Cardiovascular Research (2014) 101, 145–154
doi:10.1093/cvr/cvt234
Multifunctional regulation of angiogenesis
by high-density lipoproteins
Hamish C.G. Prosser 1,2, Joanne T.M. Tan1,2, Louise L. Dunn 1,2, Sanjay Patel1,2,3,
Laura Z. Vanags 1,2, Shisan Bao 2, Martin K.C. Ng1,2,3†, and Christina A. Bursill 1,2†*
1
The Heart Research Institute, 7 Eliza Street, Newtown, Sydney, NSW 2042, Australia; 2Department of Medicine, University of Sydney, Sydney, Australia; and 3Department of Cardiology,
Royal Prince Alfred Hospital, Sydney, Australia
Received 11 April 2013; revised 19 September 2013; accepted 2 October 2013; online publish-ahead-of-print 15 October 2013
Time for primary review: 7 days
Aims
High-density lipoproteins (HDL) exert striking anti-inflammatory effects and emerging evidence suggests that they may
augment ischaemia-mediated neovascularization. We sought to determine whether HDL conditionally regulates angiogenesis, depending on the pathophysiological context by (i) inhibiting inflammation-induced angiogenesis, but also; (ii)
enhancing ischaemia-mediated angiogenesis.
.....................................................................................................................................................................................
Methods
Intravenously delivered apolipoprotein (apo) A-I attenuated neovascularization in the murine femoral collar model of
inflammation-induced angiogenesis, compared with phosphate-buffered saline infused C57BL6/J mice (58%), P , 0.05.
and results
Conversely, apoA-I delivery augmented neovessel formation (75%) and enhanced blood perfusion (45%) in the murine
hindlimb ischaemia model, P , 0.05. Reconstituted HDL (rHDL) was tested on key angiogenic cell functions in vitro.
rHDL inhibited human coronary artery endothelial cell migration (37.9 and 76.9%), proliferation (15.7 and 40.4%), and tubulogenesis on matrigel (52 and 98.7%) when exposed to two inflammatory stimuli: tumour necrosis factor-a (TNF-a) and
macrophage-conditioned media (MCM). In contrast, rHDL significantly augmented hypoxia-stimulated migration (36.9%),
proliferation (135%), and tubulogenesis (22.9%), P , 0.05. Western blot and RT–PCR analyses revealed that these
divergent actions of rHDL were associated with conditional regulation of hypoxia-inducible factor-1a (HIF-1a), vascular
endothelial growth factor (VEGF) and VEGF receptor 2, which were attenuated in response to TNF-a (40.4, 41.0, and
33.2%) and MCM (72.5, 30.7, and 69.5%), but augmented by rHDL in hypoxia (39.8, 152.6, and 15.7%%), all P , 0.05.
.....................................................................................................................................................................................
Conclusion
HDL differentially regulates angiogenesis dependent upon the pathophysiological setting, characterized by suppression of
inflammation-associated angiogenesis, and conversely, by the enhancement of hypoxia-mediated angiogenesis. This has
significant implications for therapeutic modulation of neovascularization.
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
High-density lipoprotein † Angiogenesis † Inflammation † Ischaemia
1. Introduction
Epidemiological studies have reported inverse associations between
high-density lipoprotein (HDL) levels and risk of developing cardiovascular disease.1,2 In vitro studies and pre-clinical models have revealed that
HDL and its main protein constituent, apolipoprotein (apo) A-I, possess
potent vasculo-protective properties, providing anti-oxidant, antiinflammatory, and anti-apoptotic actions on the endothelium.3 – 6
Nevertheless, recent clinical trials of HDL-raising therapies have been
disappointing, demonstrating neutral effects on cardiovascular events
in high-risk patients.7 Despite a significant amount of work, the vascular
†
biological effects of HDL remain incompletely understood and further
efforts to translate HDL modulation into therapy require a fuller elucidation of its properties.
Angiogenesis is critical in postnatal physiological processes such as
wound healing and in tissue neovascularization in response to ischaemic
injury. In contrast to the advantages bestowed by neovascularization in
response to ischaemia, neovessel formation in inflammatory pathologies, such as atherosclerosis, accelerates their progression.8,9 Increasing
evidence also indicates that plasma HDL/apoA-I levels are associated
with a number of pathologies in which angiogenesis plays a key role.
For example, infusions of reconstituted HDL (rHDL) into humans
Equal senior authorship.
* Corresponding author. Tel: +61 2 8208 8905; fax: +61 2 9565 5584, Email: [email protected]
Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2013. For permissions please email: [email protected].
146
reduce plaque size and lipid content, the expression of the inflammatory
marker vascular cell adhesion molecule-1 (VCAM-1), and reduce the
presence of macrophages.2,10,11 In support of the human intervention
studies, epidemiological and animal intervention studies show that
higher HDL is associated with reduced atherosclerosis and macrophage
content, and improved plaque stability.12 – 15 Plaque macrophages
secrete numerous pro-angiogenic growth factors and cytokines that
promote plaque neovascularization, which subsequently accelerate
plaque growth and increase plaque instability/rupture.9,16 Therefore,
the inhibitory actions of HDL on plaque size and inflammation indicate
that HDL may be associated with the attenuation of inflammatory angiogenesis. In contrast, there is emerging evidence that HDL augment
physiological, ischaemia-driven angiogenic processes. For example, infusions of rHDL into mice augment blood flow recovery following ischaemia.17 Consistent with this, prospective studies report that elevated
serum HDL is associated with improved survival and prognosis following
myocardial infarction, indicating improved ischaemia-driven neovascularization.18,19 Taken together, these studies suggest that HDL may
modulate angiogenesis in a multifunctional manner, depending on the
pathophysiological context.
Here, we present for the first time a comprehensive comparative
study that demonstrates HDL differentially regulates angiogenesis
in vivo and in vitro—such that it inhibits inflammation-induced angiogenesis, but also augments hypoxia-driven neovascularization. We also
identify that the key intracellular angiogenic modulators such as hypoxiainducible factor-1a (HIF-1a), vascular endothelial growth factor
(VEGF), and its receptor VEGFR2 are conditionally regulated, which
can mechanistically explain these unique vascular biological effects of
HDL.
2. Methods
2.1 Preparation of native HDL (nHDL),
apoA-I, and rHDL
In the preparation of nHDL, the HDL fraction (1.063–1.21 g/mL) was
isolated from pooled samples of normal human plasma (.5 donors,
Red Cross Blood Service, Alexandria, New South Wales, Australia) by
ultracentrifugation then dialysed against phosphate-buffered saline
(PBS) and conformed with the principles outlined in the Declaration
of Helsinki. For the isolation of apoA-I, HDL was delipidated and subjected to anion-exchange chromatography using a fast protein liquid chromatography system.20 For the preparation of rHDL, purified apoA-I was
complexed with 1-palmitoyl-2-linoleoyl-phosphatidylcholine (PLPC)
using the cholate dialysis method.21 The final PLPC:apoA-I ratio was 80:1.
Assessment of rHDL oxidation was determined by measuring the
conversion of methionine to methionine sulphoxide by HPLC, as previously described.22 Using this method, it was found that rHDL had undergone nominal oxidation of 15.2% per 100 mg protein.
H.C.G. Prosser et al.
was used for analgesia, administered subcutaneously daily for 4 days following surgery.
2.2.1 Murine peri-arterial cuff model
Under inhalation of methoxyflurane, a non-occlusive silicon cuff was
placed around the right femoral artery of mice. A sham operation was
performed on the left femoral artery to provide a parallel control. Adequate anaesthesia was confirmed throughout surgery by testing the
pedal reflex. Mice were randomly divided to receive either PBS or
apoA-I (40 mg/kg) intravenously every second day following surgery.
All animals were sacrificed 3 weeks after surgery, by overdose of isoflurane via inhalation, and perfused with PBS (10 ml) via the left ventricle
before the femoral arteries of each leg (complete with cuff) were
excised for histochemical analyses. In a separate cohort of mice, RNA
was isolated from excised collared and non-collared arteries and used
to assess changes in mRNA levels of CCL2 and TNF-a. For further
details, see Supplementary materials online.
2.2.2 Murine hindlimb ischaemia model
Under inhalation of methoxyflurane, the left femoral artery was ligated
(7-0 silk) above both the epigastrica and Profunda femoris before severing them distal to the ligation. The femoral artery and vein were also
completely excised as distal as the popliteal from the hindlimb of mice
(n ¼ 11; per treatment group).23 A sham procedure was performed
on the opposite hindlimb to provide a parallel control. Mice received
i.v. injections of PBS or apoA-I (40 mg/kg) every second day following
surgery. Blood reperfusion of the hindlimb was determined by Laser
Doppler perfusion imaging (moorLDI2-IR, Moor Instruments, Devon,
UK), which was performed prior to surgery, immediately following
surgery, then at Days 3, 7, and 10 post-surgery under inhalation of
isoflurane (2–3%). Following sacrifice by overdose of isoflurane via
inhalation, the gastrocnemius muscles of both hindlimbs were collected
for histological and gene expression analyses. For further details, see
Supplementary materials online.
2.3 Angiogenic cell function assays
The effects of rHDL, apoA-I, and nHDL on the key angiogenic assays of
tubulogenesis, cell proliferation and cell migration were assessed in two
inflammatory conditions (TNF-a and macrophage-conditioned media,
MCM) or hypoxia (1% O2/5% CO2) using male human coronary
artery endothelial cells (HCAECs). For details, see Supplementary
materials online.
2.4 Immunohistochemistry
2.2 Animal studies
2.4.1 Cuff model
Femoral arteries were excised following cuff implantation and probed
for capillaries (vWF, Dako), as well as macrophages (F4/80, Abcam).
Intima and media thickness was determined on sections stained with
Masson’s trichrome.
All experimental procedures and protocols were conducted with approval from the Central Sydney Area Health Service Animal Welfare
Committee (#2008– 002C), and conformed to the Guide for the Care
and Use of Laboratory Animals (United States National Institute of
Health). Male C57BL/6J mice, 8 weeks of age (22 –25 g), were housed
under controlled conditions (12 h light/dark cycle, 228C) with access
to water and standard mouse chow ad libitum. Carprofen (5 mg/kg)
2.4.2 Hindlimb ischaemia model
The gastrocnemius muscle from six ischaemic and six sham hindlimbs
was OCT-embedded (Tissue-Tek, Sakura, Torrance, USA) and frozen
in liquid nitrogen, sectioned (5 mM) and incubated with primary antibodies specific for vWF (Abcam), a-actin smooth muscle (Sigma), and
laminin (Abcam). For details, see Supplementary information online.
147
Multifunctional regulation of angiogenesis by HDL
2.5 Protein and gene analyses
Changes in gene and protein expression were determined by real-time
PCR and western blotting, respectively, to determine the expression of
key hypoxic and inflammatory angiogenic mediators including HIF-1a,
VEGF, VEGF receptor 2 (VEGFR2), the NF-kB subunit p65, tumour necrosis factor-a (TNF-a), and the chemokines, CXCL12 and CCL2.
ELISA was used to determine CCL2 protein. For details, see Supplementary materials online.
2.6 Statistics
Values are expressed as the mean + SEM. Differences between
treatment groups were calculated using a one-way ANOVA and
Tukey’s post hoc or Student’s t-test when comparing two groups at a
single time point. A two-way ANOVA with Bonferroni’s comparison
test post hoc was performed when comparing data at multiple time
points. Probability values of ,0.05 were considered statistically
significant.
3. Results
3.1 ApoA-I inhibits inflammation-induced
neovascularization in vivo
Placement of a collar around the femoral artery of mice induced a
pro-angiogenic and inflammatory response, characterized by significant
increases in adventitial neovessels (identified by vWF immunostaining;
collared: 3.00 + 0.35 vs. non-collared: 0.40 + 0.02, P , 0.01,
Figure 1A), macrophage infiltration (collared: 5.00 + 0.70% vs. noncollared: 0, P , 0.01, Figure 1B), and intima-media thickness (collared:
0.42 + 0.05 vs. non-collared: 0.18 + 0.02, P , 0.01, Figure 1C). Infusion
of apoA-I had no effect on plasma lipid profiles (see Supplementary material online, Table S1); however, it significantly inhibited the
pro-angiogenic and inflammatory response induced by the collar; attenuating adventitial neovessel formation by 58% (1.25 + 0.12 vs.
3.00 + 0.35% for PBS controls, P , 0.005, Figure 1A), as well as adventitial macrophage infiltration by 60%, determined by F4/80 staining
(2.00 + 0.70 vs. 5.00 + 0.70% for PBS controls, P , 0.05, Figure 1B).
Figure 1 ApoA-I inhibits inflammation-induced neovascularization in vivo. A non-occlusive collar was placed around the right femoral artery of C57Bl/6J
mice. Mice (n ¼ 6/treatment) received i.v. infusions of PBS or apoA-I on alternate days following collar placement. Three weeks post-collar placement mice
were sacrificed and the femoral arteries sectioned for immunohistochemical determination of (A) von Willebrand Factor (vWF) to identify adventitial neovessels (brown staining on photomicrographs) and (B) F4/80 to identify macrophages (brown staining on photomicrographs). Sections also underwent
trichrome staining (C) to determine the intima/intima + media (I/I + M) ratio (blue represents collagen and red represents muscle). RT – PCR was performed on femoral arteries with or without collar implantation to assess the expression of TNF-a (D) and CCL2 (E). Photomicrographs represent
stained femoral artery sections from PBS- and apoA-I-treated mice at sacrifice (×20 magnification). *P , 0.05 **P , 0.001 vs. mice with no collar,
^P , 0.05 vs. PBS-injected mice with collared femoral arteries.
148
ApoA-I infusion had no significant effect on intima-media thickness when
compared with PBS-infused mice (0.50 + 0.04 vs. 0.42 + 0.05 for
PBS-infused mice, Figure 1C). Collar implantation caused a marked increase in the mRNA levels of the key inflammatory genes TNF-a
(40-fold) and CCL2 (5.7-fold, P , 0.001, Figure 1D and E). ApoA-I infusion into mice significantly attenuated both TNF-a (50%, P , 0.05) and
CCL2 (43%, P , 0.05) mRNA levels in response to collar placement,
when compared with PBS control mice (Figure 1D and E).
3.2 ApoA-I augments neovascularization
in the ischaemic hindlimb in vivo
Laser Doppler imaging revealed surgical induction of hindlimb ischaemia
by unilateral femoral artery ligation reduced blood flow equally in both
PBS and apoA-I-treated mice at Day 0 (Figure 2A). Compared with
PBS-infused animals, apoA-I improved blood flow recovery in the ischaemic limb at all-time points, reaching significance at Days 7 (PBS: 0.38 +
0.04 vs. apoA-I: 0.55 + 0.06 by laser Doppler perfusion index) and 10
(PBS: 0.50 + 0.06 vs. apoA-I: 0.72 + 0.05), P , 0.05, Figure 2A. Consistent with these findings, apoA-I increased capillary density in the gastrocnemius muscle of ischaemic hindlimbs (PBS: 0.16 + 0.02 vs. apoA-I:
0.28 + 0.02, vWF staining per myocyte, P , 0.01, Figure 2B), but had
H.C.G. Prosser et al.
no significant effect on arteriole diameter or density (see Supplementary
material online, Figure S1A and B). Furthermore, apoA-I increased hindlimb mRNA levels of two pro-angiogenic factors that play a key role in
ischaemia: VEGF (47.1%) and chemokine CXCL12 (168.9%), above
that of PBS controls (Figure 2C and D). ApoA-I infusions had no effect
on plasma lipid profiles (see Supplementary material online, Table S1).
3.3 rHDL attenuates
inflammatory-stimulated angiogenesis
in vitro
In HCAECs, inflammatory stimulation with either TNF-a or MCM significantly increased tubule formation (TNF-a: 50.4% and MCM:
50.9%), cell proliferation (TNF-a: 17.7% and MCM: 53.0%), and cell migration as assessed by live cell imaging (TNF-a: 35.3% and MCM: 71.1%),
compared with control cells (Figure 3A–C and see Supplementary material online, Figure S2A–C, all P , 0.05). rHDL treatment alone promoted endothelial cell proliferation (46.5%) and cell migration (32.7%)
compared with PBS-treated cells, but had no significant effect on tubulogenesis. Despite these actions, rHDL pre-treatment significantly attenuated the inflammatory TNF-a- and MCM-driven stimulation of
tubulogenesis, cell proliferation, and cell migration back to levels
Figure 2 ApoA-I augments blood flow recovery and neovascularization in ischaemic hindlimbs in vivo. Femoral artery ligation was performed on C57Bl/6J
mice (n ¼ 11/treatment group). Mice received i.v. injections of PBS or apoA-I on alternate days following ligation until sacrifice 10 days later. (A) Blood flow
perfusion was determined using a Laser Doppler; upper images represent high (red) to low (blue) blood flow in animals treated with PBS or apoA-I at Day
10. (B) Capillaries were identified in ischaemic hindlimb sections using immunocytochemistry for vWF (green), and quantified as number of vessels per
myocyte (blue). Photomicrographs represent ischaemic gastrocnemius muscle stained for capillaries (green) and myocytes (blue) from PBS- and
apoA-I mice following sacrifice. (C) RT – PCR was performed on RNA isolated from ischaemic limbs to assess expression of VEGF and (D) CXCL12.
RNA expression was normalized to Beta-2-microglobulin (B2M). *P , 0.05, **P , 0.005, ***P , 0.0005 vs. PBS controls.
Multifunctional regulation of angiogenesis by HDL
149
Figure 3 rHDL attenuates inflammatory-stimulated angiogenesis in vitro. HCAECs were pre-incubated with PBS or rHDL (16 h, 600 mg/mL) and stimulated with 1 ng/mL TNF-a (6 h). (A) Matrix gel tubulogenesis was assessed as total number of branches/well. Panels on left are representative tubule images
for each treatment condition. (B) Cellular proliferation was determined using the Click-It assay. (C) Migration was assessed using live cell imaging. (D) The
migration of HCAECs treated with PBS or rHDL towards interferon-g-stimulated human monocyte-derived macrophages (stimulated) in the lower
chamber of transwell chambers was also assessed. rHDL treatment has dose-dependent inhibitory effects (0– 250 mg/mL rHDL, inset graph). Data are
presented as % of PBS control, n ¼ 4 and repeated four to six times. *P , 0.05, **P , 0.005 vs. PBS controls; ^P , 0.05, ^^^P , 0.0005 vs. stimulated
controls.
equivalent to, or below that of control endothelial cells, and significantly
less than TNF-a- and MCM only controls (P , 0.05).
Finally, using transwells, HCAECs were placed in the upper chamber
and allowed to migrate through a porous membrane towards macrophages grown on the base of the lower chamber which, upon
IFN-g-stimulation, secrete a milieu of pro-angiogenic growth factors
and cytokines. This method closely mimics that which occurs during
inflammatory-driven angiogenesis in vivo. Pre-incubation of HCAECs
with rHDL significantly arrested HCAEC migration towards
IFN-g-stimulated macrophages by 84.6% (P , 0.001), compared with
PBS-treated control cells. Furthermore, rHDL inhibited migration at
lower doses and in a dose-dependent manner (33.8, 51.4, and 81.9%
for 50, 100, and 250 mg/mL respectively, P , 0.001, Figure 3D).
3.4 rHDL augments hypoxia-stimulated
HCAEC angiogenesis in vitro
In contrast to rHDL inhibiting angiogenesis in inflammatory conditions,
rHDL significantly augmented key angiogenic processes in hypoxia.
Hypoxia alone significantly increased tubulogenesis (17.9%, Figure 4A),
cell proliferation (34.6%, Figure 4B), and cell migration (23.0%,
Figure 4C) compared with normoxic PBS controls, P , 0.05. Strikingly,
pre-treatment with rHDL significantly augmented hypoxia-induced
tubulogenesis, cell proliferation, and cell migration by 19.4, 107.7, and
30.0%), respectively, relative to hypoxia alone (all; P , 0.05).
3.5 Conditional regulation of HIF-1a, VEGF,
and VEGFR2 in inflammation and hypoxia
by rHDL
To mechanistically elaborate the multifunctional effects of HDL in angiogenesis, we examined the effects of rHDL on the regulation of the key
angiogenic transcription factor HIF-1a, the downstream angiogenic
growth factor VEGF and its receptor VEGFR2 under inflammatory
and hypoxic conditions. Incubation with rHDL alone caused significant
elevations in HIF-1a (50.5%), VEGF (56.3%), and VEGFR2 (35.5%),
compared with PBS controls (Figure 5A–C and F–H). However, rHDL
pre-incubation significantly suppressed TNF-a- and MCM-mediated inflammatory stimulation of HIF-1a (40.4 and 72.5%), VEGF (TNF-a:
41.0% and MCM: 30.7%), and VEGFR2 (TNF-a 33.2% and MCM: 69.0%)
protein expression relative to stimulated control (Figure 5A–C, see Supplementary material online, Figure S3A–C). In contrast, rHDL augmented
hypoxia-induced stimulation of HIF-1a (24.4%), VEGF (111.5%), and
VEGFR2 (13.2%) relative to hypoxic control (P , 0.05, Figure 5F–H ).
These data demonstrate that rHDL conditionally regulates HIF-1a,
VEGF, and VEGFR2, with suppression of these factors in inflammation
but enhancement in hypoxia, consistent with our in vivo observations.
150
H.C.G. Prosser et al.
Figure 4 rHDL augments hypoxia-stimulated HCAEC angiogenesis in vitro. HCAECs pre-incubated with rHDL (16 h, 600 mg/mL) or PBS were subjected to
hypoxia (1% O2/5% CO2). (A) Matrix gel tubulogenesis was assessed as total number of branches/well. Panels on left are representative tubule images for each
treatment condition. (B) Cellular proliferation was determined using the Click-It assay and (C) migration was determined using live cell imaging. Data are
presented as % of PBS control, n ¼ 4 and repeated four to six times. *P , 0.05, **P , 0.005 vs. PBS controls; ^P , 0.05, ^^P , 0.005 vs. hypoxia controls.
3.6 rHDL regulation of intracellular
pro-angiogenic factors in endothelial cells
NF-kB is a transcription factor that plays a key role in regulating
inflammation-induced angiogenesis.24 It comprises two subunits (p65 and
p50), which upon activation translocate to the nucleus and activate the expression of genes associated with inflammation, including the key
pro-angiogenic and inflammatory chemokine CCL2. rHDL significantly attenuated TNF-a- and MCM-induced elevations of nucleic p65 protein levels
(43.5% for TNF-a and 37.1% for MCM, P , 0.05, Figure 5D and see Supplementary material online, Figure S4A). rHDL also inhibited TNF-a- and
MCM-induced expression of CCL2 (32.5% for TNF-a and 30.0% for
MCM, P , 0.05, Figure 5E and see Supplementary material online, Figure S4B).
Similar to the inflammatory stimuli, hypoxia increased nucleic p65
protein levels (11.2%). This was attenuated in cells pre-incubated with
rHDL (29.7%, P , 0.005, Figure 5I), indicating that NF-kB is not conditionally regulated by rHDL. Pre-incubation with rHDL did, however, significantly up-regulate CXCL12 mRNA under both normoxic and
hypoxic conditions (Figure 5J).
3.7 Comparative assessment of rHDL,
native HDL, and apoA-I in modulating
tubulogenesis, HIF-1a, and VEGF protein
expression in hypoxia and inflammation
HCAECs pre-incubated with rHDL, native HDL (nHDL), or lipid-free
apoA-I were subjected to hypoxia, TNF-a, or MCM. With the exception
of apoA-I, which significantly inhibited tubule formation following
TNF-a stimulation (25.4%, P , 0.05), neither apoA-I nor nHDL had significant effects on tubule formation or on HIF-1a and VEGF protein expression (see Supplementary material online, Figure S5). Only rHDL
elicited consistent and significant conditional modulation of these angiogenic parameters in response to hypoxia and inflammation.
4. Discussion
In this study, we report that (i) infusion of apoA-I inhibits neovascularization in the murine peri-arterial femoral cuff model of inflammatory
angiogenesis; (ii) in contrast, apoA-I infusions augment ischaemiainduced neovascularization in the murine model of hindlimb ischaemia;
(iii) consistent with the multifunctional regulation of neovessel formation observed in vivo, rHDL inhibits cellular proliferation, migration,
and tubulogenesis in response to inflammation in vitro, but (iv) augments
these key angiogenic processes in hypoxic conditions. Finally, the in vivo
and functional in vitro findings are supported by mechanistic data demonstrating that when followed by an inflammatory stimulus, rHDL inhibits
the key pro-angiogenic transcription factor HIF-1a, the growth factor
VEGF, and its receptor VEGFR2. Conversely, under hypoxic conditions
rHDL augments HIF-1a, VEGF, and VEGFR2. In summary, we have
demonstrated that apoA-I/rHDL inhibits pathological, inflammatorydriven angiogenesis but also promotes physiological, ischaemia-driven
neovascularization.
These findings support epidemiological studies that report positive
correlations between elevated HDL levels and improved prognosis in
Multifunctional regulation of angiogenesis by HDL
151
Figure 5 Conditional regulation of intracellular pro-angiogenic factors by rHDL. Western blotting was used to determine HIF-1a (A and F ), VEGF (B and
G), VEGFR2 (C and H ), and nucleic p65 (D and I ) protein expression after pre-incubation of HCAECs with rHDL (16 h, 600 mg/mL) and subjected to TNF-a
(1 ng/mL) or hypoxia (1% O2/5% CO2), respectively. (E) CCL2 protein was determined by ELISA, and (J ) CXCL12 mRNA levels were determined by RT –
PCR (normalized to B2M). All protein was normalized to a-tubulin. All data are presented as % of PBS control, n ¼ 4 and repeated four to six times.
*P , 0.05, **P , 0.005 vs. PBS controls; ^P , 0.05, ^^P , 0.005 vs. TNF-a or hypoxia stimulation.
both inflammatory- and ischaemia-driven pathologies, in which angiogenesis plays a key role.25 Inflammation-driven plaque neovascularization accelerates late plaque growth and instability.26 – 28 Plaque
neovessels provide a pathway for the delivery of inflammatory cells,
growth factors, and cytokines. Furthermore, plaque neovessels are thinwalled and prone to haemorrhage, which causes lipid core expansion by
152
H.C.G. Prosser et al.
Figure 6 Multifunctional regulation of angiogenesis by HDL. In the setting of inflammation, HDL inhibits the activation of NF-kB, which subsequently
suppresses HIF-1a, VEGF, and CCL2, causing a reduction in key inflammatory angiogenic processes. Conversely, in settings of hypoxia HDL augments
HIF-1a, VEGF, and CXCL12, thereby promoting ischaemia-driven angiogenesis. The conditional regulation of HIF-1a, VEGF, and VEGFR2 may be
central to the multifunctional effects of HDL on angiogenesis.
accumulation of cholesterol from erythrocyte membranes, making the
plaque unstable.29 Infusions of HDL, apoA-I, or apoA-I mimetic peptides
reduce atherosclerotic plaque size and increase plaque stability in
apoE2/2 mice and humans,10,11,30 indicating that HDL may therefore
reduce inflammatory plaque neovascularization. In contrast to these inhibitory effects of HDL in inflammation, rHDL has been shown to increase
neovascularization in mouse models of ischaemia.17 Prospective studies
also report that elevated serum HDL is associated with improved survival and prognosis, following myocardial infarction.18,19 Combining
these previous reports with our current findings supports the hypothesis that HDL can modulate angiogenesis dependent upon the pathophysiological setting.
To investigate the effect of HDL on inflammatory-driven neovascularization in vivo, a non-occlusive peri-arterial collar was implanted around
the femoral artery of mice. Implantation of the collar generates an acute,
localized inflammatory response, characterized by the proliferation of
vascular smooth muscle cells, and the infiltration of leukocytes and
macrophages into the intima. Infiltrated cells secrete cytokines and
growth factors that stimulate endothelial cell migration, proliferation,
and subsequent neovascularisation,31,32 providing a physiological
model of vascular inflammation-induced angiogenesis. Consistent with
this, our study found an increased presence of neovessels and macrophages, and elevated CCL2 and TNF-a mRNA expression following
collar placement. Strikingly, i.v. infusion of apoA-I significantly inhibited
this inflammatory response, reducing neovessel formation, macrophage
infiltration, and inflammatory genes. Consistent with this, our in vitro
studies showed that rHDL abolished endothelial cell migration
towards stimulated macrophages in a trans-well assay, inhibited
HCAEC tubulogenesis, and reduced the expression of key proangiogenic proteins (HIF-1a, VEGF, VEGFR2, and CCL2) in response
to two inflammatory stimuli.
Endothelial cell migration and proliferation are central to angiogenesis
and enable the formation of new blood vessels.33 Prior reports have indicated that HDL can promote in vitro endothelial and progenitor cell proliferation and migration, and alter tubule formation.34 – 36 These findings
are consistent with the results from our in vitro control groups (i.e.
without stimulus) and provide additional support for a role for HDL in
the regulation of these key angiogenic processes. Moreover, we have
further characterized the regulation afforded by rHDL under the pathophysiologically relevant angiogenic conditions of inflammation and
hypoxia. Significantly, we find that rHDL attenuates endothelial cell migration, proliferation, and tubule formation in response to two inflammatory stimuli (TNF-a and MCM), despite rHDL treatment increasing
tubulogenesis, cellular proliferation, and migration when there was no
stimulus. Conversely, in the setting of hypoxia, rHDL augmented all
three pro-angiogenic cell functions, having an opposite action to that
in inflammation. Our intracellular mechanistic findings provide an explanation to this conditional regulation of angiogenesis by rHDL.
The intracellular mechanisms regulating angiogenesis are complex.
NF-kB, CCL2, HIF-1a, VEGF, and VEGFR2 have all been identified to
play critical roles in promoting inflammation-induced neovessel formation.37 – 40 Physiological, ischaemia-induced angiogenesis is also regulated by HIF-1a, VEGF, VEGFR2, as well as CXCL12.37 The
overlapping regulation by HIF-1a, VEGF, and VEGFR2 in both
153
Multifunctional regulation of angiogenesis by HDL
pathophysiological conditions may enable the multifunctional regulation
of angiogenesis by HDL. We show that rHDL causes differential regulation of HIF-1a, VEGF, and VEGFR2 rHDL attenuated HIF-1a, VEGF, and
VEGFR2 in inflammatory conditions but conversely augmented their expression in hypoxia (Figure 5). HIF-1a,VEGF, and VEGFR2 all have NF-kB
response elements upstream of their promoter region, and NF-kB is a
well-described pivotal regulator of inflammation-induced angiogenesis.40,41 The present study found that rHDL significantly attenuated
NF-kB expression in both inflammatory and hypoxic conditions. This
suggests that the observed attenuation of angiogenesis by rHDL in response to inflammation is likely via inhibition of NF-kB by rHDL,
which subsequently reduces HIF-1a, VEGF, and VEGFR2 expression
(Figure 5). rHDL also inhibited CCL2, a pro-angiogenic chemokine41
that is regulated by NF-kB and expressed at sites of inflammation, including
atherosclerotic plaques.42 Taken together, it is apparent that the antiinflammatory properties of HDL are critical in attenuating pathological
angiogenesis (Figure 6).
Under hypoxic conditions, the rHDL-induced augmentation of
HIF-1a, VEGF, and VEGFR2 protein was independent of NF-kB activity
(Figure 5I). The mechanisms underlying rHDL-induced expression of
HIF-1a, VEGF, and VEGFR2 remain to be defined; however, we hypothesize that rHDL may stabilize HIF-1a in hypoxia through inhibition
of upstream post-translational regulators, such as prolyl-hydroxylases,
which would in turn stimulate VEGF synthesis.43 The augmentation of
VEGFR2 expression by rHDL in hypoxia is not directly dependent
upon HIF-1a, as VEGFR2 possesses no HIF-1a response elements.39
While the regulation of VEGFR2 expression has not been characterized,
it is suggested to be predominantly post-transcriptionally regulated.44
Furthermore, VEGFR2 expression is increased in response to elevated
VEGF levels, in a paracrine/autocrine feed-forward loop.45 rHDL may
therefore regulate VEGFR2 through its modulation of VEGF. CXCL12
also has HIF-1a response elements upstream of its promoter
region,43 and plays a pivotal role in directing the recruitment and migration of endothelial progenitor cells.46 Our finding that rHDL significantly
elevated CXCL12 mRNA under both normoxic and hypoxic conditions,
as well as in the mouse ischaemic hindlimb, is consistent with studies that
show rHDL increases the number of circulating endothelial progenitor
cells, and promotes their homing to sites of ischaemic injury, key processes for neovascularization.17,36
The decision to use lipid-free apoA-I, the predominant apo lipoprotein constituent of HDL, for the in vivo studies, and rHDL for the
in vitro studies, was made due to previous observations that apoA-I is
rapidly lipidated and incorporated into HDL immediately after i.v. delivery.47 Thus, in order to determine the true mechanisms by which HDL
exerts its multifunctional effects on angiogenesis in vivo, it was pertinent
to use rHDL in all in vitro experiments. Moreover, in vitro studies determined rHDL to be most consistent in conditionally regulating key angiogenic parameters in hypoxia or inflammation compared with nHDL or
lipid-free apoA-I. This finding is consistent with previous studies reporting rHDL to be more effective at inhibiting VCAM-1 than lipid-free
apoA-I;48 and have greater anti-inflammatory actions compared with
nHDL,49 which is likely the result of reduced HDL heterogeneity.
Despite the vast number of reports demonstrating the therapeutic
benefits of HDL on the cardiovascular system, there is no translated
use of HDL-targeted treatments to date. The current study provides a
greater understanding into the vascular biological effects of HDL that
may ultimately facilitate their translation into therapeutic application,
for not only cardiovascular disease but also for diseases associated
with angiogenesis.
5. Conclusions
We report that apoA-I/rHDL regulates angiogenesis in a multifunctional
manner, depending on the pathophysiological context. This regulation is
characterized by inhibition of pathological, inflammatory-driven angiogenesis and, conversely, augmentation of physiological, hypoxia-mediated
neovascularization. This was observed in two relevant murine models of
angiogenesis and in in vitro angiogenic functional assays. The mechanism
underlying HDL’s context-dependent regulation of angiogenesis may be
via the differential regulation of HIF-1a and VEGF in conditions of inflammation or hypoxia. Together, our findings support the use of HDL as a
novel therapy for the treatment of angiogenic diseases.
Supplementary material
Supplementary material is available at Cardiovascular Research online.
Acknowledgements
We thank VisiScience for the illustrations.
Conflict of interest: none declared.
Funding
This work was supported by the National Health and Medical Research
Council (NHMRC) of Australia (project grant #632512 to M.N. and C.B.,
and Early Career Fellowship #537537 to L.D.) and the Heart Foundation
(Career Development Fellowship #CR07S3331 to C.B.).
References
1. Gordon DJ, Rifkind BM. High-density lipoprotein—the clinical implications of recent
studies. N Engl J Med 1989;321:1311 – 1316.
2. Tardif JC, Grégoire J, L’Allier PL, Ibrahim R, Lespérance J, Heinonen TM et al. Effects of
reconstituted high-density lipoprotein infusions on coronary atherosclerosis: a randomized controlled trial. JAMA 2007;297:1675 –1682.
3. Calabresi L, Gomaraschi M, Franceschini G. Endothelial protection by high-density lipoproteins. Arterioscler Thromb Vasc Biol 2003;23:1724 –1731.
4. Nicholls SJ, Dusting GJ, Cutri B, Bao S, Drummond GR, Rye KA et al. Reconstituted highdensity lipoproteins inhibit the acute pro-oxidant and proinflammatory vascular changes
induced by a periarterial collar in normocholesterolemic rabbits. Circulation 2005;111:
1543– 1550.
5. O’Connell BJ, Genest JJ. High-density lipoproteins and endothelial function. Circulation
2001;104:1978 –1983.
6. Prosser HC, Ng MK, Bursill CA. The role of cholesterol efflux in mechanisms of endothelial protection by HDL. Curr Opin Lipidol 2012;23:182–189.
7. Barter PJ, Rye KA. Cholesteryl ester transfer protein inhibition as a strategy to reduce
cardiovascular risk. J Lipid Res 2012;53:1755 –1766.
8. Lamon BD, Hajjar DP. Inflammation at the molecular interface of atherogenesis: an anthropological journey. Am J Pathol 2008;173:1253 –1264.
9. Moulton KS, Vakili K, Zurakowski D, Soliman M, Butterfield C, Sylvin E et al. Inhibition of
plaque neovascularization reduces macrophage accumulation and progression of
advanced atherosclerosis. Proc Natl Acad Sci USA 2003;100:4736 –4741.
10. Shaw JA, Bobik A, Murphy A, Kanellakis P, Blombery P, Mukhamedova N et al. Infusion of
reconstituted high-density lipoprotein leads to acute changes in human atherosclerotic
plaque. Circ Res 2008;103:1084 –1091.
11. Nissen SE, Tsunoda T, Tuzcu EM, Schoenhagen P, Cooper CJ, Yasin M et al. Effect of recombinant ApoA-I Milano on coronary atherosclerosis in patients with acute coronary
syndromes: a randomized controlled trial. JAMA 2003;290:2292 – 2300.
12. Choudhury RP, Rong JX, Trogan E, Elmalem VI, Dansky HM, Breslow JL et al. High-density
lipoproteins retard the progression of atherosclerosis and favorably remodel lesions
without suppressing indices of inflammation or oxidation. Arterioscler Thromb Vasc Biol
2004;24:1904 –1909.
13. Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. High density lipoprotein
as a protective factor against coronary heart disease. The Framingham Study. Am J Med
1977;62:707 –714.
14. Nicholls SJ, Cutri B, Worthley SG, Kee P, Rye KA, Bao S et al. Impact of short-term administration of high-density lipoproteins and atorvastatin on atherosclerosis in rabbits.
Arterioscler Thromb Vasc Biol 2005;25:2416 –2421.
15. Rong JX, Li J, Reis ED, Choudhury RP, Dansky HM, Elmalem VI et al. Elevating high-density
lipoprotein cholesterol in apolipoprotein E-deficient mice remodels advanced
154
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
atherosclerotic lesions by decreasing macrophage and increasing smooth muscle cell
content. Circulation 2001;104:2447 –2452.
Moulton KS, Heller E, Konerding MA, Flynn E, Palinski W, Folkman J. Angiogenesis inhibitors endostatin or TNP-470 reduce intimal neovascularization and plaque growth in
apolipoprotein E-deficient mice. Circulation 1999;99:1726 –1732.
Sumi M, Sata M, Miura SI, Rye KA, Toya N, Kanaoka Y et al. Reconstituted high-density
lipoprotein stimulates differentiation of endothelial progenitor cells and enhances
ischemia-induced angiogenesis. Arterioscler Thromb Vasc Biol 2007;27:813 –818.
Al-Khalili F, Svane B, Janszky I, Ryden L, Orth-Gomer K, Schenck-Gustafsson K. Significant predictors of poor prognosis in women aged ≤65 years hospitalized for an acute
coronary event. J Intern Med 2002;252:561 –569.
Berge KE, Canner PL, Hainline AJ. High-density lipoprotein cholesterol and prognosis
after myocardial infarction. Circulation 1982;66:1176 – 1178.
Weisweiler P. Isolation and quantitation of apolipoproteins A-I and A-II from human
high-density lipoproteins by fast-protein liquid chromatography. Clin Chim Acta 1987;
169:249 – 254.
Bursill CA, Castro ML, Beattie DT, Nakhla S, van der Vorst E, Heather AK et al. Highdensity lipoproteins suppress chemokines and chemokine receptors in vitro and in
vivo. Arterioscler Thromb Vasc Biol 2010;30:1773 –1778.
Hadfield KA, Pattison DI, Brown BE, Hou L, Rye KA, Davies MJ et al. Myeloperoxidasederived oxidants modify apolipoprotein A-I and generate dysfunctional high-density
lipoproteins: comparison of hypothiocyanous acid (HOSCN) with hypochlorous acid
(HOCl). Biochem J 2013;449:531 –542.
Limbourg A, Korff T, Napp LC, Schaper W, Drexler H, Limbourg FP. Evaluation of postnatal arteriogenesis and angiogenesis in a mouse model of hind-limb ischemia. Nat Protoc
2009;4:1737 –1746.
Costa C, Incio J, Soares R. Angiogenesis and chronic inflammation: cause or consequence? Angiogenesis 2007;10:149–166.
Carmeliet P. Angiogenesis in health and disease. Nat Med 2003;9:653 –660.
Jain RK, Finn AV, Kolodgie FD, Gold HK, Virmani R. Antiangiogenic therapy for normalization of atherosclerotic plaque vasculature: a potential strategy for plaque stabilization.
Nat Clin Pract Cardiovasc Med 2007;4:491 –502.
Virmani R, Kolodgie FD, Burke AP, Finn AV, Gold HK, Tulenko TN et al. Atherosclerotic
plaque progression and vulnerability to rupture: angiogenesis as a source of intraplaque
hemorrhage. Arterioscler Thromb Vasc Biol 2005;25:2054 –2061.
Sluimer JC, Daemen MJ. Novel concepts in atherogenesis: angiogenesis and hypoxia in
atherosclerosis. J Pathol 2009;218:7 –29.
Moreno PR, Purushothaman KR, Fuster V, Echeverri D, Truszczynska H, Sharma SK et al.
Plaque neovascularization is increased in ruptured atherosclerotic lesions of human
aorta: implications for plaque vulnerability. Circulation 2004;110:2032 –2038.
Ibanez B, Giannarelli C, Cimmino G, Santos-Gallego CG, Alique M, Pinero A et al. Recombinant HDL(Milano) exerts greater anti-inflammatory and plaque stabilizing properties than HDL(wild-type). Atherosclerosis 2011;220:72–77.
Bhardwaj S, Roy H, Babu M, Shibuya M, Yla-Herttuala S. Adventitial gene transfer of
VEGFR-2 specific VEGF-E chimera induces MCP-1 expression in vascular smooth
muscle cells and enhances neointimal formation. Atherosclerosis 2011;219:84 –91.
H.C.G. Prosser et al.
32. Donetti E, Baetta R, Comparato C, Altana C, Sartore S, Paoletti R et al. Polymorphonuclear leukocyte-myocyte interaction: an early event in collar-induced rabbit carotid
intimal thickening. Exp Cell Res 2002;274:197–206.
33. Wahlberg E. Angiogenesis and arteriogenesis in limb ischemia. J Vasc Surg 2003;38:
198 –203.
34. Miura SI, Fujino M, Matsuo Y, Kawamura A, Tanigawa H, Nishikawa H et al. High density
lipoprotein-induced angiogenesis requires the activation of Ras/MAP kinase in human
coronary artery endothelial cells. Arterioscler Thromb Vasc Biol 2003;23:802 –808.
35. Seetharam D, Mineo C, Gormley AK, Gibson LL, Vongpatanasin W, Chambliss KL et al.
High-density lipoprotein promotes endothelial cell migration and reendothelialization
via scavenger receptor-B type I. Circ Res 2006;98:63–72.
36. Zhang Q, Yin H, Liu P, Zhang H, She M. Essential role of HDL on endothelial progenitor
cell proliferation with PI3K/Akt/cyclin D1 as the signal pathway. Exp Biol Med (Maywood)
2010;235:1082 –1092.
37. Carmeliet P. Manipulating angiogenesis in medicine. J Intern Med 2004;255:538 – 561.
38. De Martin R, Hoeth M, Hofer-Warbinek R, Schmid JA. The transcription factor
NF-kappa B and the regulation of vascular cell function. Arterioscler Thromb Vasc Biol
2000;20:E83 –E88.
39. Hoeben A, Landuyt B, Highley MS, Wildiers H, Van Oosterom AT, De Bruijn EA. Vascular
endothelial growth factor and angiogenesis. Pharmacol Rev 2004;56:549–580.
40. Mazor R, Alsaigh T, Shaked H, Altshuler AE, Pocock ES, Kistler EB et al. Matrix
metalloproteinase-1-mediated up-regulation of vascular endothelial growth factor-2
in endothelial cells. J Biol Chem 2013;288:598 – 607.
41. Jackson JR, Seed MP, Kircher CH, Willoughby DA, Winkler JD. The codependence of
angiogenesis and chronic inflammation. Faseb J 1997;11:457 –465.
42. Shin WS, Szuba A, Rockson SG. The role of chemokines in human cardiovascular pathology: enhanced biological insights. Atherosclerosis 2002;160:91 –102.
43. Rey S, Semenza GL. Hypoxia-inducible factor-1-dependent mechanisms of vascularization and vascular remodelling. Cardiovasc Res 2010;86:236 –242.
44. Neufeld G, Cohen T, Gengrinovitch S, Poltorak Z. Vascular endothelial growth factor
(VEGF) and its receptors. FASEB J 1999;13:9 –22.
45. Waltenberger J, Mayr U, Pentz S, Hombach V. Functional upregulation of the vascular
endothelial growth factor receptor KDR by hypoxia. Circulation 1996;94:1647 –1654.
46. Hattori K, Heissig B, Tashiro K, Honjo T, Tateno M, Shieh JH et al. Plasma elevation of
stromal cell-derived factor-1 induces mobilization of mature and immature hematopoietic progenitor and stem cells. Blood 2001;97:3354 –3360.
47. Kee P, Rye KA, Taylor JL, Barrett PH, Barter PJ. Metabolism of apoA-I as lipid-free protein
or as component of discoidal and spherical reconstituted HDLs: studies in wild-type and
hepatic lipase transgenic rabbits. Arterioscler Thromb Vasc Biol 2002;22:1912 –1917.
48. Baker PW, Rye KA, Gamble JR, Vadas MA, Barter PJ. Ability of reconstituted high density
lipoproteins to inhibit cytokine-induced expression of vascular cell adhesion molecule-1
in human umbilical vein endothelial cells. J Lipid Res 1999;40:345–353.
49. van der Vorst EP, Vanags LZ, Dunn LL, Prosser HC, Rye KA, Bursill CA. High-density lipoproteins suppress chemokine expression and proliferation in human vascular smooth
muscle cells. FASEB J 2013;27:1413 –1425.