Vascular Biology |
From the John P. Robarts Research Institute (Vascular Biology Group), London Health Science Centre, and the University of Western Ontario, Departments of Medicine (Cardiology), Biochemistry, and Medical Biophysics London, Ontario, Canada.
Correspondence to J. Geoffrey Pickering, MD, PhD, FRCP(C), London Health Science Centre, 339 Windermere Rd, London, Ontario N6A 5A5. E-mail gpickering{at}rri.on.ca
| Abstract |
|---|
|
|
|---|
Key Words: smooth muscle cells collagen heat-shock protein neointima vascular development
| Introduction |
|---|
|
|
|---|
The major collagen species in both normal and diseased
human arteries is type I collagen, a heterotrimeric, fibril-forming
collagen that comprises 2
1(I) collagen chains and 1
2(I)
collagen chain. The component
-chains are derived from precursors,
namely pro
1(I) collagen and pro
2(I) collagen chains, that
associate with each other in the lumen of the endoplasmic reticulum
(ER). This assembly process requires precise sorting and folding of the
pro
collagen chains within the ER and must occur before procollagen
can be transported out of the cell. Quality control for procollagen
assembly is dependent on a number of ER-resident enzymes and molecular
chaperones. For example, prolyl 4-hydroxylase hydroxylates proline
residues, the presence of which are necessary for winding of the long
triple helical domain. The molecular chaperones, protein disulfide
isomerase and immunoglobulin heavy-chain binding protein, transiently
bind to a target region of the propeptide and facilitate either
physiological folding or degradation of misfolded
protein.4 5
HSP47 is a 47-kDa heat-shockinducible glycoprotein that has also been found to associate with procollagen in the ER.6 7 Although its exact role in procollagen processing is unclear, data suggest that it acts as a collagen-specific chaperone. HSP47 has been localized exclusively to the ER of collagen-producing cells8 and has been found to associate with nascent type I procollagen chains as they translocate into the ER.4 9 It has also been shown to bind triple helical procollagen in the ER, where it may stabilize this conformation.10 Inhibition of HSP47 expression with antisense oligonucleotides has been associated with decreased expression of type I collagen in mouse fibroblasts9 and in a rat model of glomerulosclerosis.11
Recently, we determined that HSP47 was expressed by human arterial SMCs in culture, and we identified HSP47 in the fibrous cap of human atherosclerotic plaque.12 These findings highlight an involvement of this stress protein in vascular disease and, in particular, with fibrotic changes in the artery wall. The in vivo association between HSP47 expression and vascular fibrosis is consistent with other studies that have found abundant HSP47 in fibrotic organs but little, if any, in normal tissues.13 14 15 Thus, although the in vivo role for HSP47 is not well defined, the expression data to date suggest fibrotic repair as a dominant context. HSP47 has, however, also been detected during embryological development of the heart, lung, and kidneys.16 17 18 A developmental profile for arterial expression has not been defined; however, the data from other organs raise the possibility that HSP47 may have a broader role in arterial restructuring, beyond that of vascular scarring.
To investigate the possibility that HSP47 is involved in diverse forms of arterial remodeling, we analyzed HSP47 expression in the developing rat aorta and in the adult rat carotid artery during injury-induced neointima formation. Because SMC phenotype is a fundamental determinant of arterial restructuring, we also evaluated the relationship between expression of HSP47 and SMC differentiation status. Finally, we explored a mechanistic relationship between the termination of HSP47 expression by SMCs contributing to vascular restructuring and the appearance of nascent collagen fibrils in the extracellular space. The findings indicate that upregulation of HSP47 is a novel feature of acute vascular remodeling and that the duration of HSP47 expression under these circumstances is regulated by constituents of the extracellular matrix (ECM).
| Methods |
|---|
|
|
|---|
1(I) chain of human type I collagen, LF67 (gift
of Dr L.W. Fisher, National Institute of Dental and Craniofacial
Research, Bethesda,
Md)20 21 ; and a
mouse monoclonal antibody to human calponin (Sigma Chemical Co).
Biotinylated horse anti-mouse IgG or biotinylated goat anti-rabbit IgG
was used as a secondary antibody for immunostaining.
Vector SG peroxidase substrate was obtained from Vector Laboratories
Inc, and 3,3'-diaminobenzidine (DAB) was obtained from Sigma. Human
pro
1(I) collagen cDNA was derived from Hf677 (American Type Culture
Collection, Manassas, Va) and cloned into pGEM3 (pSP3, gift from Dr C.
Farrell, Amgen, Thousand Oaks, Calif). HSP47 mRNA expression was
detected by using a plasmid containing a partial cDNA clone for rat
HSP47 (pIP1).22 Rat tail
collagen was isolated from tendons dissected from rat tails. Tendons
were dissolved in 0.1 mol/L acetic acid at 4°C for 24 hours.
Insoluble matter was removed, and NaCl was added to the solution to a
final concentration of 1.7 mol/L to precipitate the collagen, which was
retrieved by centrifugation at
3500g at 4°C for 20 minutes.
The precipitate was resuspended in a minimal volume of 0.1 mol/L acetic
acid and then reprecipitated with 1.7 mol/L NaCl. This precipitate was
resuspended in 0.1 mol/L acetic acid and dialyzed against 17 µmol/L
acetic acid for 3 to 5 days. The amount of collagen was quantified by
Lowry assay (Bio-Rad) and the final concentration adjusted to 4 mg/mL
with 17 µmol/L acetic acid.
Fetal, Neonatal, and Adult Animal
Tissues
Timed-pregnant Sprague-Dawley rats (Charles
River Canada, St. Constant, Quebec) were killed by anesthesia
and subsequent asphyxiation with carbon dioxide at 17 and 19 days of
gestation and fetal animals were removed. Animals after birth were
killed at 1 day, 4 days, and 5 months. In each case, the descending
thoracic aorta was removed, immersion-fixed for 3 hours in 10%
formalin, and embedded in paraffin.
Balloon Injury to the Rat Carotid
Artery
Adult Sprague-Dawley rats were pretreated with
atropine (0.04 mg/kg) and bupremorphine (0.05 mg/kg) subcutaneously and
anesthetized with pentobarbital (40 mg/kg) intraperitoneally). The left
carotid artery was injured using a 2F Fogarty catheter, as
described.23 24
On designated days the animals were killed (pentobarbital, 110 mg/kg
intraperitoneally), and the carotid arteries were perfused in situ with
PBS and then perfusion-fixed with methanolCarnoys fixative
(methanol/chloroform/glacial acetic acid, 6:3:1 vol/vol/vol). The
excised arteries were placed in the same fixative for an additional 3
hours and then embedded in paraffin.
Immunohistochemistry
Deparaffinized tissue sections (4 µm) were
subjected to antigen retrieval by immersion in 10 mmol/L sodium
citrate buffer, pH 6.0; microwaving on full power for 2.5 minutes and
then on low power for 7 minutes; and maintained afterward in hot buffer
for 20 minutes. Nonspecific binding of primary antibodies was blocked
with 10% horse serum or 10% goat serum in PBS. Primary antibodies in
blocking solution were incubated with sections overnight at 4°C.
Endogenous peroxidase activity was inhibited with 3%
H2O2 in methanol.
Sections were incubated with biotinylated secondary antibody for 1 hour
at room temperature. Bound antibody was detected with the ABC Elite kit
from Vector Laboratories and visualized with DAB. The sections were
counterstained with Harris hematoxylin. Formalin-fixed sections of
human dermal granulation tissue served as a positive control for
HSP47.
Collagen Fibril Assessment by Circular
Polarization Microscopy
To visualize polymerized collagen fibrils in the
extracellular space, a polarized light microscopy approach was taken.
Deparaffinized 4-µm-thick sections were immersed in picrosirius red
solution (1% aqueous F3BA Sirius red and water-saturated picric acid,
1:9 vol/vol) for 60 minutes. The sections were then dehydrated
and coverslipped. The sections were viewed under circularly polarized
light by using a Nikon Optiphot microscope, customized for circular
polarization optics with quarter-wave plates inserted in the light path
both below the polarizer and above the
analyzer.
Western Blot Analysis
Cultured SMCs were harvested in lysis solution
(0.1 mmol/L EDTA, 1% wt/vol SDS, 1% wt/vol sodium deoxycholate,
and 0.1% Triton X-100 in PBS, pH 7.4) with PMSF (0.1 mmol/L) and
leupeptin (10 µg/mL). Carotid arteries were homogenized
in the same solution after the adventitia had been stripped away under
microscopic guidance. Equal amounts of protein were resolved on 6%
(for collagen) or 12% (for HSP47 and calponin) polyacrylamide
gels and evaluated by Western blot analysis, as described
previously.25
Northern Blot Analysis
Porcine SMCs in the third through fifth subcultures
were plated onto culture dishes precoated with monomeric collagen or in
3-dimensional collagen fibril lattices. Monomeric collagencoated
dishes were prepared by washing dishes with 0.1% acetic acid for 1
hour at room temperature and then coating them for 3 hours with rat
tail collagen diluted in 0.1% acetic acid to 100 µg/mL. For the
collagen lattice, SMCs were suspended at
2.5x105 cells/mL in medium 199 supplemented
with 4% fetal bovine serum. The collagen lattice was made by mixing 4
mL of the cell suspension with 80 µL of 1N NaOH, 0.4 mL of 10x
minimum essential medium (Gibco), 2 mL of rat tail collagen (4 mg/mL),
and 1.52 mL of 0.1% acetic acid. This mixture was plated in a 100-mm
tissue-culture dish. Total RNA was isolated (TRIzol reagent, Gibco) and
Northern blots were prepared as previously
described.25 Membranes were
exposed to a phosphor imager screen (Molecular Dynamics), and band
density was quantified and expressed relative to that of 18S
rRNA.
| Results |
|---|
|
|
|---|
|
To relate the expression profile of HSP47 to the
accumulation and organization of collagen fibers, we exploited the fact
that polymerized, fibrillar collagen is birefringent and thus visible
by polarized light microscopy. When stained with picrosirius red and
illuminated by circular polarization optics, nascent collagen
fibrils/fibers can be identified with high sensitivity and
specificity.27 28
As shown in
Figures 1F
and 1G
, by embryonic days 17 and 19, collagen
fibers were evident in the adventitial layer only, despite the
abundance of HSP47 throughout the artery wall at these times. Collagen
fiber formation in the adventitia notably progressed by 1 and 4 days
after birth, with a well-developed fibrillar network. At these neonatal
times, short collagen fibrils were evident in the media; however, these
were scant and weakly birefringent. By 5 months, however, collagen
fibrils were well developed in the media. A band of birefringent
collagen bundles was evident subjacent to the internal elastic lamina,
and a pair of collagen fiber bands straddled each medial elastic
lamella.
HSP47 Expression Is Rapidly Increased During
Injury-Induced Neointimal Formation
To determine whether HSP47 expression was also
upregulated during acute arterial remodeling/repair of the
adult artery, the left carotid artery of rats was subjected to balloon
injury. As illustrated in
Figure 2A
, in uninjured carotid arteries fixed with
methanolCarnoys reagent, there was a weak signal for HSP47 in SMCs,
consistent with the pattern observed in the aorta of adult
rats. HSP47 was also detected in adventitial cells. HSP47 expression
was less obvious in the endothelial cells, although
this was fixation dependent because endothelial cell
staining was evident in carotid arteries fixed with 4%
paraformaldehyde (data not shown). Four days after
injury, HSP47 was strongly expressed in SMCs of the primordial
neointima, as well as in SMCs within the inner layers of
the media
(Figure 2B
). At 14 days, there was intense cellular staining
throughout the expanded neointima, as well as in cells of
the inner media and in adventitial cells
(Figure 2C
). By 28 days, HSP47 expression levels had declined
notably, with weaker expression in scattered intimal cells
(Figure 2D
).
|
Collagen fibril formation was assessed in adjacent
sections by using polarization microscopy. As shown in
Figures 2E
through 2G, birefringent fibers were detectable in
the neointima by day 14, although they were thin, loosely organized,
and evident only in the deeper regions. By day 28
(Figure 2H
), at which time HSP47 expression was relatively
low, there was a relative abundance of thicker and more densely packed
fibrils throughout the neointima
(Figure 2H
).
HSP47 Marks Less-Differentiated SMCs
To assess the relationship between HSP47 and calponin
expression in vivo, control and balloon-injured rat carotid arteries
were assessed by Western blot analysis after the adventitia had
been removed. As shown in
Figure 3
, calponin expression underwent a striking decline 6
days after injury, consistent with the emergence of a
less-differentiated SMC phenotype. This was followed by a
gradual increase in calponin abundance at 14 and 28 days. In contrast,
HSP47 expression, which was detectable at low levels in the uninjured
artery, increased substantially 6 days after injury and declined
thereafter. Type I procollagen expression was not detectable in the
uninjured artery but displayed a transient rise after injury, similar
to HSP47. A similar relationship between HSP47 expression and SMC
differentiation status was established by tracking SMCs in vitro after
their acute dispersion from the porcine carotid artery. An
8-fold
increase in HSP47 coincided with the initiation of DNA synthesis and
decline in calponin expression (Figure
I; published online at
http://atvb.journals.aha).
|
HSP47 Expression by SMCs Is Inhibited by a
Polymerized, 3-Dimensional Collagen Lattice
The in vivo expression profile of HSP47, during both
vascular development and vascular repair, was inversely correlated with
the emergence of collagen fibrils in the ECM. To determine whether
collagen in either its monomeric or fibrillar form could in fact
regulate HSP47 expression by SMCs, they were cultured either on
monomeric type I collagen or in a 3-dimensional collagen fibril
lattice. As shown in
Figure 4
, expression of HSP47 mRNA by SMCs within a
fibrillar collagen lattice fell to
0.4 of basal levels within 24
hours. In contrast, expression of HSP47 in SMCs cultured on monomeric
collagen declined minimally. Consistent with previous reports,
polymerized collagen also inhibited expression of pro
1(I) collagen
mRNA levels in SMCs. After 24 hours of stimulation with fibrillar
collagen, pro
1(I) collagen mRNA levels fell to
0.2 of basal
levels. A more modest decline was observed for SMCs on monomeric
collagen (
0.5 of basal level).
|
| Discussion |
|---|
|
|
|---|
The late fetal and early postpartum period is
characterized by a surge in ECM deposition in the vasculature. Collagen
content has been recognized to rapidly increase after birth in the rat
and sheep
aorta.29 30 At
the gene expression level, pro
1(I) collagen mRNA was found to
substantially increase in the rat coronary artery immediately
after birth.26 Our finding
of HSP47 expression in the late stages of embryonic development and in
the first few neonatal days is thus consistent with the concept
that HSP47 is a participant in procollagen processing in vivo.
Furthermore, it suggests that the intracellular machinery for
optimizing procollagen production is in place at the time of
the perinatal surge in collagen production.
HSP47 expression also increased shortly after vascular
injury. As with aortic development, the early stages of artery repair
and neointimal formation after balloon injury have been
associated with collagen production. Abundance of pro
2(I)
collagen mRNA was found to be markedly increased 7 days after balloon
injury to the rat carotid
artery,31 and pro
1(I)
collagen mRNA and total collagen synthesis rates were found to increase
1 week after balloon injury to rabbit iliac
arteries.32 33
Our observation that type I procollagen peptides are detectable in the
vessel wall 6 days after balloon injury further indicates that the
early stages of neointima formation entail increased
fibrillar collagen synthesis. This in vivo relationship between HSP47
expression and procollagen production in diverse contexts
provides strong support to the hypothesis that HSP47 plays an important
role in collagen metabolism by SMCs.
The pronounced increase in HSP47 in the primordial neointima and inner medial SMCs early (4 days) after balloon injury suggests that HSP47 expression is correlated with the presence of less-differentiated SMCs. This was verified by documenting a reciprocating relationship between HSP47 expression and that of calponin, a marker of relatively mature SMCs, after balloon injury. HSP47 expression also increased in newly cultured SMCs acutely dispersed from the porcine carotid artery, coincident with the initiation of SMC DNA synthesis. Thus, regardless of its role in procollagen processing, HSP47 expression appears to be a marker of less-differentiated SMCs, including those SMCs that manifest increased ECM synthesis.
The finding that HSP47 was constitutively expressed at a low level in the adult rat aorta and carotid artery was not predicted. Prior studies of HSP47 expression in rats 69 days after birth found little, if any, HSP47 mRNA or protein in homogenates of rat heart, kidney, or lung.16 17 Similarly, in an assessment of human coronary arteries, we did not detect HSP47 expression by immunohistochemistry in normal arteries.12 Because the only available antibody to HSP47 is raised against rat HSP47, low-level expression in the human artery wall may not be optimally immunodetectable. The basal expression of HSP47 observed in the current study of rat arteries was further noteworthy because it was even evident in endothelial cells, which are generally considered to express little, if any, collagen under normal conditions. The presence of constitutively expressed HSP47 could suggest that there is in fact a very low level of collagen turnover in the normal aorta and carotid artery. Alternatively, it may suggest that HSP47 and collagen production are not inextricably linked. In this regard, it is noteworthy that HSP47 expression has been found to be increased by certain stresses, including heat and oxidized LDL, without a concomitant increase in type I collagen expression.6 12
The increases in HSP47 expression observed during vascular development and repair were transient. During aortic development, HSP47 expression fell during the first few neonatal days. After vascular injury it declined between 14 and 28 days. In both cases, the relatively precipitous decline in HSP47 expression was correlated with the early accumulation of polymerized collagen fibrils in the artery wall. Although traditionally difficult to visualize, these early collagen fibrils were identified on the basis of their acquisition of a birefringent structure. The findings thus established that HSP47 expression is inhibited as soon as or shortly after de novo collagen fibrils are formed in the local ECM. Furthermore, the in vitro studies suggested that these newly polymerized fibrils were in fact responsible for the downregulation of HSP47 expression.
Polymerized collagen fibrils have previously been
shown to regulate gene expression in several cell
types.2 34
Interestingly, many of the genes regulated by fibrillar collagen encode
proteins that are intimately involved in collagen fibril turnover. In
fibroblasts, fibrillar collagen has been shown to induce the expression
of
collagenase-135
and
collagenase-3,36
increase expression of the collagen-binding
2ß1
integrin,37 and inhibit the
expression of type I
collagen.38 The current
finding that fibrillar, but not monomeric, collagen inhibits expression
of HSP47 by SMCs thus not only identifies a new gene target for
fibrillar collagen but also strengthens the paradigm that the abundance
and architecture of collagen fibrils provide feedback to the SMCs
responsible for remodeling the collagen matrix. Because abundant
collagen deposition is typically a late stage of tissue remodeling, the
downregulation of HSP47 expression by collagen fibrils may be integral
to a mechanism that effectively terminates the remodeling processes
itself.
In summary, we have determined that the stress protein HSP47 is abundantly expressed in the artery wall under diverse circumstances of restructuring, including acute neointimal formation. In these contexts, HSP47 expression was closely linked to the emergence of less-mature SMCs. Furthermore, the findings indicate that HSP47 expression by SMCs can be inhibited by newly polymerized collagen fibrils in the extracellular space, implicating a novel feedback mechanism that may be critical to self-termination of vascular development and lesion growth.
| Acknowledgments |
|---|
Received July 17, 2000; accepted October 26, 2000.
| References |
|---|
|
|
|---|
1ß1 and
2ß1
integrin receptors by human vascular smooth muscle cells:
2ß1
integrin is required for chemotaxis across type I collagen coated
membranes. Am J Pathol. 1994;145:10701081.[Abstract]
1(I) chains of type
I procollagen. J Biol
Chem. 1994;269:39413946.
1(I) and
1(III) collagen genes in carbon tetrachloride-induced rat liver
fibrosis. J Clin Invest. 1994;94:24812488.
-smooth muscle actin in renal allograft failure.
Nephrol Dial Transplant. 2000;15:529535.
5ß1
integrin expression and luminal edge fibronectin matrix assembly by
smooth muscle cells after arterial injury.
Am J Pathol. 2000;156:453465.
2ß1 is upregulated in fibroblasts and highly
aggressive melanoma cells in three-dimensional collagen lattices and
mediates the reorganization of collagen I fibrils.
J Cell Biol. 1991;115:14271436.This article has been cited by other articles:
![]() |
T. W. Small, Z. Bolender, C. Bueno, C. O'Neil, Z. Nong, W. Rushlow, N. Rajakumar, C. Kandel, J. Strong, J. Madrenas, et al. Wilms' Tumor 1-Associating Protein Regulates the Proliferation of Vascular Smooth Muscle Cells Circ. Res., December 8, 2006; 99(12): 1338 - 1346. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. de Nooijer, J.H. von der Thusen, C.J.N. Verkleij, J. Kuiper, J.W. Jukema, E.E. van der Wall, Th.J.C. van Berkel, and E.A.L. Biessen Overexpression of IL-18 Decreases Intimal Collagen Content and Promotes a Vulnerable Plaque Phenotype in Apolipoprotein-E-Deficient Mice Arterioscler. Thromb. Vasc. Biol., December 1, 2004; 24(12): 2313 - 2319. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. F. Rocnik, E. van der Veer, H. Cao, R. A. Hegele, and J. G. Pickering Functional Linkage between the Endoplasmic Reticulum Protein Hsp47 and Procollagen Expression in Human Vascular Smooth Muscle Cells J. Biol. Chem., October 4, 2002; 277(41): 38571 - 38578. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. G. Pickering Regulation of Vascular Cell Behavior by Collagen : Form Is Function Circ. Res., March 16, 2001; 88(5): 458 - 459. [Full Text] [PDF] |
||||
![]() |
S. Li, Y.-S. Fan, L. H. Chow, C. Van Den Diepstraten, E. van der Veer, S. M. Sims, and J. G. Pickering Innate Diversity of Adult Human Arterial Smooth Muscle Cells: Cloning of Distinct Subtypes From the Internal Thoracic Artery Circ. Res., September 14, 2001; 89(6): 517 - 525. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2001 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |