Articles |
From the Departments of Pathology (H.S.S., D.M.L., P.P., S.M.S., M.E.R.) and Pathobiology (M.E.R.), University of Washington, Seattle, Wash, and the Cardiovascular Research Department, Genentech, Inc, South San Francisco, Calif (L.P-B., S.B.)
Correspondence to Michael E. Rosenfeld, Department of Pathobiology, Box 353410, University of Washington, Seattle, WA 98195. E-mail ssmjm{at}u.washington.edu
| Abstract |
|---|
|
|
|---|
Key Words: stenosis compensatory remodeling medial necrosis adventitial inflammation apolipoprotein E
| Introduction |
|---|
|
|
|---|
Genetically engineered mice are potential models for studying changes in advanced lesions, which may cause stenosis. One such model contains a targeted disruption in the apolipoprotein (apo) E gene (apoE-deficient) and results in a plasma lipoprotein profile that resembles type III hypercholesterolemia in humans.7 8 Previous studies of atherosclerotic lesion development in the apoE-deficient mouse have demonstrated the presence of advanced, atheromatous lesions in the major muscular arteries and aorta.9 10 11 12 13 The aims of this study were to assess the distribution, characteristics, and severity of advanced lesions throughout the entire vasculature of apoE-deficient mice with a particular focus on whether stenotic lesions occur reproducibly at any sites. We report that stenotic lesions consistently develop in the external carotid and popliteal arteries in these animals and that the failure to remodel may be associated with the presence of extracellular lipid deposits, adventitial inflammation, and medial atrophy.
| Methods |
|---|
|
|
|---|
|
The mice were anesthesized with ketamine/xylazine (85.7 and 7.7 µg/g body weight, respectively), blood samples were collected for analysis of serum cholesterol, and the majority of the animals were perfusion-fixed with 10% formalin at 90 to 100 mm Hg for 4 to 5 minutes. After perfusion, most of the vascular tree was dissected intact. Each arterial tree included the carotid bifurcation and small branches, the thoracic and abdominal aorta with the renal arteries, and the iliac, femoral, and popliteal arteries. The internal carotid artery was not dissected out with the rest of the arterial tree as it runs up to the brain through an inaccessible bony canal. Each of the intact vascular trees was placed into 10% formalin for immersion fixation. An additional five apoE-deficient mice on the chow diet and three wild-type mice were killed for frozen sections. After perfusion fixation, the arterial trees were dissected as above, quick-frozen in ethanol with dry ice, and mounted in OCT compound (Miles Inc). Frozen sections were cut on a Reichert 2800 cryostat (Reichert Scientific Instruments) and kept in a -70°C freezer for less than 1 month.
Determination of the Location, Severity, and Predictability of
Atherosclerotic Lesions
Six of the intact arterial trees of the
perfusion-fixed chow-fed and high-fat-fed apoE null mice were
analyzed under a dissecting microscope with appropriate back
lighting. The exact distribution of grossly visible atherosclerotic
lesions was mapped and scored for the apparent degree of severity. This
degree of severity was later verified using morphological,
morphometric, and immunohistochemical criteria in
histological cross sections of the lesions of all 12 of
the chow-fed apoE-deficient mice, which had been embedded in paraffin
and by plastic casting of the vasculature of an additional group of six
9-month-old, chow-fed apoE-deficient mice.
Analysis of Luminal Remodeling and Medial Atrophy
Sections of each lesion were stained with Harris
hematoxylin-eosin or Movat's pentachrome
stain,14 and the plaque and original lumen areas
were measured using digital image analysis (Optimas 5.2,
Optimas Corp). The original lumen area was determined under
phase-contrast microscopy by tracing the entire length of the internal
elastic lamina (IEL). The resulting value was compared with the lumen
area measured in the wild-type mice at the exact same locations in the
external carotid and popliteal arteries and in the aorta. Successful
luminal remodeling of the atherosclerotic artery occurred if the area
of the lumen and the area circumscribed by the IEL were greater than or
equal to the corresponding area values for the wild-type animals. An
artery was considered to be stenotic (loss of remodeling) if
the lumen area of the atherosclerotic artery (original lumen area minus
the area of the plaque) was <60% of the area circumscribed by the IEL
(original lumen area). The thickness of the media was measured in the
peripheral arteries with stenotic lesions at sites
where cholesterol clefts were deposited along the IEL,
where there were no cholesterol clefts, where there was no
lesion (in sections with eccentric lesions), and in the wild-type
animals. The strength of the statistical associations between
morphometric parameters was assessed using regression and
2 analyses according to standard
procedures (NCSS, v5.x).
Immunohistochemical Analyses
Paraffin-embedded and/or frozen sections of atherosclerotic
lesions in the aortic arch, thoracic and abdominal aortas, and external
carotid and popliteal arteries of the chow-fed apoE null mice were used
for immunohistochemical analyses of the cellular components of
the lesions. Except where noted, a biotin-streptavidin-immunoperoxidase
procedure was used according to the manufacturer's specifications
(Vector Laboratories). In those cases where mouse monoclonal antibodies
were used as the primary antibodies, the tissue was preblocked with
goat anti-mouse antiserum (Dako) before addition of the primary
antibodies. The smooth muscle cell and macrophage contents of
the lesions in the paraffin and frozen sections were assessed using an
anti-
-actin mouse monoclonal antibody (Boehringer Mannheim;
1:400 dilution) and anti-F4/80 rat monoclonal
antibody15 (BioSource International; 1:20
dilution), respectively. For the characterization of the elastin,
collagen, and proteoglycan contents of the lesions, paraffin sections
were stained with the Movat's pentachrome
histological staining procedure according to standard
techniques.14
Electron Microscopy
The atherosclerotic lesions in the external carotid artery and
lesser curvature of the aortic arch were examined by transmission
electron microscopy (JEM-1200EX II, JEOL). In brief, segments of the
arteries that had previously been perfusion-fixed with 10% formalin
were immersed in 1/2 strength Karnovsky's fixative (2%
paraformaldehyde and 2.5%
glutaraldehyde in 0.1 mol/L cacodylate buffer,
pH 7.3) for 1 hour at room temperature, washed in 0.1 mol/L
cacodylate buffer containing 4 to 6% sucrose, and postfixed in 1%
OsO4 in 0.1 mol/L cacodylate buffer for 1
hour at room temperature. The tissues were dehydrated with gradient
ethanol and propylene oxide and infiltrated and embedded in Medcast
(Pella). Toluidine blue dye was applied to 1-µm-thick sections for
preliminary examination. Areas of each lesion that were
representative of the lesion as a whole were chosen for
transmission electron microscopy evaluation. These included areas
exhibiting a high degree of cellularity and a predominance of
connective tissue as well as areas along the internal and external
elastic lamina. After trimming of each plastic block, thin sections (80
to 100 nm) were cut and stained with uranyl acetate (6% aqueous) and
lead citrate (Reynolds).
Casting of the Arterial Tree
Batson's plastic kit (Polysciences) was used for casting
of the arterial tree of six of the chow-fed and three of
the control wild-type mice. Casting was done to obtain additional
verification of the distribution of stenotic lesions as
determined previously under the dissecting microscope. Casting was also
used as an additional means to quantitate the reduction in lumen size
because of the presence of stenotic lesions. Unfortunately, due
to inconsistent shrinkage of the casting material, it was
impossible to obtain highly accurate measures of lumen size using this
technique. In brief, each mouse was heparinized with approximately 100
U/kg of heparin intravenously via the tail vein. Each mouse
was catheterized via the left ventricle, and the promotor and catalyst
in monomer solution were mixed well for 3 minutes. The complete
solution was injected into the left ventricle at a pressure of between
100 and 120 mm Hg. Pressure was maintained for 15 to 30 minutes
until the solution was set. The whole mouse was then put into saturated
KOH solution (20%) for 48 hours and fresh KOH solution for another
day. The contracted volume of the plastic was also measured outside of
the mice, and further demonstrated inconsistencies in the degree to
which shrinkage occurred.
| Results |
|---|
|
|
|---|
|
|
In 9 of 12 chow-fed animals, stenotic lesions formed in the
right and left external carotid artery and less frequently in the
common and internal carotid, iliac, femoral, and popliteal arteries.
The most severely stenotic lesions were located in the external
carotid arteries (Table 2
). The distribution of stenotic
lesions in the mice was further verified by evaluating the locations of
large depressions in vascular casts made in 6 of the chow-fed
apoE-deficient mice. As shown in Fig 2
, the casts also demonstrated that the most severely stenotic
lesions were predominantly located in the external carotid arteries and
less frequently in other peripheral arteries. We have not
observed stenotic lesions in any arteries other than the
peripheral arteries listed in Table 2
.
|
Characteristics of the Stenotic Lesions
Extracellular Matrix
There were no apparent differences in the composition and
distribution of the extracellular matrix in the stenotic
lesions (external carotid) compared with those with normal lumen size
(aorta). Staining of cross sections with Movat's pentachrome
stain showed that the extracellular matrix contained a sparse network
of collagen and elastic fibers dispersed throughout most of the plaque
including the core. However, there was a more dense network of
connective tissue fibers associated with the thin smooth muscle layer
situated immediately beneath the endothelium (fibrous
cap) and along the shoulders of the plaques (Fig 3
).
|
Cellular Composition
Immunohistochemical staining with cell type-specific antibodies
showed the presence of
-actin-stained smooth muscle cells arrayed in
a multilayered pattern resembling a thin fibrous cap immediately
beneath the endothelium. There were also scattered
smooth muscle cells throughout the entire lesion (Fig 4
). F4/80-positive macrophages
were abundant throughout the intima, but were mostly concentrated
within the core of the lesions (Fig 5A
).
In the adventitia, an inflammatory cell infiltration was observed in
most of the animals (Fig 5B
). In some cases, this inflammatory cell
insudate appeared to invade and replace the outer layers of the media
(Figs 3 to 5![]()
![]()
).
|
|
Ultrastructure
Electron microscopic evaluation of the lesions in the external
carotid arteries revealed a high degree of cellularity, especially the
abundance of foam cells (Fig 6
). At high
resolution, large amounts of extracellular lipid in the form of
cholesterol clefts and dense osmiophilic deposits were
evident (Fig 7
). The
cholesterol clefts were especially abundant at the base of
the lesions adjacent to the IEL in areas where there was also a
thinning of the media (Fig 8
). We also
observed that the smooth muscle cells situated beneath the
endothelial layer contained relatively small amounts of
lipid within the cytoplasm (Fig 6
). In contrast, there were significant
numbers of lipid-laden smooth muscle cells situated along the IEL and
within the media (Fig 9
).
|
|
|
|
Morphometric Analysis
Measurement of the extent of vascular narrowing was based on the
following equation:
![]() |
![]() |
Morphometric measurements were made of the average lumen area and
the areas circumscribed by the external elastic lamina and IEL of
arteries from both wild-type C57Bl/6 mice without
atherosclerosis and the corresponding vessels from
apoE-deficient mice. Table 3
shows that
in the thoracic aorta at the point of maximum lesion thickness there
was maintenance or even possibly an increase in the lumen size
compared with the wild-type arteries devoid of lesions. There was also
an increase in the length of the IEL. In contrast, there was a
significant reduction in the area of the lumen compared with the
wild-type arteries at the peripheral sites and no increase
in the length of the IEL. In 9 of 12 apoE-deficient mice (75%), there
was a reduction in lumen area in the external carotid artery compared
with 5 of 12 (41.6%) in the femoral and popliteal arteries and 4 of 12
(33.3%) in the common carotid artery (Table 2
). The prevalence of
atherosclertoic lesions was also highest in the external carotid artery
(Table 2
).
|
Regression analyses of a combined sample of
peripheral arteries (external carotid, popliteal, femoral,
and common carotid arteries) demonstrated a positive correlation
between the size of the lesions and the area of the lumen
(r=0.40, P=.004) (Fig 10A
). The correlation between the size
of the lesion and the area circumscribed by the IEL was also positive
but reached a much higher r value than the corresponding
association between lesion size and lumen area (r=0.92,
P<.001, Fig 10B
). Similarly, there was a stronger positive
correlation between the size of the lumen and the area circumscribed by
the IEL (r=0.72, P<.001, Fig 10C
). Thus,
although we did not observe an average increase in the length of the
IEL in the external carotid arteries, the analyses of the
combined sample of lesions in all of the peripheral
arteries suggest that there may be some compensatory enlargement of
lumen size at peripheral sites due to dilation of the
IEL.
|
As our histological and ultrastructural
analyses revealed a significant degree of extracellular lipid
deposition along the IEL, medial atrophy, and the presence of
adventitial inflammation at many of the peripheral sites,
we evaluated whether these pathological processes were statistically
associated with the failure to remodel (Table 4
). The correlation between the degree of
stenosis and medial atrophy, the presence of
cholesterol clefts or adventitial inflammation. and the
presence of cholesterol clefts or adventitial inflammation
and medial atrophy were all significant (P<.01). There was
no correlation between the presence of adventitial inflammation or
cholesterol clefts and the degree of medial atrophy in the
thoracic aorta (data not shown). The relationship between the presence
of extracellular cholesterol clefts and thinning of the
media was further analyzed by measuring the thickness of the
media (1) underlying sites where cholesterol clefts were
deposited along the IEL, (2) underlying areas of the lesion where no
cholesterol clefts were observed, (3) underlying sites
without any overlying lesion (eccentric lesions), and (4) in the
wild-type animals. As shown in Table 5
,
the media underlying those locations where lipid was deposited along
the IEL was considerably thinner than the media at any other
locations.
|
|
| Discussion |
|---|
|
|
|---|
Our morphometric analyses of advanced lesions in the chow-fed,
9-month-old apoE-deficient mouse demonstrate a consistent
pattern of narrowing in specific peripheral vessels. As in
human atherosclerosis, this narrowing is not a simple
function of encroachment by an enlarging plaque. Instead, the data
suggest that some remodeling of the vessel wall occurs at these sites,
since there is a positive correlation between increases in the area
circumscribed by the IEL and lumen or lesion size (Fig 10
). This amount
of remodeling, however, inadequately compensates for the extent of
formation of the lesions. Glagov and
associates18 19 reported that lumen enlargement
occurs in the human coronary artery up to the point at which
plaque area occupies approximately 40% of the area circumscribed by
the IEL, suggesting that plaque size does have an impact on the
remodeling process in the coronary arteries. In contrast, there
was no obvious correlation of lesion size with lumen loss in the
apoE-deficient mouse. These data, along with data from human and monkey
vessels,1 suggest that vessel narrowing is due to
pathological remodeling where the media and adventitia actively change
to narrow the lesion, independent of lesion size.
Pathological remodeling of an atherosclerotic vessel and impairment of the capacity of the artery to dilate, ie, failure of normal remodeling, are likely to depend on the composition of the lesion and the properties of the media and adventitia. For example, we have observed that intimal smooth muscle cells situated close to the lumen in the stenotic lesions of the apoE-deficient mice do not contain significant amounts of intracellular lipid while smooth muscle cells situated along the IEL and in the media appear to be routinely engorged with lipid. Furthermore, it appears that pools of extracellular lipid in the form of large aggregates of cholesterol clefts occur predominantly along the IEL. It is possible that excessive lipid accumulation along the IEL and within the media has an impact on the elastic properties of the media and inhibits chronic dilation and remodeling. This may be analogous to what occurs during the transitional stage of lesion development in humans whereby increasing lipid deposition contributes to the formation of the necrotic core,24 25 but the effects in a mouse vessel would be both more rapid and more marked.
Adventitial inflammation and medial atrophy may also underlie the
failure to remodel or pathological remodeling. In our studies, medial
atrophy was associated with inflammatory cell infiltration of the
adventitia in the stenotic peripheral arteries and
the degree of lumen narrowing was significantly correlated with the
presence of medial atrophy (Table 4
). It is possible that the presence
of inflammatory cells in the adventitia promotes death of medial cells
and a loss of the normal capacity to remodel.
There have been several reports of an association between the thinning
of the media and the presence of adventitial inflammation in human
atherosclerotic lesions and in allograft
arteriosclerosis in
rats.26 27 28 29 30 31 32 The mediators of such thinning have
not been determined. However, recent studies have shown that smooth
muscle cells have apoptotic receptors including fas and tumor
necrosis factor receptor-1 (D.K.M. Han et al, unpublished
observations). Thus fas ligand or tumor necrosis factor
released from leukocytes at sites of inflammation might thin the
media by inducing apoptosis. Although we did not
immunohistochemically analyze the mouse arteries for the
presence of lymphocytes, others have demonstrated the presence of
lymphocytes in lesions from the apoE-deficient
mouse.29 Our studies do show extensive staining
for macrophages in the adventitia, and electron microscopic
analyses showed that lymphocytes are present although the
type and relative numbers could not be determined.
Case reports on vasculitis in humans have shown that in some cases the lumen narrowing in small arteries and arterioles, thought to be due to vasculitis, is actually due to cholesterol emboli. It is believed that the presence of the cholesterol emboli stimulates the infiltration of inflammatory cells.33 34 Thus, in the apoE-deficient mice, the presence of large amounts of extracellular lipid and cholesterol clefts situated along the IEL and within the media may act similarly to cholesterol emboli and, thus, may be a partial explanation for the simultaneous presence of adventitial inflammation. Oxidized lipids are known to be chemotactic for both monocytes and lymphocytes in vitro and can activate these cells to express additional chemotactic factors.35 Furthermore, oxidation-specific epitopes have been demonstrated to be associated with the adventitia and with cholesterol clefts in atherosclerotic lesions in the apoE-deficient mouse36 and, thus, may also help explain our present observations in these mice.
Why is it that remodeling is successful at 9 months in the aorta and not in the external carotid artery? In the apoE-deficient mouse, the most stenotic lesions form at sites in the external carotid and popliteal arteries that have approximately the same lumen diameter and medial thickness. This suggests that the size of the artery may be more important than the location in dictating whether remodeling will be successful. For example, Zarins et al37 have shown that greater pressure is required to dilate a small artery than a large artery. To increase the diameter of a 1-mm vessel by 25% requires approximately 6 times the pressure that would be required to increase a 3-mm vessel by the same percentage. Furthermore, the toxic effects of medial lipid deposition or adventitial inflammation could be more critical in thinner arteries where loss of even a small amount of mass would have a more immediate effect on the capacity to dilate than in arteries such as the aorta where additional layers of smooth muscle cells and connective tissue could continue to dilate and remodel despite the loss of mass.
In summary, the apoE-deficient mouse shows a consistent pattern of vascular stenosis at reproducible arterial sites. This narrowing, as with the narrowing seen in human atherosclerosis, occurs either via a failure of normal remodeling mechanisms or perhaps active pathological remodeling associated with the presence of cytotoxic lipids or cytokines released from adventitial inflammation. It is also intriguing to note that medial atrophy, usually thought of as leading to aneurysm and dissection, is associated with stenosis. Perhaps loss of medial function, secondary to cell death induced by inflammation, leads to the loss of ability of atherosclerotic vessels to accommodate growing lesions.
| Acknowledgments |
|---|
Received May 27, 1997; accepted August 18, 1997.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. C.H. Clarke, T. D. Littlewood, N. Figg, J. J. Maguire, A. P. Davenport, M. Goddard, and M. R. Bennett Chronic Apoptosis of Vascular Smooth Muscle Cells Accelerates Atherosclerosis and Promotes Calcification and Medial Degeneration Circ. Res., June 20, 2008; 102(12): 1529 - 1538. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Aikawa, M. Nahrendorf, J.-L. Figueiredo, F. K. Swirski, T. Shtatland, R. H. Kohler, F. A. Jaffer, M. Aikawa, and R. Weissleder Osteogenesis Associates With Inflammation in Early-Stage Atherosclerosis Evaluated by Molecular Imaging In Vivo Circulation, December 11, 2007; 116(24): 2841 - 2850. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Zhang, K. Peppel, P. Sivashanmugam, E. S. Orman, L. Brian, S. T. Exum, and N. J. Freedman Expression of Tumor Necrosis Factor Receptor-1 in Arterial Wall Cells Promotes Atherosclerosis Arterioscler. Thromb. Vasc. Biol., May 1, 2007; 27(5): 1087 - 1094. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Falk, S. M. Schwartz, Z. S. Galis, and M. E. Rosenfeld Putative Murine Models of Plaque Rupture Arterioscler. Thromb. Vasc. Biol., April 1, 2007; 27(4): 969 - 972. [Full Text] [PDF] |
||||
![]() |
S. M. Schwartz, Z. S. Galis, M. E. Rosenfeld, and E. Falk Plaque Rupture in Humans and Mice Arterioscler. Thromb. Vasc. Biol., April 1, 2007; 27(4): 705 - 713. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-H. Wu, R. Goswami, X. Cai, S. T. Exum, X. Huang, L. Zhang, L. Brian, R. T. Premont, K. Peppel, and N. J. Freedman Regulation of the Platelet-derived Growth Factor Receptor-beta by G Protein-coupled Receptor Kinase-5 in Vascular Smooth Muscle Cells Involves the Phosphatase Shp2 J. Biol. Chem., December 8, 2006; 281(49): 37758 - 37772. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. P.W. Moos, N. John, R. Grabner, S. Nossmann, B. Gunther, R. Vollandt, C. D. Funk, B. Kaiser, and A. J.R. Habenicht The Lamina Adventitia Is the Major Site of Immune Cell Accumulation in Standard Chow-Fed Apolipoprotein E-Deficient Mice Arterioscler. Thromb. Vasc. Biol., November 1, 2005; 25(11): 2386 - 2391. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Rattazzi, B. J. Bennett, F. Bea, E. A. Kirk, J. L. Ricks, M. Speer, S. M. Schwartz, C. M. Giachelli, and M. E. Rosenfeld Calcification of Advanced Atherosclerotic Lesions in the Innominate Arteries of ApoE-Deficient Mice: Potential Role of Chondrocyte-Like Cells Arterioscler. Thromb. Vasc. Biol., July 1, 2005; 25(7): 1420 - 1425. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Mercer, N. Figg, V. Stoneman, D. Braganza, and M. R. Bennett Endogenous p53 Protects Vascular Smooth Muscle Cells From Apoptosis and Reduces Atherosclerosis in ApoE Knockout Mice Circ. Res., April 1, 2005; 96(6): 667 - 674. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. J. Merched and L. Chan Absence of p21Waf1/Cip1/Sdi1 Modulates Macrophage Differentiation and Inflammatory Response and Protects Against Atherosclerosis Circulation, December 21, 2004; 110(25): 3830 - 3841. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. F. Bentzon, G. Pasterkamp, and E. Falk Expansive Remodeling Is a Response of the Plaque-Related Vessel Wall in Aortic Roots of ApoE-Deficient Mice: An Experiment of Nature Arterioscler. Thromb. Vasc. Biol., February 1, 2003; 23(2): 257 - 262. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Palinski and C. Napoli Unraveling Pleiotropic Effects of Statins on Plaque Rupture Arterioscler. Thromb. Vasc. Biol., November 1, 2002; 22(11): 1745 - 1750. [Full Text] [PDF] |
||||
![]() |
M. R. Bennett Breaking the Plaque: Evidence for Plaque Rupture in Animal Models of Atherosclerosis Arterioscler. Thromb. Vasc. Biol., May 1, 2002; 22(5): 713 - 714. [Full Text] [PDF] |
||||
![]() |
V. V. Kunjathoor, D. S. Chiu, K. D. O'Brien, and R. C. LeBoeuf Accumulation of Biglycan and Perlecan, but Not Versican, in Lesions of Murine Models of Atherosclerosis Arterioscler. Thromb. Vasc. Biol., March 1, 2002; 22(3): 462 - 468. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Braun, B. L. Trigatti, M. J. Post, K. Sato, M. Simons, J. M. Edelberg, R. D. Rosenberg, M. Schrenzel, and M. Krieger Loss of SR-BI Expression Leads to the Early Onset of Occlusive Atherosclerotic Coronary Artery Disease, Spontaneous Myocardial Infarctions, Severe Cardiac Dysfunction, and Premature Death in Apolipoprotein E-Deficient Mice Circ. Res., February 22, 2002; 90(3): 270 - 276. [Abstract] [Full Text] [PDF] |
||||
![]() |
|