Basic Science Reports
Angiogenesis Inhibitors Endostatin or TNP-470 Reduce
Intimal Neovascularization and Plaque Growth in
Apolipoprotein E–Deficient Mice
Karen S. Moulton, MD; Eric Heller, MS; Moritz A. Konerding, MD; Evelyn Flynn, MS;
Wulf Palinski, MD; Judah Folkman, MD
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Background—Neovascularization within the intima of human atherosclerotic lesions is well described, but its role in the
progression of atherosclerosis is unknown. In this report, we first demonstrate that intimal vessels occur in advanced
lesions of apolipoprotein E– deficient (apoE 2/2) mice. To test the hypothesis that intimal vessels promote
atherosclerosis, we investigated the effect of angiogenesis inhibitors on plaque growth in apoE 2/2 mice.
Methods and Results—ApoE 2/2 mice were fed a 0.15% cholesterol diet. At age 20 weeks, mice were divided into 3
groups and treated for 16 weeks as follows: group 1, recombinant mouse endostatin, 20 mg z kg21 z d21; group 2,
fumagillin analogue TNP-470, 30 mg/kg every other day; and group 3, control animals that received a similar volume
of buffer. Average cholesterol levels were similar in all groups. Plaque areas were quantified at the aortic origin. Median
plaque area before treatment was 0.250 mm2 (range, 0.170 to 0.348; n510). Median plaque areas were 0.321 (0.238 to
0.412; n510), 0.402 (0.248 to 0.533; n515), and 0.751 mm2 (0.503 to 0.838; n512) for the endostatin, TNP-470, and
control groups, respectively (P#0.0001). Therefore, endostatin and TNP-470 inhibited plaque growth during the
treatment period by 85% and 70%. Intimal smooth muscle cell contents of plaques from control and treated mice were
similar.
Conclusions—Prolonged treatment with either angiogenesis inhibitor reduced plaque growth and intimal neovascularization in apoE 2/2 mice. Although the mechanism of plaque inhibition induced by these agents is not established, these
results suggest that intimal neovascularization may promote plaque development. (Circulation. 1999;99:1726-1732.)
Key Words: angiogenesis n atherosclerosis n apolipoproteins
I
n normal vessels, the microvascular network of vasa
vasorum is confined to the adventitia and outer media. In
vessels with atherosclerotic involvement, these networks
become more abundant and extend into the intima of atherosclerotic lesions.1,2 Casting studies and confocal microscopy
have shown that intimal vessels mostly branch from the
native adventitial vasa vasorum.3 The proliferation rates of
endothelial cells in plaque vessels range from undetectable to
43%, which indicates these vessels are found in various
stages of development.4
vessels both suggest that these vessels may recruit inflammatory cells into lesions and initiate a positive-feedback mechanism.8 It is also conceivable that the supply of oxygen and
nutrients provided via plaque vessels is a precondition for
growth beyond a certain stage, after which diffusion from the
artery lumen is insufficient to meet the metabolic demands of
the plaque.
The clinical importance of plaque neovascularization is
suggested by studies that show a higher prevalence of
neovascularization in lesions with plaque rupture, mural
hemorrhage, or unstable angina.9,10 Angiogenesis occurs in
association with remodeling and protease activation in the
surrounding tissues.11,12 Therefore, factors that stimulate
plaque angiogenesis could also contribute to activities that
promote plaque disruption, the event often responsible for
myocardial infarction and ischemic stroke.
To test the hypothesis that intimal neovascularization
promotes the progression of atherosclerosis, we investigated
See p 1653
Plaque vessels are often found in areas rich in macrophages, T cells, and mast cells— cell types that can activate
angiogenesis.5–7 Their close proximity to inflammatory infiltrates and the expression of adhesion molecules (such as
vascular cell adhesion molecule-1, intercellular adhesion
molecule-1, and E-selectin) on the endothelium of plaque
Received September 25, 1998; revision received December 14, 1998; accepted December 29, 1998.
From the Surgical Research Laboratory (K.S.M., E.F., J.F.), Children’s Hospital, Boston Mass; Cardiovascular Division (K.S.M.), Brigham and
Women’s Hospital, Boston, Mass; Harvard Medical School (E.H.), Boston, Mass; the Department of Anatomy (M.A.K.), University of Mainz, Mainz,
Germany; and the Department of Medicine (W.P.), University of California, San Diego, Calif.
Correspondence to Karen S. Moulton, Children’s Hospital, Surgical Research Laboratory Enders 10, 300 Longwood Ave, Boston, MA 02115. E-mail
[email protected]
© 1999 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org
1726
Moulton et al
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
whether treatments with potent angiogenesis inhibitors reduce plaque growth in an animal model of the disease. The
limited availability of such agents has previously impeded the
experimental verification of this hypothesis in significant
numbers of animals.
The fumagillin analogue TNP-470 is a selective inhibitor
of endothelial cell proliferation and migration and is under
investigation in clinical trials as an anticancer agent. TNP470 inhibits primary tumor growth and demonstrates 70%
inhibition of angiogenesis in the corneal micropocket assay.13
Endostatin, a C-terminal 20-kDa fragment of the basement
membrane protein collagen XVIII, is a potent inhibitor of
primary tumor growth and endothelial cell proliferation and
migration.14,15
In this report, we first demonstrate that intimal capillaries
occur in advanced lesions of apolipoprotein E– deficient
(apoE 2/2) mice by immunohistochemistry with antibodies
against CD31 and von Willebrand factor (vWF) and by
transmission electron microscopy (TEM). We then show that
long-term treatment with recombinant murine endostatin or
TNP-470 significantly reduces the further growth of atherosclerosis without affecting cholesterol levels.
Methods
Animals and Experimental Design
Male apoE 2/2 mice (Jackson Labs, Bangor, Me) were fed a 0.15%
cholesterol diet (Western type No. 88137, Teklad) from age 6 to 8
weeks. At 20 weeks of age, 10 animals were killed to evaluate the
baseline extent of atherosclerosis. The remaining littermates were
divided into 3 groups and treated for 16 weeks as follows: group 1
was treated with recombinant murine endostatin 20 mg z kg21 z d21 by
subcutaneous injection, a dose shown to produce .95% inhibition of
primary tumor growth in mice14; group 2 was treated with the
fumagillin analogue TNP-470 at 30 mg/kg SC every other day; and
group 3 consisted of control animals that received a similar volume
of saline (TNP-470 buffer) or PBS (endostatin buffer).
After 16 weeks of treatment, animals were euthanized with
methoxyflurane. A blood sample was obtained from the right
ventricle for the analysis of serum cholesterol, which was performed
by an automated colorimetric assay on a Hitachi 917 instrument. The
heart and aorta were perfusion fixed with 2% paraformaldehyde and
dissected as described previously.16,17 Atherosclerotic plaques in the
aortas were used for histology, and plaque areas were only determined at the aortic origin.18 However, to document the extent of
disease in the total aorta, the unopened aortas were photographed.
The effects of these angiogenesis inhibitors on late-stage lesions
were then compared with their effects on early lesions, in which
intimal vessels were rarely observed. Animals were treated with
endostatin or TNP-470 from age 32 to 48 weeks (late-stage experiment) or from 6 to 22 weeks (early-stage experiment). Plaque
involvement was measured at the aortic origin. Because the aortas
from early-stage animals were not used for immunohistochemistry in
these experiments, the extent of atherosclerosis was also measured in
the entire aorta (percent surface area of Sudan IV–stained lesions).17
Animal health and weight was monitored throughout treatment.
The percentages of animals available for analysis in all studies were
92% (range, 86% to 100%), 89% (80% to 94%), and 91% (86% to
100%) for the endostatin, TNP-470, and control groups, respectively.
Chronic dermatitis and dental malocclusion were the most common
reasons for exclusion.
Immunohistochemistry
To detect intimal capillaries, the hearts and portions of the descending aorta with substantial lesions were embedded in paraffin,
sectioned (10 mm), digested with 0.02% protease XXIV (Sigma) for
April 6, 1999
1727
6 minutes at room temperature, and incubated with rabbit polyclonal
anti-human vWF antibody (Dako; 1:500 dilution) or rat monoclonal
anti-mouse CD31 antibody (Pharmigen; 20 mg/mL). When acetonefixed frozen sections were stained, the CD31 antibody was used at 5
mg/mL. The bound antibodies against vWF or CD31 were detected
with biotinylated goat anti-rabbit or rabbit anti-rat (mouse absorbed)
antibodies (Vector; 1:250 dilution) and the avidin-biotin peroxidase
complex (ABC standard kit, Vector). A red reaction product was
produced with 3-amino-9-ethyl carbazole substrate (AEC, Dako),
and sections were counterstained with Gill’s hematoxylin (Sigma).
Vascular sections incubated with nonimmune serum served as
negative controls. Positive staining of the endothelium on the lumen
and in adventitial capillaries served as an internal control. Intimal
vessels were identified under high power (3400) and counted when
both an endothelial cell nucleus and lumen were seen and when the
vessel was also observed in an adjacent section.
Smooth muscle cells were identified with a monoclonal IgG2a
antibody against human smooth muscle cell a-actin (Dako,
M0851).19 Sections were immersed in 0.1 mol/L sodium citrate, pH
6.0, and heated for 10 minutes in the microwave. Primary antibody
was bound to sections at 3 mg/mL and then visualized with
biotinylated anti-mouse IgG2a antibody (1:200 dilution, Amersham
RPN1181) and the ABC method described above. The adjacent
medial layer of vessels served as positive control.
Transmission Electron Microscopy
Aortas with atherosclerotic lesions were isolated from retired breeder
apoE 2/2 mice that were fed the Western diet for 20 weeks.
Extensive lesions in the descending aorta were divided transversely
into 2 segments. One segment was frozen and cryosectioned for
CD31 immunohistochemistry; the other was processed for TEM.
Only lesions that screened positive for intimal vessels were analyzed
on an electron microscope (EM 10, Zeiss).
Recombinant Murine Endostatin and
TNP-470 Treatments
Recombinant murine endostatin was prepared as described with the
expression plasmid TB01#8 transformed into the Escherichia coli
strain BL21:DE3.14,15 Induction results in a fusion protein with the
amino acid sequence MARRASVGTDHHHHHH at the N-terminus
followed by the sequence of endostatin, which corresponds to the
C-terminal, 184 amino acids of mouse collagen XVIII. Murine
endostatin was purified under denaturing conditions on an Ni21-NTA
column (QiaExpressionist Handbook, Qiagen). Purity was analyzed
by SDS-PAGE. Before use in apoE 2/2 mice, endostatin batches
were tested for inhibition of Lewis lung carcinoma growth in
C57Bl6/J mice.14
The apoE 2/2 mice received endostatin 20 mg/kg SC every day.
Animals tolerated the injections and showed regular weight gain,
activity levels, and no ulceration at the injection sites.
The fumagillin analogue TNP-470 was donated for these studies
by TAP Holdings, Deerfield, Ill. An injection solution of TNP-470 (3
mg/mL) was freshly prepared and mice received 30 mg/kg SC every
other day. Control animals received a similar volume of saline or
endostatin buffer.
Plaque Morphometry
Plaque Area
To determine the extent of atherosclerosis at the aortic origin, 40
serial sections (8 to 10 mm thick) of the aortic sinus were collected
on 10 slides.18 Every second slide was stained with hematoxylin and
eosin for morphometry. The rest were used for immunohistochemistry to count intimal vessels and determine intimal smooth muscle
cell contents. Plaque images were captured at 3100 magnification
with a Hitachi HV-C20 3CCD digital camera and measured with the
Leica Q500 MC image-analysis program.20 Lesion area for each
animal was reported as the mean intimal cross-sectional area
(in mm2). In the early-stage experiment, plaque involvement in the
entire aorta was also measured by image analysis as the percentage
of aortic surface area covered with lesions.17,20
1728
Angiogenesis Inhibitors Reduce Atherosclerosis
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Figure 1. Intimal neovascularization in
advanced lesions of apoE 2/2 mice. A,
Lesion from descending aorta shows vessels in intima identified by antibodies
against CD31 (magnification 3100). B,
Higher magnification (3250) of inset region
shows multiple capillaries. C, Marked intimal neovascularization detected by antibodies to CD31 is located near the medial
layer in a separate lesion (3400). D, Intimal
vessels in this plaque appear to run perpendicular to the media. E, An intimal capillary (large arrow), which extends along
the intima-media border (small arrow), is
identified by antibodies to vWF. F, Atherosclerotic plaque at aortic origin shows
large capillary-like vessel (large arrow) filled
with red blood cells and identified by antibodies against vWF. Bar in E and F at
3400 represents 20 mm.
Plaque Cell Density
Cell densities in aortic sinus plaques were evaluated to determine
whether endostatin or TNP-470 treatments affected lesion cellularity.
Cell nuclei in the intima of the aortic sinus plaques were counted at
3 midlesion levels (slides 4, 6, and 8). Cell density was reported as
the number of cells per intimal area (cells/mm2).
Intimal Smooth Muscle Cell Content
To determine whether angiogenesis inhibitor treatments affected
smooth muscle cell migration or proliferation in lesions, 2 slides
from the middle of the aortic sinus were stained for smooth muscle
cell a-actin. For each animal, the numbers of intimal smooth muscle
cells and total cell nuclei were counted in 8 nonoverlapping fields. A
smooth muscle cell was counted only if it stained for smooth muscle
cell a-actin and its cell nucleus was seen. The mean percentage of
smooth muscle cells relative to total cells was determined for each
animal.
Statistical Analysis
The inhibitory effects on median plaque growth induced by treatment
with TNP-470 and endostatin were compared with the KruskalWallis test. Data presented in Tables 1 and 2 were evaluated with the
Fisher exact test. Intimal cell density and smooth muscle cell content
of lesions from animals in treatment and control groups were
compared by nonparametric ANOVA analysis with adjustment for
multiple measurements from each animal (Table 3).
Results
ApoE 2/2 Mice Have Intimal Neovascularization
Advanced atherosclerotic lesions of apoE 2/2 mice were
first examined for the presence of intimal capillaries. Intimal
vessels were identified by immunohistochemistry with antibodies against the endothelial cell markers CD31 (Figure 1, A
through D) or vWF (Figure 1, E and F). Thin-walled,
capillary-like vessels were observed in lesions harvested from
the descending aorta (Figure 1, A through E) and the aortic
sinus (Figure 1F). Intimal neovascularization was also observed when advanced atherosclerotic lesions were examined
by TEM (Figure 2A). Increased numbers of capillaries
(Figure 2B) and endothelial cells with signs of early lumen
formation (Figure 2C) were seen in the adventitia adjacent to
atherosclerotic lesions.
The incidence of lesions with intimal vessels was generally
low in apoE 2/2 mice but increased in more extensive
lesions. No vessels were observed in fatty streak lesions and
only a few in lesions ,250 mm thick. Intimal vessels were
detected in 15 (13%) of 114 advanced aortic lesions from
cholesterol-fed apoE 2/2 mice aged 36 to 60 weeks (Table
1). The majority of plaques that contained intimal vessels (13
of 15; 87%) were .250 mm thick. The incidence of intimal
vessels in this group was increased 9-fold compared with
100- to 250-mm-thick lesions (P#0.0005).
Plaque Growth Is Reduced by Long-Term
Treatment With Endostatin or TNP-470
To demonstrate a role of plaque vessels in the progression of
atherosclerosis, apoE 2/2 mice were treated with an angiogenesis inhibitor, either endostatin or TNP-470, for a prolonged period. A prolonged treatment schedule was chosen
because of uncertainty about the onset of neovascularization
in murine lesions. On the basis of our initial survey of mice
younger than 20 weeks, plaques thicker than 250 mm were
Moulton et al
April 6, 1999
1729
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Figure 2. TEM images of intimal and adventitial vessels associated with lesions in apoE 2/2 mice. A, Cross section through a plaque (pl) in
the lower thoracic aorta. An intimal vessel is located adjacent to the elastic membrane (em) of the aortic wall. The capillary lumen (L) contains
lymphocytes (lc) that are in close contact to the endothelial cell (ec) and platelets (p) with a-granules. Bar52 mm. B, Periaortic vessels in the
adventitia associated with an atherosclerotic plaque. The medial layer (ml) with smooth muscle cells (sm) is eroded by the plaque (pl) to a
thickness of '6 mm. The lacunar capillary (c) that is composed of 2 endothelial cells (ec) is void of pericytes. cf indicates collagenous fiber
bundles. Bar55 mm. C, Capillary sprout in adventitial layer of lesion shown in B. Two endothelial cells (ec) interconnected by punctiform (circled) and more complex cell contacts (arrows) form an early lumen (L) filled with amorphous substance. Arrowheads point to the basal lamina
and large numbers of pinocytotic vesicles. The extensions of the pericytic cell (pc) have not yet reached the presumed newly formed vessel.
Bar51 mm.
infrequent. We therefore induced substantial atherosclerosis
by feeding apoE 2/2 mice a diet that contained 0.15%
cholesterol, and at 20 weeks, we initiated treatments for 16
weeks.
The median aortic origin plaque area in 10 mice killed at
baseline was 0.250 mm2 (range, 0.170 to 0.348 mm2; n510),
represented by the dashed line in Figure 3. The remaining 3
groups of animals were treated with endostatin, TNP-470, or
buffer as described in Methods. The median plaque areas
were 0.321 (0.238 to 0.412; n510), 0.402 (0.248 to 0.533;
n515), and 0.751 mm2 (0.503 to 0.838; n512) for the
endostatin, TNP-470, and control groups, respectively
(P#0.0001). Despite the considerable variability in plaque
areas, there was little overlap between the treatment and
control groups (Figure 3). Because the extent of lesions
before treatment was 0.250 mm2, long-term treatment with
endostatin or TNP-470 appeared to inhibit plaque growth by
85% and 70%, respectively. Percent inhibition was calculated
by the formula 1003[12(median plaque area treated20.25)/
(median plaque area control20.25)].
The cholesterol levels in treatment and reference groups
were not different (Figure 4). Median cholesterol levels at the
time the animals were killed were 877 (range, 542 to 1045;
n510), 823 (461 to 1164; n515), and 918 mg/dL (558 to
1366; n512) in the endostatin, TNP-470, and control groups,
TABLE 1.
Maximum
Intimal Depth
.250 mm
100–250 mm
Total
respectively (P50.62). In a subset of animals, cholesterol
levels were similar before and after 4 weeks of treatment.
Median weights after treatment were 36 (31 to 37), 31 (24
to 34), and 35 g (26 to 49) for the endostatin, TNP-470, and
control groups, respectively. All animals increased weight
during treatment, but TNP-470 –treated animals gained only
4 g (P50.001) compared with 8 g for control and 9 g for
endostatin-treated animals.
Significant inhibitions of plaque growth by endostatin and
TNP-470 were also observed in a second experiment that
treated mice from ages 20 to 36 weeks. Median plaque areas
were 0.475 (0.308 to 0.598; n512), 0.370 (0.181 to 0.521;
n512), and 0.690 mm2 (0.591 to 0.720; n512) for the
animals in the endostatin, TNP-470, and control groups,
respectively (P50.001).
Intimal Vessels in Advanced Lesions
Vessel (1)
Vessel (2)
Plaque Total
Incidence
13
33
46
0.2826
2
66
68
0.0294
15
99
114
0.1316
P#0.0005, Fisher’s exact test.
Figure 3. Size of lesions at aortic origin after treatment with
angiogenesis inhibitors in mice aged 20 to 36 weeks. Control
(▫), endostatin (‚), and TNP-470 (V) animals were treated for 16
weeks as described in Methods. Dashed line centered at
0.25 mm2 represents median plaque area of aortic sinus lesions
measured in a cohort (n510) analyzed at 20 weeks.
1730
Angiogenesis Inhibitors Reduce Atherosclerosis
Figure 4. Serum cholesterol levels. Blood samples were collected at 36 weeks from controls (▫) and animals treated with
endostatin (‚) or TNP-470 (E).
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
For these experiments, plaque measurement at the aortic
origin was selected to detect a change in plaque thickness
rather than surface area. However, photographed images of
the entire aorta, taken before the aortas were dissected for
histology, also demonstrated a noticeable reduction of plaque
surface area in endostatin or TNP-470 –treated animals (not
shown).
Inhibition of Plaques by Endostatin or TNP-470 Is
Less Prominent in Early Lesions
It was then evaluated whether endostatin and TNP-470 were
as effective in animals with predominantly advanced lesions
as in animals with predominantly early lesions. In this
experiment, cholesterol-fed apoE 2/2 mice were treated
with endostatin or TNP-470 from age 32 to 48 weeks. Plaque
areas at the aortic origin were determined after the same
16-week treatment period (Figure 5). Median plaque areas for
the endostatin, TNP-470, and control groups were 0.422
(0.283 to 0.637; n512), 0.448 (0.260 to 0.566; n514), and
0.584 mm2 (0.426 to 0.911; n513), respectively. The inhibitions of plaque growth were smaller than in the previous
experiments but still significant (P50.002). Again, cholesterol levels were similar in all groups (P50.13). Final weights
were 39 (32 to 50), 33 (25 to 37), and 35 g (32 to 44) for the
Figure 6. Effect of angiogenesis inhibitors on early-stage lesions
in entire aorta. Extent of atherosclerosis was measured as percent surface area of aorta covered with Sudan IV–positive
lesions. Treatment with endostatin (‚) or TNP-470 (E) from age
6 to 22 weeks resulted in no significant difference in lesion
severity compared with controls (▫).
endostatin, TNP-470, and control mice, respectively (first
level test P#0.0001; TNP-470 versus control, P50.003;
endostatin versus control, P50.04, not significant by adjusted
probability value 0.017).
The effect of endostatin or TNP-470 treatment was then
evaluated in mice that had primarily early lesions without
significant intimal neovascularization. ApoE 2/2 mice were
started on the diet at 6 weeks and divided into endostatin,
TNP-470, and control groups. After treatment for 16 weeks,
lesions in the entire aorta were not very extensive, and no
significant difference was seen between treatment and control
groups (Figure 6; P50.88). Median percent plaque areas
were 9.6% (6.2% to 16.4%; n510), 11.7% (4.4% to 16.6%;
n58), and 14.4% (5.4% to 16.1%; n512) for the endostatin,
TNP-470, and control animals, respectively. Plaque measurements at the aortic origin also showed no difference (not
shown). Therefore, the inhibitory effects of endostatin and
TNP-470 were less pronounced during early atherogenesis.
Incidence of Intimal Vessels in Endostatin- or
TNP-470 –Treated Lesions Is Reduced
Aortic sinus plaques isolated from control and treated animals
were examined for the presence of intimal vessels. The
percentage of plaques that contained any intimal vessels was
significantly smaller in treated mice (5% for endostatin,
P50.032; 0% for TNP-470, P50.003) than in controls (29%)
(Table 2).
Smooth Muscle Cell Content and Cellular Density
Of Lesions
The magnitude of plaque inhibition observed with endostatin
or TNP-470 raised the possibility that these agents not only
TABLE 2. Intimal Neovascularization in Control and
Angiogenesis-Inhibitor Treated Lesions
Treatment
Figure 5. Size of lesions at aortic origin after treatment with
angiogenesis inhibitors in mice aged 32 to 48 weeks. Median
plaque areas for endostatin-treated (‚) and TNP-470-treated (E)
apoE 2/2 mice are reduced compared with controls (▫).
P†
Aortic Sinus Plaque, n
Neovascularization (1)*
24
7 (29%)
Endostatin
22
1 (5%)
0.032
TNP-470
27
0
0.003
Control
*Plaques with $1 intimal capillary were counted positive (see Methods).
†Fisher’s exact test.
Moulton et al
TABLE 3. Intimal Smooth Muscle Content and Cell Density
of Lesions
Treatment
SMC, %
P*
Cell Density,
cells/mm2
P*
Control (n512)
20.4 (14.3–30.5)
Endostatin (n510)
19.4 (16.2–22.7)
zzz
0.20
1183 (867–1640)
1646 (952–2361)
zzz
#0.0005
TNP-470 (n515)
19.8 (14.6–23.5)
0.55
1261 (855–1926)
0.22
Median values and range of values for individual animals are listed.
*Compared with control by nonparametric ANOVA.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
affect plaque angiogenesis but may also affect other cell types
found within atherosclerotic lesions. To evaluate whether
TNP-470 and endostatin influence smooth muscle cell migration and proliferation in atherosclerotic lesions, the smooth
muscle cell contents of lesions from treatment and control
mice were compared (Table 3). The median smooth muscle
cell contents of aortic sinus lesions from control mice were
20.4% and were similar for endostatin-treated (19.4%;
P50.20) and TNP-470 –treated mice (19.8%; P50.55).
The cellular density of atherosclerotic lesions from treated
and control mice were determined at 3 levels of the aortic
sinus, as described in Methods. The calculated median cell
densities were similar between the control and TNP-470
groups but were increased in the endostatin group (1645
cells/mm2; P#0.0005).
Discussion
The present study demonstrates for the first time that atherosclerotic lesions in the aorta of apoE 2/2 mice contain
intimal vessels, as do human lesions. This observation is
remarkable because the absolute dimensions of murine lesions are much smaller. The incidence of intimal vessels in
advanced murine lesions was 13%, whereas the incidence in
human lesions has been reported at 40% to 53%.4 Prior
studies showed infrequent vasa vasorum in normal mouse
aortas. Comparisons of the lamellar structure across many
species suggested that a minimal vessel wall thickness was
required for the presence of vasa vasorum.21 Intravital studies
of tissue oxygenation in tumors demonstrated that significant
hypoxia and acidosis occur when the distance from a capillary exceeds 100 mm.22 Given the thin media of the murine
aorta, only relatively large lesions may exceed the limit at
which growing plaques require additional sources of perfusion beyond the artery lumen and adventitial vessels.
Our results are consistent with this assumption. However,
although the intimal thickness of many plaques from apoE
2/2 mice exceeded 100 mm, only 13% of these lesions
showed intimal neovascularization. It is possible that the
metabolic requirements of plaque tissues depend on lesion
composition and differ from tumors. In addition, the contribution from adventitial vessels was not measured.
The vessel density in vascularized plaques ranged from 1
to 17 capillaries per high-power field, but the vessel number
in these lesions did not correlate with intimal thickness. For
example, 1 plaque (Figure 1A) had .30 vessels and a
maximal intimal thickness of 312 mm, whereas a 486-mmthick lesion (not shown) had 8 to 10 capillaries. It is likely
that factors other than size, such as cell density, leukocyte
April 6, 1999
1731
infiltrates, and matrix composition, also influence the development of plaque vessels. Despite this lack of a linear
correlation, plaque size may be an indicator of the presence of
intimal vessels, because their incidence in lesions .250 mm
was increased 9-fold compared with smaller lesions.
The observation of intimal vessels in lesions from apoE
2/2 mice provided the opportunity to test whether potent
angiogenesis inhibitors affect the progression of atherosclerosis in these animals. The results of 2 separate experiments
conducted on animals treated from ages 20 to 36 weeks
showed significant inhibition of atherosclerosis with no effect
on cholesterol levels (Figure 3). Compared with animals
killed at baseline, endostatin and TNP-470 treatments inhibited atherosclerosis by 85% and 70%, respectively. In the
second experiment, either treatment significantly inhibited
plaque growth, but the degree of inhibition by endostatin was
less than that by TNP-470.
Significant inhibition of plaque growth by endostatin or
TNP-470 was seen even when the treatment was delayed until
32 weeks, although the degree of inhibition was smaller
(Figure 5). One potential explanation for the smaller inhibition could be that the plaque growth rate from age 20 to 36
weeks is different than that from age 32 to 48 weeks. Second,
both endostatin and TNP-470 are reversible inhibitors of
endothelial cell proliferation and appear to exert few effects
on quiescent nonproliferating endothelium.15,23 Therefore, the
effects of these inhibitors on plaques with established intimal
vessels might be different than those on plaques that develop
intimal neovascularization during the treatment period.
Interestingly, we saw little effect when treatments were
performed during early stages of atherosclerosis (Figure 6). In
apoE 2/2 mice fed the Western diet, fatty streaks are
typically seen from 8 to 20 weeks, and plaques with smooth
muscle cells are initially observed at 15 weeks.19 Our results
therefore suggest that neither endostatin nor TNP-470 significantly affected foam cell and early fibromuscular lesions.
When lesions from endostatin- or TNP-470 –treated animals were examined, few intimal vessels were observed.
However, this correlation does not prove that inhibition of
plaque growth occurred because intimal neovascularization
was decreased. The reduced neovascularization could merely
be a consequence of reduced plaque size rather than an effect
of angiogenesis inhibitors and the cause of reduced
atherosclerosis.
The process of atherosclerosis involves multiple factors
that control inflammation, cell proliferation and migration,
cholesterol metabolism, and interactions between cells,
blood, and matrix. It is possible these agents altered the
functional characteristics of the endothelium that influence
leukocyte adhesion, transmigration, or activation. TNP-470
has been shown to enhance E-selectin expression, but this
would not be predicted to inhibit lesion development.23
Smooth muscle cell proliferation and migration are inhibited
by TNP-470 in vitro, but this requires doses that are 30- to
70-fold greater than the doses for endothelial cells.13,24
Despite these in vitro findings, no significant differences in
the smooth muscle cell contents of lesions from treated and
control animals were observed in the present study.
1732
Angiogenesis Inhibitors Reduce Atherosclerosis
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Currently, only limited information on the effects of
TNP-470 and endostatin is available. TNP-470 regulates
cyclin activity in endothelial cells and forms a covalent bond
to methionine aminopeptidase-2, a cobalt-dependent metalloprotease.25,26 Additional investigations of the mechanisms of
action for TNP-470 and endostatin may therefore provide
insights toward understanding their effects on plaque growth.
Although the present study cannot provide conclusive
evidence for causality, the combined findings provide strong
support for the hypothesis that intimal vessels contribute to
the progression of atherosclerosis. First, similar inhibition of
atherosclerosis was observed with 2 very different agents that
share a potent inhibitory effect on endothelial cell proliferation. Second, the inhibition of plaque growth by these agents
was associated with a decreased incidence of intimal neovascularization. Finally, these inhibitors showed little effect
during early stages of plaque development, when intimal
neovascularization was unlikely to occur.
The endothelium of plaque vessels may be qualitatively
different from the arterial endothelium that covers the plaque. In
future studies, characterization of these potential differences
may distinguish pathways for leukocyte exchange in the plaque
that are selective for the plaque microvasculature.
If studies in other models confirm our present observations,
new treatments directed at intimal vessels might be considered to augment established interventions that reduce atherosclerosis. Clinical trials of angiogenesis inhibitors for the
treatment of tumors, macular degeneration, and other diseases
characterized by neovascularization may provide opportunities to evaluate the effects of these agents on concurrent
atherosclerosis. Finally, the recent demonstrations of therapeutic angiogenesis in ischemic myocardium and peripheral
limbs raise questions as to whether endothelial cell growth
factors will promote plaque angiogenesis, growth, or vulnerability. Exogenous endothelial cell growth factors might not
induce such effects because these factors are already produced in plaque tissue.27,28
Acknowledgments
Studies were supported by a Physician Scientist Award (K11 HL02563) from the National Heart, Lung, and Blood Institute and the
Harvard Medical School 50th Anniversary Program (K.S.M.). We thank
Thomas Boehm for advice on endostatin production, Dipak Panigrahy
for testing endostatin preparations, Allison Brown for assistance with
image analysis, Kerstin Bahr for TEM, and Elizabeth Allred for
statistical analyses. We are grateful to Peter Libby and Joyce Bischoff
for helpful discussions and review of this manuscript.
References
1. Barger AC, Beeuwkes R, Lainey L, Silverman KJ. Hypothesis: vasa
vasorum and neovascularization of human coronary arteries. N Engl
J Med. 1984;310:175–177.
2. Koester W. Endarteritis and arteritis. Berl Klin Wochenschr. 1876;13:
454 – 455.
3. Zhang Y, Cliff WJ, Schoefl GI, Higgins G. Immunohistochemical study
of intimal microvessels in coronary atherosclerosis. Am J Pathol. 1993;
143:164 –172.
4. O’Brien ER, Garvin MR, Dev R, Stewart DK, Hinohara T, Simpson JB,
Schwartz SM. Angiogenesis in human coronary atherosclerotic plaques.
Am J Pathol. 1994;145:883– 894.
5. Polverini PJ, Cotran RS, Gimbrone MA Jr, Unanue ER. Activated macrophages induce vascular proliferation. Nature. 1977;269:804 – 806.
6. Kaartinen M, Penttila A, Kovanen PT. Mast cells accompany
microvessels in human coronary atheromas. Atherosclerosis. 1996;123:
123–131.
7. Auerbach R, Sidky YA. Nature of the stimulus leading to lymphocyteinduced angiogenesis. J Immunol. 1979;123:751–754.
8. O’Brien KD, McDonald TO, Chait A, Allen MD, Alpers CE. Neovascular
expression of E-selectin, ICAM-1, and VCAM-1 in human atherosclerosis and their relation to intimal leukocyte content. Circulation. 1996;93:
672– 682.
9. Paterson JC. Capillary rupture with intimal hemorrhage as a causative
factor in coronary thrombosis. Arch Pathol. 1938;25:474 – 487.
10. Tenaglia AN, Peters KG, Sketch MH Jr, Annex BH. Neovascularization
in atherectomy specimens from patients with unstable angina. Am
Heart J. 1998;135:10 –14.
11. Gross JL, Moscatelli D, Rifkin DB. Increased capillary endothelial cell
protease activity in response to angiogenic stimuli in vitro. Proc Natl
Acad Sci U S A. 1983;80:2623–2627.
12. Galis ZS, Sukhova GK, Lark MW, Libby P. Increased expression of matrix
metalloproteinases and matrix degrading activity in vulnerable regions of
human atherosclerotic plaques. J Clin Invest. 1994;94:2493–2503.
13. Ingber D, Fujita T, Kishimoto S, Sudo K, Kanamaru T, Brem H, Folkman
J. Synthetic analogues of fumagillin that inhibit angiogenesis and
suppress tumour growth. Nature. 1990;348:555–557.
14. Boehm T, Folkman J, Browder T, O’Reilly MS. Antiangiogenic therapy
of experimental cancer does not induce acquired drug resistance. Nature.
1997;390:404 – 407.
15. O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, Flynn E,
Birkhead JR, Olsen BR, Folkman J. Endostatin: an endogenous inhibitor
of angiogenesis and tumor growth. Cell. 1997;88:277–285.
16. Palinski W, Ord VA, Plump AS, Breslow JL, Steinberg D, Witztum JL.
ApoE-deficient mice are a model of lipoprotein oxidation in atherogenesis. Arterioscler Thromb. 1994;14:605– 616.
17. Tangirala RK, Rubin EM, Palinski W. Quantitation of atherosclerosis in
murine models. J Lipid Res. 1995;36:2320 –2328.
18. Paigen B, Morrow A, Holmes PA, Mitchell D, Williams RA. Quantitative
assessment of atherosclerotic lesions in mice. Atherosclerosis. 1987;68:
231–240.
19. Nakashima Y, Plump AS, Raines EW, Breslow JL, Ross R. ApoEdeficient mice develop lesions of all phases of atherosclerosis throughout
the arterial tree. Arterioscler Thromb. 1994;14:133–140.
20. Dong ZM, Chapman S, Brown A, Frenette P, Hynes RO, Wagner DD.
Combined role of P- and E-selectins in atherosclerosis. J Clin Invest.
1998;102:145–152.
21. Wolinsky H, Glagov S. A lamellar unit of aortic medial structure and
function in mammals. Circ Res. 1967;20:99 –111.
22. Torres Filho I, Leunig M, Yuan F, Intaglietta M, Jain R. Noninvasive
measurement of microvascular and interstitial oxygen profiles in a human
tumor in SCID mice. Proc Natl Acad Sci U S A. 1994;91:2081–2085.
23. Budson AE, Ko L, Brasel C, Bischoff J. The angiogenesis inhibitor
AGM-1470 selectively increases E-selectin. Biochem Biophys Res
Commun. 1996;225:141–145.
24. Koyama H, Nishizawa Y, Hosoi M, Fukumoto S, Kogawa K, Shioi A,
Morii H. The fumagillin analogue TNP-470 inhibits DNA synthesis of
vascular smooth muscle cells stimulated by PDGF and IGF-I. Circ Res.
1996;79:757–764.
25. Sin N, Meng L, Wang MQ, Wen JJ, Bornmann WG, Crews CM. The
anti-angiogenic agent fumagillin covalently binds and inhibits the
methionine aminopeptidase, MetAP-2. Proc Natl Acad Sci U S A. 1997;
94:6099 – 6103.
26. Abe J, Zhou W, Takuwa N, Taguchi J, Kurokawa K, Kumada M, Takuwa
Y. A fumagillin derivative angiogenesis inhibitor, AGM-1470, inhibits
activation of cyclin-dependent kinases and phosphorylation of retinoblastoma gene product in vascular endothelial cells. Cancer Res. 1994;
54:3407–3412.
27. Brogi E, Winkles JA, Underwood R, Clinton SK, Alberts GF, Libby P.
Distinct patterns of expression of fibroblast growth factors and their
receptors in human atheroma and nonatherosclerotic arteries. J Clin
Invest. 1993;92:2408 –2418.
28. Couffinhal T, Kearney M, Witzenbichler B, Chen D, Murohara T,
Losordo DW, Symes J, Isner JM. VEGF(VPF) in normal and atherosclerotic human arteries. Am J Pathol. 1997;150:1673–1685.
Angiogenesis Inhibitors Endostatin or TNP-470 Reduce Intimal Neovascularization and
Plaque Growth in Apolipoprotein E−Deficient Mice
Karen S. Moulton, Eric Heller, Moritz A. Konerding, Evelyn Flynn, Wulf Palinski and Judah
Folkman
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Circulation. 1999;99:1726-1732
doi: 10.1161/01.CIR.99.13.1726
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1999 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circ.ahajournals.org/content/99/13/1726
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/