally viewed as the main mechanisms leading to intimal hyperplasia, which plays a role in atherogenesis and even more so in restenosis after angioplasty and stenting.1 Cellular components of intimal hyperplasia include phenotypically altered SMCs and activated macrophages,1 but many cells in human lesions do not fit this simple classification. An important characteristic component of human intimal hyperplasia is the so-called myxomatous tissue, also named "loose neointima" and "fibroproliferative tissue." It is defined as regions of moderate cellularity in which stellate-shaped cells are dispersed in a random fashion in abundant ECM, or "myxoid" stroma, composed mainly of proteoglycans and sparse, loosely arranged collagen fibrils.2 3 4 5 6 7 Myxomatous tissue comprises large parts of restenotic lesions of human coronary arteries after angioplasty, but it is also present in atherectomy specimens from primary lesions.4 5 8 9 The matrix of the myxomatous tissue is positive for the proteoglycans versican and biglycan, and its pattern of unique collagen staining suggests a collagen type I and type III fiber network.6 7 10 Because many cells in intimal hyperplasia are positive for SMC-specific
-actin, desmin, and vimentin,8 9 11 stellate
cells are thought to be typical SMCs. However, despite its unique
morphology, in most analyses of atherectomy specimens the
myxomatous tissue is not distinguished from other types of intimal
hyperplasia,4 5 8 9 12 and the nature of stellate
cells remains unclear. Besides SMCs, immunohistochemical analysis shows significant accumulation of macrophages and T lymphocytes in the regions of intimal hyperplasia.13 14 15 These cells are the local source of multiple cytokines, which affect SMC proliferation and ECM production in the lesion.1 16 In this respect, it is interesting that cells of stellate morphology can be found in the regions of organizing thrombus, in the peripheral vasculature, and especially in the atrium, ie, in the areas usually invaded by blood-borne and endothelial cells.2 17 18 Therefore, an alternative source of stellate cells could be macrophage/monocyte–related cells that could invade thrombi or inflammatory sites and acquire a stellate morphology. Obvious candidates are migrating, star-shaped dendritic cells (DCs) that arise from bone marrow progenitors and share a common ancestor with monocytes/macrophages.19 They are widely distributed in the skin epidermis (Langerhans cells [LCs]), the cardiovascular system (interstitial DCs), the cortical zones of lymph nodes (interdigitating follicular cells), and also in the peripheral blood.20 21 22 23 DCs express a number of unique markers that distinguish them from regular monocytes/macrophages.21 24 25 26 27 DCs function as antigen-presenting cells and are involved in various immune system–mediated, neoplastic, and infectious diseases.21 25 28
Another alternative source of origin of stellate cells could be mesenchymal cells with temporary differentiation as pericytes or myofibroblasts. There are reports that migrating adventitial myofibroblasts may contribute to the process of lesion formation after balloon overstretch injury in a swine model.29 30 31 Studies in the rat model suggest that pericytes could be a normal component of the arterial media.32 33 These cells have very high phenotypic plasticity and are able to express major phenotypic markers of SMCs on activation.34 35 Although the relationship between myofibroblasts/pericytes and SMCs is unclear, it has been suggested that pericytes are specific precursors of SMCs.33 35
In the present study, focusing on the myxomatous tissue only and applying various complementing technical approaches, we made an attempt to elucidate the origin of stellate cells in human coronary lesions. On the basis of results of immunohistochemistry, in situ hybridization, and electron microscopy, we found that stellate cells, although expressing all SM markers tested, in several respects differ from both quiescent medial SMCs and phenotypically altered SMCs of intimal hyperplasia. We demonstrate that in addition to SMC markers, they express some of the antigens that are considered typical for macrophages and DCs. Although it is difficult to unequivocally attribute stellate cells to 1 single lineage, we conclude that stellate cells represent a specific functional phenotype of mesenchymal cells that may be related to pericytes/myofibroblasts, with a possibly important function of antigen presentation.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Tissue Processing and Characterization
Coronary atherectomy specimens were obtained from 64 patients (mean age, 64±11 years) who underwent conventional directional coronary atherectomy at the Washington Hospital Center in 1993 to 1995. Specimens were flushed from the instrument chamber with saline solution and fixed immediately with 10% buffered formalin for 1 hour, followed by immersion in 70% ethanol. After dehydration, the tissue was embedded in paraffin. Criteria for identification of myxomatous tissue were the following: (1) predominance of cells with a characteristic stellate shape after hematoxylin and eosin staining for general morphology (Figure 1C
); (2)
loose, myxomatous ECM staining strongly positive for proteoglycans
(blue) in a modification of Movat's pentachrome stain that
employs a saffron/Alcian blue combination36
(Figure 1A and 1B); and (3) absent or weak staining for mature collagen
fibers in Movat's stain (yellow) and Fraser-Lendrum stain for fibrin
(green).37 The regions that met these criteria
were outlined manually on screen with direct microscopic visual
verification, and the areas were subjected to computer morphometry
using Image-1 software.38 For this study, we
selected 64 of 200 specimens, in which the myxomatous tissue occupied a
substantial part (at least a quarter) of the cross-sectional area of
the entire atherectomy. In the rare instances of equivocal boundaries,
the typical star-shaped cell morphology determined inclusion into this
study because it focused on stellate cells. Most of the specimens (55
of 64) were primary lesions; 9 specimens were restenotic
lesions.
|
Control Tissues
To compare stellate cells with the contractile phenotype
of SMCs, we used the medial layer of human aortas (3 different
individuals); ie, the tissue composed of the cells that by definition
are quiescent SMCs (according to their anatomic position within the
internal and external elastic laminas, spindle shape, close parallel
arrangement, and separation by continuous elastic laminas and collagen
fibers). These intact comparison specimens also displayed a typical
complicated atherosclerotic plaque with a sclerotic core and a fibrous
cap, which was used as a known staining control for activated
macrophages and foam cells. In each experiment, 1 of the 3
aortas was included as a standard tissue with the staining pattern
known over several years of use. In addition, almost all atherectomy
specimens had regions of normal media that could be distinguished by a
dense, parallel arrangement of spindle-shaped cells after
hematoxylin/eosin staining and the presence of the elastica interna as
shown by Verhoeff–Van Gieson's elastin stain. These regions served as
built-in controls for quiescent SMCs.
In experiments comparing stellate cells with DCs, we used as a positive control the tissue from which Langerhans first described and defined the DC, ie, the suprabasal layer of skin epidermis.39 Because it is known that the number of LCs is increased with inflammation, we used a specimen of human skin characterized by mild inflammation with leukocyte infiltration in the dermis. This specimen was also run in each experiment as a standard for the intensity and quality of the staining known over several years of use.
Electron Microscopy
Atherectomy tissues were flushed from the atherectomy device
with 0.9% saline, immediately immersion-fixed in 1% formaldehyde and
2.5% glutaraldehyde, rinsed with PBS, and stored in
PBS containing 5% sucrose. The tissues were postfixed in 1% aqueous
OsO4 and stained en bloc with 0.2% uranyl
acetate. After dehydration, embedding in Epon, sectioning, and staining
with uranyl acetate and lead citrate, the tissues were examined with a
Phillips CM12 transmission electron microscope.
Immunohistochemistry
The indirect avidin-biotin horseradish peroxidase method (ABC
Standard and Elite, Vector Laboratories) was used.
Endogenous peroxidase activity was exhausted in
methanol–0.3% H2O2.
Antigen retrieval was performed in a microwave oven (800 W on
"high") for 15 minutes in the solution prepared as follows. Nine
milliliters of 0.1 mol/L citric acid was mixed with 41 mL of 0.1 mol/L
sodium citrate, the volume was brought to 500 mL, and pH was adjusted
to 6.0. The source and specificity of the primary antibodies are
summarized in Table 1. Mayer's
hematoxylin was used to counterstain the sections.
|
Hybridization In Situ
Nonradioactive in situ hybridization was performed on paraffin
sections with digoxigenin-labeled DNA oligoprobes (40 and 50 bp): human
vascular SM -actin, AGGCGCTGATCCACAAAACATTCACAGTTGTGTGCTAGAG; human
SM22, CATAAACCAGTTGGGATCTCCACGGTAGTGCCCATCATTC; and human CD1a,
GGAAGGCCCTCTGGAGTAAGGGATATCTAAAACAGGCCTGATGATCCATT.
Human fibroblast growth factor (FGF)-1 was detected using 450-bp
digoxigenin-labeled riboprobe. The oligoprobes were designed using GCG
software to select the unique regions and "tailed" with
digoxigenin-11-dUTP by using terminal transferase according to
established procedures.40 Digoxigenin-labeled
FGF-1 riboprobe (antisense and sense, courtesy of Dr J.A. Thompson) was
synthesized in a standard in vitro transcription
reaction.41 In some experiments, instead of
enzymatic treatment of the rehydrated sections, we used heat in a
variation of the antigen retrieval procedure to facilitate access of
the probe to the target. The antigen retrieval solution (see the
Immunohistochemistry section) was brought to boiling in a microwave
oven and heated for an additional 4 minutes, and the slides were left
to cool for 15 minutes. The slides were postfixed for 10 minutes in
freshly prepared 4% paraformaldehyde in PBS.
Prehybridization was performed at 37°C for 1 to 2 hours in a buffer
containing 50% formamide, 10% dextran sulfate, 5x SSC, 5x
Denhardt's solution, and 100 µg/mL blocking DNA (all reagents
purchased from Sigma Chemical Co). Hybridization buffer contained 50%
formamide, 10% dextran sulfate, 5x SSC, 1x Denhardt's solution, 100
µg/mL DNA, and 1 ng/µL of digoxigenin-labeled probe. Hybridization
was performed at 37°C for oligoprobes and at 50°C for the riboprobe
overnight in a moist chamber saturated with 4x SSC. To control
nonspecific binding of the probes, in the negative controls we either
omitted digoxigenin-labeled oligoprobes or used a sense FGF-1 riboprobe
that should not hybridize with mRNA. To detect the hybridization
product, sections were incubated with anti-digoxigenin antibody
(Fab fragment coupled with alkaline phosphatase; Boehringer
Mannheim) at a dilution of 1:500 in the blocking solution (1% blocking
reagent, Boehringer Mannheim, in 0.1 mol/L Tris-HCl, pH 7.5,
and 0.15 mol/L NaCl) at room temperature for 2 hours. The hybridization
products were visualized in color reaction for alkaline phosphatase
(45 µL nitro blue tetrazolium chloride, 35 µL
5-bromo-4-chloro-3-indolyl phosphate disodium salt, 2.4 mg levamisole
in 10 mL of 0.1 mol/L Tris-HCl (pH 9.5), 0.1 mol/L NaCl, and 0.05 mol/L
MgCl2) performed in the dark from 2 hours to
overnight at 4°C. The reaction was stopped by rinsing the sections in
10 mmol/L Tris-HCl and 1 mmol/L EDTA, pH 8.0. Nuclear fast
red was used to counterstain the sections.
Double Staining
To colocalize 2 markers, in several experiments we performed
immunohistochemistry after in situ hybridization. The solutions of
primary antibodies were applied after the color developing reaction
step of the first procedure (see above). These sections were not
counterstained.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
General Histology of Myxomatous Tissue
The myxomatous tissue retrieved by directional coronary atherectomy from lesions of human coronary arteries formed areas that could be clearly distinguished from other types of intimal constituents on hematoxylin and eosin staining by their predominant cell type. This tissue was composed of randomly oriented cells with a characteristic branching, "stellate" shape that formed a network in abundant, loosely organized ECM (Figure 1C
). This morphological pattern was
clearly distinct from that of densely cellular areas of intimal
hyperplasia with spindle-shaped cells in a dense fibrous matrix, and it
also differed clearly from the typical sclerotic plaque areas with
sparse, elongated cells in a matrix of very dense, tendonlike collagen.
Movat's stain of ECM components showed that stellate cells in the
myxomatous regions were separated by wide, proteoglycan-rich spaces
that stained intensely blue, in contrast to collagen-rich areas of
fibrocellular and sclerotic tissue, which stained yellow (Figure 1A and 1B). Mixtures of proteoglycan-rich and collagenous matrix were
occasionally seen, which stained green; some of these areas retained
multidirectional matrix architecture following the shapes of the
stellate cells and thus were included in this study, whereas areas with
mixed proteoglycan/collagen matrix containing spindle-shaped cells in a
parallel arrangement were counted as maturing fibrocellular SMC areas
and were not included. Transition between stellate and spindle cell
areas was infrequent. Normal media in atherectomy specimens was
recognized by the internal elastic boundary's staining black (Figure 1A). In trichrome staining, myxomatous areas were identified by
dissolution or "fuzzy" staining of the collagenous fibrous layers,
which were stained blue where present (data not shown).
Ultrastructural Features of Myxomatous Tissue
Stellate cells of myxomatous tissue had a characteristic
ultrastructural appearance: they showed a complex branching pattern of
condensed cytoplasm and an exaggerated, often multilayered, basal
membranes (Figure 2A). Many of the gray
patches in the matrix represented detached layers of
excessive basal membrane material. The punctate pattern of matrix
indicated abundant proteoglycans that could also be seen in some
cellular vacuoles. In contrast to typical fibrocellular lesions, mature
collagen fibrils were sparse. Stellate cells invariably showed bizarre
shapes of the nucleus (Figure 2B). They contained few synthetic and
metabolic organelles, which were concentrated in the
perinuclear area; instead, they displayed abundant actin filaments in
the cell body and in the multiple branches of the cell periphery
(Figure 2B). Occasional "dense bodies" were present within the
actin filaments, but no typical SMC adhesion plaques were observed.
|
SMC Markers
All stellate cells stained positive with a monoclonal antibody
recognizing SM -actin, which is considered the major phenotypic
marker of SMCs42 (Figure 3B). Quiescent medial cells of an aorta
(Figure 3A) and occasional areas of normal media in atherectomy
specimens (data not shown) that were used as positive controls also
displayed strong reactivity. In situ hybridization confirmed these
findings and showed that stellate cells expressed -actin mRNA
(Figure 3E) at a level comparable to or even higher than that in medial
SMCs in the control aortic media tissue (Figure 3D). Stellate cells
showed a heterogeneous pattern of expression (approximately
half of the cells were positive) for another marker of SMCs, myosin
heavy chain (Figure 3C). A similar pattern of myosin staining was
observed for quiescent medial SMCs found in some atherectomy specimens
(data not shown). SM22, a calponin-related protein, is specifically
expressed in adult SM and is recognized as one of the markers of
differentiated SMCs.43 44 In situ hybridization
showed 100% expression of SM22 by quiescent medial SMCs, both from
control aortas and from atherectomy specimens (Figure 3G). On the
contrary, <10% of stellate cells were positive for the SM22 probe
(Figure 3H); large areas of myxomatous tissue were completely negative.
Among intermediate filaments, desmin-type filaments are specific for
muscle cells and have not been observed in other cell
types,45 46 47 whereas vimentin-type filaments are
present in all kinds of mesenchymal cells, including vascular
cells.45 46 Approximately a third of stellate
cells were positive for desmin (Figure 3F), and 
80% of the stellate
cells were positive for vimentin (Figure 3I).
|
Expression of FGF-1 by Stellate Cells
In contrast to quiescent medial SMCs, all stellate cells
contained high levels of FGF-1 protein (Figure 4A). Lipid-laden foam cells and
monocytes/macrophages that had accumulated in inflammatory
regions were strongly FGF-1–positive (Figure 4C), whereas quiescent
medial SMCs (Figure 4B) as well as the spindle-shaped cells in the
typical densely cellular nonmyxomatous areas of intimal hyperplasia
(data not shown) were negative for FGF-1 protein. As shown by in situ
hybridization with FGF-1 riboprobe, a strong hybridization signal was
observed in the myxomatous areas outlining the characteristic shape of
stellate cells (Figure 4D). The specificity of hybridization was
confirmed by the absence of the signal with the sense probe. The
macrophages in the aortic control tissue also expressed FGF-1
mRNA, whereas medial SMCs were "silent" for FGF-1 message (Figure 4E). Thus, in situ hybridization demonstrated that the stellate cells
not only contained FGF-1 but also synthesized this growth factor.
|
Although FGF-1 is known to be produced in different cell types, our finding of FGF-1 in stellate cells is a major difference from quiescent medial SMCs and spindle-shaped intimal SMCs that were reported to express hardly detectable levels of FGF-1 mRNA or protein.1 48 49 50 51 52 The level of FGF-1 mRNA is increased in atheromas compared with normal human arteries, and in atheromas, immunoreactivity for FGF-1 is associated with areas of neovascularization (ie, endothelial cells and/or pericytes) and macrophage-rich regions.48 49 53 Stellate cells were negative with respect to CD31 antigen, a specific marker of endothelial cells,54 whereas capillary endothelium in the same section was strongly positive (data not shown). On the basis of these results, we ruled out a relationship between the 2 cell types. Thus, expression of FGF-1 by stellate cells could be an argument in favor of a monocyte/macrophage origin for these cells.
To test this hypothesis, we studied the expression of phenotypic
markers of differentiated macrophages, HAM56 and CD68. However,
the myxomatous tissue in atherectomy specimens proved to be negative
for HAM56 (Figure 4G) and so were the quiescent SMCs, whereas
monocytes, macrophages, and foam cells in the positive control
tissue (inflammatory region of an atherectomy specimen) stained
strongly (Figure 4H). Only 15% of the stellate cells expressed CD68
(Figure 4F), whereas monocytes, macrophages, and foam cell
macrophages in the inflammatory areas were strongly positive
(Figure 4I). Thus, by these criteria, stellate cells do not
represent mature monocytes or differentiated
macrophages.
DC Markers
We then studied expression in the myxomatous tissue of
several phenotypic antigens of another type of bone marrow–derived
cell sharing a common ancestor with monocytes, ie, DCs. This cell
family is a phenotypically and functionally diverse population of cells
that actively migrate from bone marrow via the blood circulation to the
peripheral organs, where they function as
antigen-presenting cells responsible for induction of primary
immune responses.21 25 28 A typical member of
this family, LCs, are usually distributed in the suprabasal layer of
the epidermis.21 These cells display major
histocompatibility complex (MHC) class I and II antigens (CD1a and
human leukocyte antigen [HLA]-DR), S100 antigen, and several adhesion
molecules.21 24 25 26 27
Approximately 40% of the stellate cells were HLA-DR–positive
(Figure 5A). In the control tissues both
monocytic and foam cell macrophages expressed high levels of
HLA-DR (not shown), whereas the highly organized, spindle-shaped,
quiescent SMCs of a piece of normal media in the same specimen were
negative (Figure 5B). In the inflamed skin used as a positive control
specifically for LCs, anti–HLA-DR antibody stained these cells as well
as macrophages and monocytes (Figure 5C).
|
The stellate cells uniformly and strongly expressed another phenotypic
marker of LCs, S100 antigen (Figure 5D), whereas quiescent SMCs in
normal media from the same specimen did not (Figure 5E). In the skin,
anti-S100 antibody stained LCs as expected, as well as melanocytes and
Schwann cells, but did not stain vascular SMCs,
endothelial cells, leukocytes, and plasma cells (Figure 5F).
However, it is recognized that the major phenotypic marker
defining LCs is CD1a antigen. We first applied the in situ
hybridization technique for detection of CD1a mRNA. The oligoprobe was
selected from the 5'-untranslated region specific for CD1a mRNA that is
absent in CD1b or CD1c mRNA.55 In 60% of the
specimens (22 of 36) analyzed, we detected the hybridization
product in the myxomatous areas in variable numbers of cells
(Figure 6D). In contrast, quiescent SMCs
in the areas of normal media from the same atherectomy specimens were
CD1a-negative (Figure 6E). As expected in the skin (positive control
tissue), LCs in the suprabasal layer of the epidermis showed the
hybridization signal (Figure 6F). This pattern was confirmed by
immunohistochemistry with the anti-CD1a monoclonal antibody O10. We
observed distinct cytoplasmic staining of many, though not all,
stellate cells of the myxomatous regions (Figure 6A). Similarly in the
skin, positive LCs scattered between keratinocytes in
the epidermis could be recognized by their characteristic shape (Figure 6C). As expected, areas of normal media of the atherectomy specimens
were always negative with respect to this antigen (Figure 6B).
|
Double Staining
Immunohistochemical staining of specimens for HLA-DR after in situ
hybridization for CD1a revealed a small subpopulation (10% to 15%) of
double-positive cells (Figure 7A).
However, for the remainder of labeled cells, these two signals did not
overlap. Interestingly, staining for -actin after CD1a in situ
hybridization demonstrated coexpression of both antigens in every
CD1a-positive stellate cell (Figure 7B).
|
If the expression of the markers of antigen-presenting LCs in
stellate cells is functionally relevant, then some evidence of antigen
presentation, such as an accumulation of T cells in
myxomatous tissue, should be found. Indeed, staining for the
T-lymphocyte marker CD45RO showed a frequent lymphocyte infiltration in
myxomatous tissue, albeit not as cellular as in the
macrophage-rich, inflammatory areas. Interestingly,
CD45RO-positive lymphocytes were always present in the vicinity of
the stellate cells, which also expressed CD1a (Figure 7C).
Pericyte Markers
We also investigated a possible relationship between stellate
cells and pericytes. Immunohistochemical identification of pericytes is
complicated owing to their phenotypic, biochemical, and
physiological
heterogeneity.34 35 Although a
number of specific markers have been detected on pericytes, they are
not continually expressed and are dependent on either a functional or
metabolic state of the cells.34 We
tested stellate cells for expression of a nonmuscle isoform of actin,
which is dominant in pericytes of "true"
capillaries,56 and melanoma-associated antigen
(MAA), discovered on pericytes in tumors.57
Stellate cells were positive for ß-actin; however, the signal was
comparable or weaker than that of medial SMCs (data not shown). With
respect to MAA, in our archival atherectomy specimens the antigenic
reactivity was inhibited by overfixation and the paraffin-embedding
procedure. Therefore, we studied expression of this antigen in freshly
frozen sections of a coronary artery of a human donor heart
that had been rejected for organ transplantation. The left anterior
descending coronary artery, incidentally, had a lesion
containing almost exclusively stellate cells. In these frozen, left
anterior descending coronary artery sections, the antibody
MEL-2 against MAA reacted with pericytes of adventitial capillaries,
arterioles, and postcapillary venules as expected, and we also observed
a strong signal in the stellate cells. We also found abundant stellate
cells in myxomatous endocardial tissues covering infarcted
myocardium in an autopsy specimen (Figure 8A), which gave the same distinct
cytoplasmic staining for MAA, outlining their stellate morphology
(Figure 8B). The deeper layer of the thickened endocardium, mainly
composed of spindle-shaped cells in more collagenous scar tissue, was
negative for this marker. A similar staining pattern was observed for
ß-actin (Figure 8C). Thus, expression of these antigens by stellate
cells may be an argument for their relationship to pericytes.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Myxomatous tissue represents a subtype of intimal hyperplasia in human coronary lesions. It is composed of stellate-shaped cells randomly arranged in abundant, loosely organized, proteoglycan-rich ECM. Several reports connect this type of intimal hyperplasia with restenotic lesions.4 5 8 9 Our findings on 200 atherectomy specimens show that myxomatous tissue can be found in 85% of restenotic lesions versus 50% of primary lesions and that in restenotic lesions it occupies a larger cross-sectional area (C.C.H. et al, unpublished data, 1998). Cells with a similar stellate shape are occasionally observed in the subendothelium of normal and atherosclerotic intima of human aortas, suggesting that these cells may be a normal, albeit rare, component of the vessel wall.58 59 60 Although this peculiar stellate cell type was first described in the normal aortic intima as early as in 186661 and was found to be increased there in the inflammatory state,62 the nature of arterial myxomatous tissue is still poorly understood. As a component of intimal hyperplasia, stellate cells could originate from SMCs. This hypothesis is based on the observation that many cells of hyperplastic tissue are positive for the SMC markers actin, desmin, and vimentin. However, in most studies the myxomatous tissue was not studied separately from other types of hyperplasia.4 5 8 9 12
When focusing on myxomatous tissue only, our study provided many
arguments that stellate cells cannot be considered typical SMCs,
although they do express phenotypic SM markers (Table 2). We demonstrated that -actin, the
major phenotypic marker of quiescent SMCs,42 was
indeed abundant in all stellate cells both as message and as protein.
Other SM markers such as SM-specific heavy-chain myosin, SM22, desmin,
and vimentin showed a heterogeneous pattern of
distribution. A similar distribution of these markers was observed in
quiescent SMCs of normal aortic media and media fragments in
atherectomies used as positive controls and was consistent with
earlier reports.44 45 46 47 63 64 65 However, a
significant difference was observed for SM22: although this marker was
strongly expressed by all medial SMCs, only 10% of stellate cells
were positive.
|
Electron microscopy analysis demonstrated that the cytoplasm of stellate cells is filled with bundles of thin filaments, showing occasional "dense bodies." The cells had prominent basement membranes, which were often multilayered. These ultrastructural features also point to a SM nature for stellate cells. However, they clearly differ from those of the so-called "contractile" phenotype described for differentiated, quiescent SMCs of the media that also have abundant, densely packed actin filaments but are spindle-shaped and surrounded by a thin basement membrane and mature, collagenous, and elastic matrix.66 67 On the other hand, stellate cells differ ultrastructurally from SMCs with the so-called "synthetic" phenotype,66 67 which is characterized by a distinct lack of contractile filaments and abundant synthetic organelles. Thus, the pattern of expression of SMC markers in stellate cells and their ultrastructural features suggest either another (non-"contractile," non-"synthetic") phenotype of SMCs or a different origin for these cells.
One of the striking peculiarities of stellate cells that distinguishes
them from typical SMCs is constantly high levels of FGF-1 message and
protein (Table 2). FGF-1 is expressed by several cell types involved in
atherogenesis. Although vascular SMCs can produce and respond to FGF-1
in vitro,68 the most abundant growth factor of
SMCs is FGF-2 but not FGF-1.1 48 49 50 51 Freshly
isolated rat SMCs express a very low level of
FGF-152; FGF-1 mRNA is hardly detectable in SMC
cultures from adult rat aortas, whereas in newborn SMC cultures, FGF-1
mRNA is expressed to a much greater degree.51
There is a differential pattern of FGF-1 expression in human
atherosclerotic and normal arteries. In particular, the level of FGF-1
mRNA is increased in human atheroma compared with
nonatherosclerotic arteries.48 49 53 In normal
human arteries, FGF-1 immunoreactivity was reported predominantly for
adventitial fibroblasts but not SMCs, whereas in atheromas
FGF-1 protein is associated with plaque microvessels (ie,
endothelial cells) and macrophage-rich
regions.49 53 Indeed, in our study, neither
medial SMCs, nor -actin–positive, spindle-shaped cells of
nonmyxomatous intimal hyperplasia in the atherectomy specimens were
FGF-1–positive at the threshold that could be detected with our
antibody. At the same time, with the strong expression of FGF-1 in
stellate cells, we detected high levels of FGF-1 in built-in control
areas of monocytic cells within macrophage-rich, inflammatory
regions of lesions as well as in lipid-laden, "foamy"
macrophages. This finding is a strong argument that stellate
cells could be related to cells of monocyte/macrophage lineage.
However, a classic marker of human tissue macrophages, HAM56,
was not present in the stellate cells, and CD68 was seen in only a
few stellate cells that made a clear distinction of stellate cells from
mature macrophages. We also excluded a possible relationship of
stellate cells to endothelial cells because the former
were not reactive to anti-CD31 antibody,54 which
specifically recognizes endothelial cells.
Next, we compared stellate cells of human atherectomies with DCs, which comprise a diverse family of bone marrow–derived cells able to migrate via the blood circulation to many peripheral locations of the body.20 21 22 24 Morphologically, DCs strikingly resemble stellate cells of myxomatous tissue: they are irregularly shaped, have long, cytoplasmic processes, and possess a lobulated nucleus. The earliest progenitor of DCs is a CD34-positive pluripotent stem cell of the bone marrow; the more mature intermediate cell precursors found in the peripheral blood give rise to either DCs or monocytes.22 27 The hypothesis that stellate cells of myxomatous tissue could represent DCs is especially attractive because vascular DCs were described in the human aortic intima: in normal aorta, they are mainly localized in the subendothelial layer, whereas in atheromas these cells are distributed throughout the lesion.23 69 70 71
We analyzed several of the most frequently expressed phenotypic
markers of DCs: CD1a, HLA-DR, and S100. We found expression of these
antigens in all or in a subpopulation of arterial stellate
cells in our atherectomy specimens (Table 2). One of the classic
markers of DCs, S100 antigen, was uniformly expressed by all cells of
myxomatous tissue and was present in neither quiescent medial SMCs
nor in monocytes/macrophages. Besides within LCs, S100 is found
in a number of cell types of the nervous system, melanocytes, and in
various tumor cells, including SM tumors.72 Two
other antigens, CD1a and HLA-DR, displayed a heterogeneous
pattern of expression ranging from completely negative areas to
predominantly positive regions of myxomatous tissue. While expression
of HLA-DR is normally restricted to cells of the immune system, such as
macrophages and T lymphocytes, it can also be induced in other
cell types, such as SMCs, endothelial cells,
fibroblasts, and various tumor cells.73 74 75 CD1a
belongs to the CD1 family of proteins that are primarily expressed on
the surface of cortical thymocytes and some T-cell
leukemias.76 77 Extrathymically, CD1a has been
observed on DCs, mainly LCs.78
Thus, we found a striking antigenic similarity between stellate cells
in human atherosclerotic lesions and DCs in their expression of
functionally relevant molecules, including CD1a and HLA-DR (Table 2).
However, for several reasons we believe that stellate cells rather
belong to a mesenchymal cell lineage characterized by expression of the
markers typical for macrophages and antigen-presenting DCs.
First, they express most SM markers tested, such as -actin, myosin,
desmin, and vimentin, except SM22, for which only 10% of stellate
cells were positive. There are no reports in the literature that
blood-borne cells, such as macrophages, or other nonmuscle
cells can express the SM markers -actin and desmin even under
pathological conditions or in culture.45 46 63 We
did not observe a signal for -actin in skin LCs. Second,
ultrastructural features of stellate cells most closely resemble those
of SMCs. Indeed, the ultrastructural appearance of these cells is very
different from that of macrophages, DCs, or
endothelial cells. Obviously, the ability of stellate
cells to synthesize basement membranes and ECM is more typical for SMCs
than for monocytic cells or DCs, which are characterized by a thin ring
of cytoplasm and lack dense bodies and a basement
membrane.20 71 We did not find Birbeck granules,
which are a characteristic ultrastructural feature of LCs, although at
certain stages of maturation DCs also do not have these
structures.21 22 27 79 Third, Hansson et
al80 and Jonasson et al81
reported that SMCs of atherosclerotic lesions can express HLA-DR
antigen and that this expression can be induced in vitro by
interferon-
from T lymphocytes.73 The finding
of S100 antigen in stellate cells also does not contradict the
hypothesis of their mesenchymal origin: besides in LCs, expression of
this antigen was reported for various tumor cells, including SM
tumors.72 Last, in our study, double staining for
CD1a and -actin confirmed that all CD1a-positive stellate cells were
also positive for -actin, arguing against a bone marrow origin, at
least for this part of the stellate cell population. In atherosclerotic
lesions, vascular DCs are reported to be mainly accumulated near the
neovasculature or within inflammatory infiltrates, establishing
contacts with each other or with macrophages, foam cells,
endothelial cells, or T
lymphocytes.82 83 Indeed, in inflammatory areas
of our specimens as well as in regions of angiogenesis, we could
confirm the presence of CD1a-positive cells. However, unlike scattered
DCs, the stellate cells of our study formed a tissue that was clearly
distinct from areas of both inflammatory infiltrates and
neovascularization.
Thus, by the ultrastructural peculiarities of stellate cells and their expression of SMC markers simultaneously with expression of several antigens that are not typical for SMCs, we consider stellate cells to be either an atypical (non-"contractile" and non-"synthetic") phenotype of SMCs or to belong to another, SM-like mesenchymal cell lineage. In this respect, an intriguing possibility for the origin of stellate cells is that they could be pericytes. Pericytes, or myofibroblasts, are pluripotent mesenchymal cells that can differentiate into adipocytes, osteoblasts, and phagocytes; it is also suggested that pericytes can be specific precursors of SMCs.33 35 Pericytes are morphologically, biochemically, and physiologically heterogeneous, and the only reliable criterion for their identification is localization next to a capillary.34 35 However, these cells share many features with SMCs, such as expression of actin, desmin, and vimentin and synthesis of collagens and proteoglycans.34 84 Pericytes contain both SM and non-SM isoforms of actin and myosin.56 85 On the other hand, it has been shown that brain pericytes can express some of the markers of macrophages.86 There are reports indicating the possible presence of pericytes within arteries.32 33 It has been suggested that migrating adventitial myofibroblasts may contribute to the process of lesion formation.29 30 31 Morphologically, typical pericytes have a stellate shape, although they can assume a fibroblast-like shape.34 87 Pericytes synthesize and release structural constituents of the basement membrane and ECM.34 In our study, we observed a heterogeneous staining of stellate cells for the nonmuscle isoform of actin that is supposed to be one of the markers of pericytes. This finding does not contradict earlier data: pericytes from "true" capillaries mainly have nonmuscle actin and myosin isoforms, whereas pericytes from postcapillary venules contain mainly SM isoforms.34 85 The stellate cells of human endocardium were shown to express MAA that is considered another surface marker of pericytes. Taken together, our findings, such as heterogeneous antigenic pattern and peculiar ultrastructural features of stellate cells, make a strong argument for a pericyte origin for myxomatous tissue.
Despite the arguments against a bone marrow origin, finding of HLA-DR and CD1a expression in stellate cells implies that functionally, they could be compared with DCs. Both molecules are known to be directly related to the antigen-presenting function of DCs: they capture antigenic peptides, and such antigen–MHC molecule complexes can be specifically recognized by T lymphocytes via T-cell receptors.21 24 25 26 88 We hypothesize that in atherosclerotic lesions, stellate cells might be involved in a local immune response, where they function as antigen-presenting cells. It has been shown that endothelial cells expressing HLA-DR are capable of presenting foreign antigens to T cells.74 89 90 The same ability has been proposed for HLA-DR–bearing SMCs.91 Interestingly, the CD1 family of molecules has been proposed as an important mechanism for the presentation of lipid antigens to T lymphocytes.92 On the other hand, it is known that activated memory T lymphocytes, to which the antigen is presented, invade atherosclerotic lesions in high numbers.93 Indeed, in most of our specimens we detected massive infiltration of CD45RO-positive lymphocytes. The highest cellularity was detected in inflammatory areas densely populated by macrophages, which are "professional" antigen-presenting cells. It is noteworthy that the myxomatous tissue was the second tissue type with frequent CD45RO-positive T-lymphocyte infiltration, whereas in all other arterial tissue types, normal and atherosclerotic, only rare CD45RO-positive T lymphocytes could be found. Although this observation does not directly prove that T lymphocytes are the inducers of CD1a and HLA-DR expression by stellate cells nor does it prove that stellate cells attract T cells, it suggests an interaction between these two cell types in the atherosclerotic/restenotic lesions. Further investigations are necessary to elucidate what the local immune response specifically implies and what the consequences are. One might speculate that the normal self-limitation of vascular wound healing may be affected by such interactions, leading to accumulation of myxomatous tissue with abnormal matrix in these foci.
Thus, on the basis of results of the present study, we propose that stellate cells of myxomatous tissue from human coronary lesions represent a specific phenotype of mesenchymal cells, which could originate from pericytes. Stellate cells acquire some features of antigen-presenting cells such as DCs and macrophages and could be involved in local immune reactions in atherosclerotic lesions. This type of activation may be an important mechanism of atherogenesis and restenosis.
Note added in proof. Randolph et al have recently reported that "DCs arise from monocytes that migrate across endothelium in an ablumenal-to-lumenal direction, whereas macrophages develop from monocytes that remain in the subendothelial matrix." These findings emphasize the existence of a common precursor for DCs and macrophages and the importance of local vascular wall interactions for their differentiation pathways. (Science. 1998;282:480–483.)
|
Acknowledgments |
|---|
This study was supported by National Institutes of Health (NIH) grant HL 54246. We are indebted to Dr John A. Thompson and Qiang Ding (Department of Surgery, University of Alabama at Birmingham) for supplying us with the digoxigenin-labeled FGF-1 riboprobe and help with in situ hybridization experiments.
Received February 20, 1998; accepted June 4, 1998.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature. 1993;362:801–809.[Medline] [Order article via Infotrieve]
- Rosenschein U, Ellis SG, Haudenschild CC, Yakubov SJ, Muller DW, Dick RJ, Topol EJ. Comparison of histopathologic coronary lesions obtained from directional atherectomy in stable angina versus acute coronary syndromes. Am J Cardiol. 1994;73:508–510.[Medline] [Order article via Infotrieve]
- Escaned J, de Jong M, Violaris AG, MacLeod DC, Umans VA, van Suylen RJ, de Feyter PJ, Verdouw PD, Serruys PW. Clinical and histological determinants of smooth-muscle cell outgrowth in cultured atherectomy specimens: importance of thrombus organization. Coron Artery Dis. 1993;4:883–890.[Medline] [Order article via Infotrieve]
- Miller MJ, Kuntz RE, Friedrich SP, Leidig GA, Fishman RF, Schnitt SJ, Baim DS, Safian RD. Frequency and consequences of intimal hyperplasia in specimens retrieved by directional atherectomy of native primary coronary artery stenoses and subsequent restenoses. Am J Cardiol. 1993;71:652–658.[Medline] [Order article via Infotrieve]
- Waller BF, Johnson DE, Schnitt SJ, Pinkerton CA, Simpson JB, Baim DS. Histologic analysis of directional coronary atherectomy samples: a review of findings and their clinical relevance. Am J Cardiol. 1993;72:80E–87E.[Medline] [Order article via Infotrieve]
- Riessen R, Isner JM, Blessing E, Loushin C, Nikol S, Wight TN. Regional differences in the distribution of the proteoglycans biglycan and decorin in the extracellular matrix of atherosclerotic and restenotic human coronary arteries. Am J Pathol. 1994;144:962–974.[Abstract]
- Wight TN, Lara S, Riessen R, Le Baron R, Isner J. Selective deposits of versican in the extracellular matrix of restenotic lesions from human peripheral arteries. Am J Pathol. 1997;151:963–973.[Abstract]
- Garratt KN, Edwards WD, Kaufmann UP, Vlietstra RE, Holmes DR Jr. Differential histopathology of primary atherosclerotic and restenotic lesions in coronary arteries and saphenous vein bypass grafts: analysis of tissue obtained from 73 patients by directional atherectomy. J Am Coll Cardiol. 1991;17:442–448.[Abstract]
- Schnitt SJ, Safian RD, Kuntz RE, Schmidt DA, Baim DS. Histologic findings in specimens obtained by percutaneous directional coronary atherectomy. Hum Pathol. 1992;23:415–420.[Medline] [Order article via Infotrieve]
- Lin H, Kanda T, Hoshino Y, Takase S, Kobayashi I, Nagai R, McManus BM. Versican, biglycan, and decorin protein expression patterns in coronary arteries: analysis of primary and restenotic lesions. Cardiovasc Pathol. 1998;7:31–37.
- Pickering JG, Weir L, Jekanowski J, Kearney MA, Isner JM. Proliferative activity in peripheral and coronary atherosclerotic plaque among patients undergoing percutaneous revascularization. J Clin Invest. 1993;91:1469–1480.
- Strauss BH, Umans VA, van Suylen RJ, de Feyter PJ, Marco J, Robertson GC, Renkin J, Heyndrickx G, Vuzevski VD, Bosman FT. Directional atherectomy for treatment of restenosis within coronary stents: clinical, angiographic and histologic results. J Am Coll Cardiol. 1992;20:1465–1473.[Abstract]
- van der Wal AC, Das PK, Tigges AJ, Becker AE. Macrophage differentiation in atherosclerosis: an in situ immunohistochemical analysis in humans. Am J Pathol. 1992;141:161–168.[Abstract]
- Gown AM, Tsukada T, Ross R. Human atherosclerosis, II: immunocytochemical analysis of the cellular composition of human atherosclerotic lesions. Am J Pathol. 1986;125:191–207.[Abstract]
-
Jonasson L, Holm J, Skalli O, Bondjers G, Hansson GK.
Regional accumulations of T cells, macrophages, and smooth
muscle cells in the human atherosclerotic plaque.
Arteriosclerosis. 1986;6:131–138.
[Abstract/Free Full Text] - Hansson GK. Immune and inflammatory mechanisms in the development of atherosclerosis. Br Heart J. 1993;69:S38–S41.
- Escaned J, van Suylen RJ, MacLeod DC, Umans VA, de Jong M, Bosman FT, de Feyter PJ, Serruys PW. Histologic characteristics of tissue excised during directional coronary atherectomy in stable and unstable angina pectoris. Am J Cardiol. 1993;71:1442–1447.[Medline] [Order article via Infotrieve]
- Virmani R. Cardiac myxoma. In: Burke A, Virmani R, eds. Tumors of the Heart and Great Vessels. Washington, DC: Armed Forces Institute of Pathology; 1996:21–46.
- Bowers WE, Goodell EM. Dendritic cell ontogeny. Res Immunol. 1989;140:880–883.[Medline] [Order article via Infotrieve]
-
Zhang J, Yu ZX, Fujita S, Yamaguchi ML, Ferrans
VJ. Interstitial dendritic cells of the rat heart:
quantitative and ultrastructural changes in experimental myocardial
infarction. Circulation. 1993;87:909–920.
[Abstract/Free Full Text] - Teunissen MB. Dynamic nature and function of epidermal Langerhans cells in vivo and in vitro: a review, with emphasis on human Langerhans cells. Histochem J. 1992;24:697–716.[Medline] [Order article via Infotrieve]
- Reid CD. The dendritic cell lineage in haemopoiesis. Br J Haematol. 1997;96:217–223.[Medline] [Order article via Infotrieve]
- Bobryshev YV, Lord RS, Rainer S, Jamal OS, Munro VF. Vascular dendritic cells and atherosclerosis. Pathol Res Pract. 1996;192:462–467.[Medline] [Order article via Infotrieve]
- Wood GS, Turner RR, Shiurba RA, Eng L, Warnke RA. Human dendritic cells and macrophages: in situ immunophenotypic definition of subsets that exhibit specific morphologic and microenvironmental characteristics. Am J Pathol.. 1985;119:73–82.[Abstract]
- Steinman RM. The dendritic cell system and its role in immunogenicity. Annu Rev Immunol. 1991;9:271–296.[Medline] [Order article via Infotrieve]
- Jaffe R. Review of human dendritic cells: isolation and culture from precursors. Pediatr Pathol. 1993;13:821–837.[Medline] [Order article via Infotrieve]
- Cella M, Sallusto F, Lanzavecchia A. Origin, maturation and antigen presenting function of dendritic cells. Curr Opin Immunol. 1997;9:10–16.[Medline] [Order article via Infotrieve]
- Sprecher E, Becker Y. Role of Langerhans cells and other dendritic cells in disease states. In Vivo.. 1993;7:217–227.[Medline] [Order article via Infotrieve]
-
Scott NA, Cipolla GD, Ross CE, Dunn B, Martin FH,
Simonet L, Wilcox JN. Identification of a potential role for the
adventitia in vascular lesion formation after balloon overstretch
injury of porcine coronary arteries. Circulation. 1996;93:2178–2187.
[Abstract/Free Full Text] -
Shi Y, O'Brien JE, Fard A, Mannion JD, Wang D,
Zalewski A. Adventitial myofibroblasts contribute to
neointimal formation in injured porcine coronary
arteries. Circulation. 1996;94:1655–1664.
[Abstract/Free Full Text] -
Shi Y, O'Brien JE Jr, Mannion JD, Morrison RC, Chung
W, Fard A, Zalewski A. Remodeling of autologous saphenous vein grafts:
the role of perivascular myofibroblasts. Circulation. 1997;95:2684–2693.
[Abstract/Free Full Text] - Nicosia RF, Villaschi S. Rat aortic smooth muscle cells become pericytes during angiogenesis in vitro. Lab Invest. 1995;73:658–666.[Medline] [Order article via Infotrieve]
- Meyrick B, Reid L. Ultrastructural features of the distended pulmonary arteries of the normal rat. Anat Rec. 1979;193:71–97.[Medline] [Order article via Infotrieve]
- Shepro D, Morel NM. Pericyte physiology. FASEB J. 1993;7:1031–1038.[Abstract]
- Hirschi KK, D'Amore PA. Pericytes in the microvasculature. Cardiovasc Res. 1996;32:687–698.[Medline] [Order article via Infotrieve]
- Schmidt RA, Wirtala J. A modification of Movat's pentachrome stain with improved reliability of elastin staining. J Histotech. 1996;19:325–327.
- Luna LG, ed. Manual of Histologic Staining Methods of the Armed Forces Institute of Pathology. New York, NY: McGraw-Hill Book Co; 1968:81–82.
- Tjurmin AV, Hong MK, Haudenschild CC, Kent KM, Pichard AD, Leon MB. Histopathologic observations of directional coronary atherectomy (DCA) specimens after excimer laser angioplasty in complex coronary lesions. J Invasive Cardiol. 1996;8:90. Abstract.
- Langerhans P. Über die nerven der menschlichen havt. Virchows Arch Pathol Anat.. 1868;44:325–337.
- Bagasra O, Seshamma T. In: Hansen J, ed. Protocol: In Situ Amplification and Hybridization. Philadelphia, Pa: Thomas Jefferson University, Omar Bagasra; 1993:1–16.
- Miller MA, Kolb PE, Raskind MA. A method for simultaneous detection of multiple mRNAs using digoxigenin and radioisotopic cRNA probes. J Histochem Cytochem. 1993;41:1741–1750.[Abstract]
-
Skalli O, Ropraz P, Trzeciak A, Benzonana G, Gillessen
D, Gabbiani G. A monoclonal antibody against -smooth muscle
actin: a new probe for smooth muscle differentiation. J Cell
Biol. 1986;103:2787–2796.
[Abstract/Free Full Text] -
Lees-Miller JP, Heeley DH, Smillie LB, Kay CM.
Isolation and characterization of an abundant and novel 22-kDa protein
(SM22) from chicken gizzard smooth muscle. J Biol Chem. 1987;262:2988–2993.
[Abstract/Free Full Text] - Weissberg PL, Cary NR, Shanahan CM. Gene expression and vascular smooth muscle cell phenotype. Blood Press Suppl. 1995;2:68–73.[Medline] [Order article via Infotrieve]
- Osborn M, Weber K. Intermediate filaments: cell-type-specific markers in differentiation and pathology. Cell. 1982;31:303–306.[Medline] [Order article via Infotrieve]
- Rungger-Brandle E, Gabbiani