Content uploaded by Brian J. Van Lenten
Author content
All content in this area was uploaded by Brian J. Van Lenten on May 03, 2015
Content may be subject to copyright.
222
|
APRIL 2011
|
VOLUME 8 www.nature.com/nrcardio
Department of
Medicine, David Geffen
School of Medicine,
UCLA, 10833 Le Conte
Avenue, Los Angeles,
CA 90095-1679, USA
(M. Navab, S.T. Red dy,
B.J. Van Lenten,
A.M.Fogelman).
Correspondence to:
M. Navab
mnavab@
mednet.ucla.edu
HDL and cardiovascular disease: atherogenic
and atheroprotective mechanisms
Mohamad Navab, Srinivasa T. Reddy, Brian J. Van Lenten and Alan M. Fogelman
Abstract | The lipoprotein HDL has two important roles: first, it promotes reverse cholesterol transport, and
second, it modulates inflammation. Epidemiological studies show that HDL-cholesterol levels are inversely
correlated with the risk of cardiovascular events. However, many patients who experience a clinical event have
normal, or even high, levels of HDL cholesterol. Measuring HDL-cholesterol levels provides information about
the size of the HDL pool, but does not predict HDL composition or function. The main component of HDL,
apolipoproteinA-I (apo A-I), is largely responsible for reverse cholesterol transport through the macrophage
ATP -binding cassette transporter ABCA1. Apo A-I can be damaged by oxidative mechanisms, which render
the protein less able to promote cholesterol efflux. HDL also contains a number of other proteins that are
affected by the oxidative environment of the acute-phase response. Modification of the protein components
of HDL can convert it from an anti-inflammatory to a proinflammatory particle. Small peptides that mimic
some of the properties of apo A-I have been shown in preclinical models to improve HDL function and reduce
atherosclerosis without altering HDL-cholesterol levels. Robust assays to evaluate the function of HDL are
needed to supplement the measurement of HDL-cholesterol levels in the clinic.
Navab, M. etal. Nat. Rev. Cardiol. 8, 22 2–232 (2011); publi shed onl ine 8 Feb ru ary 2 011; doi:10.1038/nrcardio.2010.222
Introduction
Epidemiological studies from the past five decades have
consistently shown that levels of HDL cholesterol are
inversely correlated with clinical events resulting from
atherosclerosis, whereas LDL-cholesterol levels are
directly related to the rate at which these events occur.1–5
HDL protects against cardiovascular disease (CVD) by
regulating cholesterol efflux from tissues and modula-
ting inflammation. Other properties include its anti-
oxidant and vasoprotective effects (Figure1). However,
clinical events often manifest in patients with normal
HDL-cholesterol levels in the range of 40–50 mg/dl.
In an analysis of data from the original Framingham
Heart Study,1 our research group calculated that 44% of
clinical events in men occurred in those with ≥40 mg/dl
of HDL cholesterol, and 43% of events in wome n
occurred in those with ≥50 mg/dl of HDL cholesterol.6
Therefore, although clinical events were associated
with low HDL-cholesterol levels when considering the
study population as a whole, nearly half of the men and
women who suffered a clinical event had normal HDL-
cholesterol levels. This finding suggests that measur-
ing HDL-cholesterol levels is not the most-appropriate
method for determining risk of cardiovascular disease
(CVD). Alternative means for determining CVD risk
have been tested, such as the ratio of total cholesterol to
HDL cholesterol, which has been shown to be 40% more
informative than non-HDL-cholesterol levels (total
cholesterol minus HDL) at predicting mortality from
ischemic heart disease, and more than twice as informa-
tive as total cholesterol levels.2 Of note, the investigators
of a large meta-analysis found that lipid assessment in
patients with CVD can be simplified by measuring levels
of total cholesterol and HDL cholesterol or apolipo-
proteins without need of fasting, and without having to
determine triglyceride levels.3 Data from another large
meta-analysis indicated that simply increasing HDL-
cholesterol levels will not reduce the risk of events or
deaths related to coronary heart disease (CHD), or
all-cause deaths, and the investigators suggested that
the primary go al of lipid-modifying interventions
should be to decrease levels of LDL.4 After adjustment
for apolipo proteinA-I (apo A-I) and apolipoproteinB
(apo B) levels, very high levels of HDL choles terol and
large HDL-particle size were associated with increased
risk of coronary artery disease (CAD), whereas very
high levels of apo A-I were not associated with CAD.5
These findings could have important implications for
the assessment and treatment of risk associated with
CAD, indicating that high HDL-cholesterol levels alone
should not be the determining factor in designing a
treatment strategy in these patients. In this Review, we
focus on the mechanisms by which HDL protects against
CVD and on the relationship between impaired HDL
function andCVD. We also discuss approaches to assess
and modul ate HDL function that could potentially be
applied in the clinicalsetting.
Competing interests
M. Navab, S.T. R e d d y, a n d A . M. Fogelman declare associations
with the following company: Bruin Pharmaceuticals. See the
article online for full details of the relationships. B.J. VanLenten
declares no competing interests.
REVIEWS
© 2011 Macmillan Publishers Limited. All rights reserved
NATURE REVIEWS
|
CARDIOLOGY VOLUME 8
|
APRIL 2011
|
223
HDL structure
Lipoproteins consist of an outer layer of phospholipids
and apolipoproteins, and an inner core that carries lipids.
Normal HDL also contains high levels of antioxidant
molecules, such as paraoxonase/arylesterase1 (PON 1)
and lec ithin–ch olesterol acyltransferas e (LC AT).
Approximately 40–60% of HDL is composed of lipids,
such as cholesterol, cholesteryl esters, phospholipids, and
triglycerides. HDL proteins include apo A-I (approxi-
mately 70%), apo A-II (20%), and other apolipoproteins
and enzymes (10%). Apo B is not present in HDL; by
contrast, the only apolipoprotein associated with LDL is
apo B. One apo B molecule is associated with each LDL
particle and remains integrated with LDL as this lipo-
protein travels in the circulation and enters tissues. By
contrast, apo A-I can leave the HDL particle and interact
with tissues on its own, mainly through the ATP-binding
cassette (ABC) transporter ABCA1. This transporter
transfers lipids to lipid-poor apo A-I, which results in
the formation of new HDL particles by the liver and
intestine.7 Indeed, all the apolipoproteins associated with
HDL are exchangeable, meaning that they are constantly
being transferred between lipoprotein particles in the
circulation, and distinct HDL particles might only exist
for a few seconds.8 Almost three decades ago, Eisenberg
reflected that HDL should not be considered only as a
vehicle for lipid transport, as this lipoprotein has many
diversefunctions.9
HDL and cholesterol efflux
Regulation of cholesterol levels
Cholesterol is a fairly inert molecule with an important
role in membrane structure. Dietary cholesterol or chol-
es terol synthesized denovo by the liver or intestine can
be removed from the body by three methods. First, the
liver can convert cholesterol to bile acid, which, together
with phospholipids, is excreted into the bile. Bile is then
either reabsorbed or excreted in feces.10 Second, a very
small amount of cholesterol is converted to steroid hor-
mones.11 Third, cholesterol can also be removed from
the body through the shedding of cells from the skin and
intestine.12 Levels of free cholesterol are maintained in a
very narrow range because excess amounts are harmful to
normal membrane function.13 Any excess free chol esterol
is esterified and stored in lipid droplets to prevent it from
accumulating in cell membranes.13 The chol esterol syn-
thesis pathway and the LDL-receptor pathway regulate the
levels of cholesterol in cells, and both of these pathways
are exquisitely sensitive to the presence of chol esterol.
One of the main functions of HDL is to promotechol-
esterol efflux from peripheral tissues and transport
chol esterol to the liver. Lipid homeostasis is maintained
via multiple downstream feedback responses by adjust-
ing the rate of cholesterol biosynthesis, esteri fication,
ingestion, and export (Figure2).13
Macrophages have a number of receptors that recog-
nize modified LDL produced in atherosclerotic lesions,
and uptake of this modified LDL leads to the forma-
tion of macrophage foam cells, which are a hallmark
of athero sclerosis.14 HDL can remove cholesterol from
Key points
■HDL-cholesterol levels are inversely correlated with the risk of clinical events
resulting from atherosclerosis, although many patients who experience
cardiovascular events have normal, or even high, levels of HDL cholesterol
■HDL-cholesterol levels provide a measure of the size of the HDL pool, but do
not predict the composition or function of HDL
■A major function of HDL is to promote cholesterol efflux and reverse cholesterol
transport, which are mediated by interaction with the macrophage ATP-binding
cassette transporters ABCA1 and ABCG1
■The main protein component of HDL, apolipoproteinA-I, can be damaged by
oxidative mechanisms that render it less able to promote cholesterol efflux
■HDL contains several other proteins that are affected by the oxidative
environment of the acute-phase response, causing HDL to change from an
anti-inflammatory particle to a proinflammatory particle
■Peptides that mimic the lipid-binding properties of apolipoproteinA-I have
been shown in preclinical models to improve HDL function and reduce
atherosclerosis without altering HDL-cholesterol levels
macro phages, which is thought to be the mechanism
by which this lipoprotein is protective against athero-
sclerosis. The major mediators of cholesterol trans-
port from macro phages invivo seem to be ABCA1 and
ABCG1, which act synergistically.15,16 Deletion of both
of these transporters in a mouse model produced more-
severe athero sclerosis than deletion of either one alone.17
For this reason, serum samples with comparable levels
of HDL cholesterol (after removal of apo B-containing
lipoproteins) could be expected to promote similar rates
of cholesterol efflux from macrophages. However, de la
Llera-Moya etal. showed that the best predictor of chol-
esterol efflux from macrophages was not the amount of
HDL cholesterol in serum samples, but rather the level
of a small subfraction of HDL known as preβ-1 HDL.18
Preβ-1 HDL interacts with the ABCA1 transporter and
removes cholesterol and phospholipids from cells.15,16,19,20
Although preβ-1-HDL levels were shown to be the stron-
gest predictor ofthe ability of serum to promote ABCA1-
mediated efflux, thecorrelation coefficient indicated that
preβ-1-HDL levels only accounted for 43% of the vari-
ability in chol esterol efflux, indicating that other factors
must also have an important role.18
Figure 1 | Functions and properties of HDL. Important
functions of HDL include reverse cholesterol transport
and cholesterol efflux, and its role in anti-inflammatory
processes. Abbreviations: LPS, lipopolysaccharide;
NO,nitric oxide.
Antithrombotic Anti-inammatory
Antibrotic HDL Antioxidant
Vasoprotective Increases NO
production
Promotes cholesterol efux
Protects against LPS
FOCUS ON LIPIDS AND CVD RISK
© 2011 Macmillan Publishers Limited. All rights reserved
224
|
APRIL 2011
|
VOLUME 8 www.nature.com/nrcardio
As the amount of cholesterol and phospholipids associ-
ated with apo A-I increases, the HDL particle changes
from a disc-like structure to become more spherical.
LCAT, which is activated by apo A-I, esterifies free chol-
esterol to cholesteryl esters, which are then stored in the
growing HDL particle.21,22 The HDL particles continue
to grow in size by interacting with ABCG1 on macro-
phages. ABCG1 only delivers cholesterol to lipid-rich
HDL parti cles, in contrast to ABCA1, which delivers
both chol esterol and phospholipids to lipid-poor preβ-1
HDL.15,16,23,24 The lipid-rich HDL delivers esterified
choles terol to the liver via receptors, such as the scaven-
ger receptor classB member1 (SRB1).25 Although SRB1
is important in removing esterified cholesterol from the
circulation and transporting it to the liver, it is not critical
for exporting cholesterol from macrophages.26
As HDL particles enlarge and mature in the circulation,
some apo A-I dissociates from these particles, creating
new lipid-poor apo A-I particles. These particles restart
the cycle of HDL formation, in which cholesterol is
transported from peripheral cells, such as macrophages,
to the liver for excretion in the bile or repackaging into
lipoproteins that are then released into the circulation.27
Although studies of HDL biogenesis have usually focused
on the role of ABCA1 in the liver and intestine, HDL can
also be generated postprandially when triglyceride-rich
lipoproteins undergo core lipid hydrolysis, generating
surface phospholipids.16
Oxidative stress and cholesterol efflux
Although preβ-1-HDL levels could be expected to be
inversely correlated with the risk of clinical events result-
ing from atherosclerosis, this is not the case. In one study,
preβ-1-HDL levels were actually higher in patients with
ischemic heart disease than in normal individuals, and
the activity of LCAT in these patients was diminished.28–30
One explanation for this observation, which has not been
directly tested, is that the preβ-1 HDL from the patients
included in this study was dysfunctional. Two groups
independently demonstrated that oxidative modifi-
cation of apo A-I, particularly by chlorination through
the myeloperoxidase pathway, dramatically reduced the
ability of apo A-I to promote cholesterol efflux through
the ABCA1 pathway.31,32 Both groups also demonstrated
that apo A-I in patients with CVD was oxidatively modi-
fied and defective in promoting cholesterol efflux from
macrophages via ABCA1.29,31 Furthermore, the LCAT
binding site on apo A-I was shown to be preferentially
Figure 2 | Lipid biosynthesis, storage, and elimination. HDL has a central role in the reverse cholesterol transpor t pathway.
Lipid-poor apo A-I is secreted by the liver and rapidly acquires cholesterol via the hepatocyte ABCA1 transporter and
promotes cholesterol efflux from macrophages. Free cholesterol is esterified to cholesteryl esters by LCAT to form mature
HDL, which transfers its cholesterol to apo B-containing lipoproteins, such as VLDL and LDL, via CETP-mediated transfer.
This cholesterol is subsequently taken up by the liver via the LDL receptor. PLTP transfers phospholipids from triglyceride-
rich lipoproteins to HDL, which promotes HDL remodeling. Hepatic cholesterol can be excreted into the bile after
conversion to bile acid or expelled directly into the bile as cholesterol. Bile and its components are either reabsorbed by
the intestine or ultimately excreted in feces. HDL can be remodeled by lipases, such as endothelial lipase and hepatic
lipase, which hydrolyze HDL phospholipids and HDL triglycerides, respectively. The kidneys are an important site of apo A-I
catabolism. Abbreviations: ABC, ATP-binding cassette; apo A-I, apolipoproteinA-I; CETP, cholester yl ester transport protein;
LCAT, lecithin–cholesterol acyltransferase; LDLR, LDL receptor; PLTP, phospholipid transfer protein; SRB1, scavenger
receptor classB member1.
Free cholesterol
Cholesteryl ester
Phospholipid
PLTP
CETP
LCAT
VLDL, LDL
ABCG1
ABCA1
Macrophage
SRB1
LDLR
ABCA1
ABCG1
HDL,
triglycerides,
cholesteryl
ester
HDL (discoidal)
endothelial lipase/
hepatic lipase
Apo
A-I
Bile
HDL (discoidal)
Kidneys
Intestine
Liver
REVIEWS
© 2011 Macmillan Publishers Limited. All rights reserved
NATURE REVIEWS
|
CARDIOLOGY VOLUME 8
|
APRIL 2011
|
225
targeted by oxidative modification, which resulted in
diminished LCAT activity.33,34
Chlorination, but not nitration, by myeloperoxidase
impaired the ability of apo A-I to interact directly with
ABCA1 and to activate the Janus-kinase-2 signaling
pathway, which prevented apo A-I from promoting cel-
lular cholesterol efflux in a cell-based model.35 ABCA1
is normally degraded in the absence of apolipoproteins
and, interestingly, this modification of apo A-I had little
effect on the ability of this apolipoprotein to stabilize the
ABCA1 protein or to solubilize phospholipids.35
HDL contains higher levels of lipid-oxidation products
than LDL.36 HDL might acquire lipid-oxidation prod-
uctsat sites of inflammation and transport them back
to the liver for disposal.37 This putative function of HDL
could protect endothelial cells and macrophages from
toxic lipid- oxidation products much as is thought to be
the case for reverse cholesterol transport.37 Therefore, the
increased levels of preβ-1 HDL in patients with CVD
that are associated with decreased LCAT activity could
be due to oxidative modification of their apo A-I at sites
of inflammation in the arteries. The increased levels of
preβ-1 HDL observed in patients with CVD,28 however,
are difficult to reconcile with the work of Cavigiolio and
colleagues, who showed that oxidative modification
ofapo A-I by myeloperoxidase limits the dissociationof
apo A-I from HDL, and, therefore, prevents it from
becoming a lipid-poor particle.38 More work is needed to
address this ap parent paradox.
On the basis of invitro studies, lysines in LDL were
proposed to be modified by malondialdehyde produced
by lipid peroxidation at sites of arterial injury.39 Shao
etal. reported that treatment of HDL invitro with mal on-
dialdehyde modified lysines in apo A-I and, when added
to cells, it considerably reduced the ability of apo A-I to
promote cholesterol efflux by the ABCA1 pathway.40
Moreover, HDL isolated from human atherosclerotic
lesions had elevated levels of malondialdehyde–protein
adducts, indicating that lipid peroxidation might render
HDL dysfunctional invivo.40
A study in which preβ-1-HDL levels predicted the
ability of a serum sample to promote ABCA1-mediated
cholesterol efflux better than did HDL-cholesterol levels,
was carried out using samples from normal indivi duals
who would be expected to have minimal oxidative
modifi cation of apo A-I.18 By contrast, patients with CHD
have substantial oxidative modification of their apo A-I,
and the high levels of preβ-1 HDL in these patients
might, therefore, not be protective. Simple measure-
ments of preβ-1 HDL levels are not likely to be useful in
understanding relative risk without the knowledge that
apo A-I in these particles is normal.
Relationship with LDL levels
Mice that were genetically modified to lack the two
major proteins that mediate cholesterol efflux, ABCA1
and ABCG1, had very low levels of HDL cholesterol
(<10 mg/dl), which was in contrast to normal mice
that had levels of about 60 mg/dl.17 These genetically
modified mice developed foam cells in almost all tissues
containing macrophages, but not in their arteries. The
reason for this observation was that these mice also had
low levels of LDL cholesterol. The investigators sug-
gested that these low levels of LDL were a consequence
of immuno modulation result ing from the double
knockout of the ABC transporters.17 Binding of LDL to
the subendo thelial matrix in arteries is critical for the
development of athero sclerosis. Accumulation of LDL in
the subendo thelial space and oxidative modification of
this lipo protein contribute to the inflammatory reaction,
which results in monocyte transmigration into the artery
wall and further lipid accumu lation.41 In the absence of
LDL or other cholesterol- rich lipoproteins containing
apo B, macro phages do not accumu late in arteries, and
athero sclerosis does not develop, even in the presence
of a severe defect in cholesterol efflux.7 This result high-
lights that, in the presence of high levels of LDL, protec-
tive HDL is neces sary to prevent inflammation, whereas
the role of HDL becomes less critical when low levels of
LDL are present.
Genetic determinants of cholesterol efflux
Allelic variations in ABCA1, APOA1, and LCAT have
been identified as causes of low HDL-cholesterol levels.42
Apo A-IMilano is a mutant that results in a genetic condi-
tion associated with very low levels of apo A-I and HDL
cholesterol, and high triglyceride levels, but without
increased risk of clinical events resulting from athero-
sclerosis.43 Studies in animals indicated that the muta-
tion in apo A-IMilano might increase the ability of this
lipo protein to promote reverse cholesterol transport.44
However, a direct comparison of the ability of normal
apo A-I and apo A-IMilano to promote reverse cholesterol
transport in mice failed to demonstrate any superiority
of the mutant protein.45
Mutations in two genes have been found to increase
levels of HDL cholesterol: the gene encoding chol esteryl
ester transferase protein (CETP),46 and the gene encod-
ing endothelial lipase.47 CETP is responsible for the
transfer of phospholipids and cholesteryl esters from
HDL to VLDL,46 which results in reductions in the chol-
esterol content of HDL. Inhibiting CETP could lead to a
beneficial increase in HDL-cholesterol levels. However,
increased levels of CETP were inversely associated with
the risk of coronary outcomes in patients taking statins,
particularly in those with low LDL-cholesterol levels.48
CETP might facilitate reverse cholesterol transport in
the setting of robust LDL clearance, making inhibition
of CETP in patients overexpressing this protein problem-
atic.48 The results of ongoing clinical trials will determine
whether the increase in HDL-cholesterol levels caused by
CETP inhibition reduces the rate of coronary outcomes.
Endothelial lipase is an HDL candidate gene, in which
loss-of-function mutations can result in elevated concen-
trations of HDL cholesterol. This lipase is a member of
the triglyceride lipase family of proteins, which includes
hepatic lipase and lipoprotein lipase.47
Castrillo etal. demonstrated that bacterial and viral
infections inhibited gene expression dependent on
the liverX receptor, through activation of the toll-like
FOCUS ON LIPIDS AND CVD RISK
© 2011 Macmillan Publishers Limited. All rights reserved
226
|
APRIL 2011
|
VOLUME 8 www.nature.com/nrcardio
receptors3 and 4 mediated by the interferon regulatory
factor3.49 One of the target genes of the liver-X- receptor
signaling pathway is ABCA1; in this way, infectious
agents might contribute to atherosclerosis by inhibit-
ing cholesterol efflux from macrophages.49 As discussed
above, SRB1 promotes reverse cholesterol transport by
transporting HDL-derived cholesteryl esters into the
liver, where they can be hydrolyzed, and transformed into
free cholesterol that is excreted as bile acids or directly
into the bile. In the setting of hyperlipidemia, cholesteryl-
ester-rich HDL accumulates in SRB1-knockout mice,
which develop severe atherosclerosis despite elevated
levels of HDL cholesterol.50 Van Eck etal. demonstrated
that loss of SRB1 in mice results in dysfunctional HDL,
as indicated by decreased activity of PON 1.51 In this
study, mice without SRB1 were under considerable oxi-
dative stress, which was made worse by feeding them a
Western-type diet.51
HDL and inflammation
Anti-inflammatory properties of HDL
Normal HDL is an effective antioxidant and anti-
inflammatory molecule (Figure3). Proteins and lipids
associated with HDL contribute to its anti-inflammatory
capacity. For example, PON 1 is capable of preventing
the formation of lipid oxidation products that result in
induction of proinflammatory molecules, such as inter-
leukin8.52 Apo A-I by itself reduces the pro inflammatory
effect of oxidized lipids.53 HDL lipids are also effective
in sequestering lipid oxidation products and reduc-
ing their proinflammatory capacity.53 A subfraction of
HDL, HDL3, has been shown to protect LDL from oxi-
dative damage by free radicals generated in the arterial
intima. The HDL3 particle is dense and protein-rich,
and is capable of picking up lipids from peripheral
cells.54 Rye etal. demonstrated that, following treat-
ment of cultured endothelial cells with tumor necrosis
factor (TNF), the expression of vascular cell adhesion
molecule1 (VCAM-1) and intercellular adhesion mol-
ecule1 (ICAM-1) was markedly increased.55 Treatment
of the cells with HDL from healthy individuals substan-
tially attenuated the expression of these proinflammatory
molecules, which are involved in leukocyte–endothelial
interactions.
In an attempt to mimic the conditions of the human
aortic wall, our group cultured aortic endothelial cells
on a multilayer of aortic smooth muscle cells isolated
from the same individual.56 LDL added to this coculture
became trapped in the space between the endothelial cells
and the smooth muscle cells, and underwent oxidative
modification. As a result, both cell types secreted mono-
cyte chemoattractant protein1 (MCP-1), which induced
monocytes to migrate between the endothelial cells and
smooth muscle cells. Addition of HDL to the coculture
together with LDL resulted in a dramatic decrease in the
production of MCP-1.56
Proinflammatory properties of HDL
During inflammation, HDL proteins and lipids become
oxidatively and enzymatically modified, and HDL loses
its protective capacity.57 High levels of dysfunctional
HDL are in fact associated with an increased risk of
CAD. Corsetti and colleagues reported that elevated
HDL levels were a risk factor for recurrent coronary
events in a subgroup of patients who had experienced
myocardial infarction and had hypercholesterolemia
and inflammation, as measured by levels of C-reactive
protein.58 Follow-up studies of this cohort confirmed
that patients with high levels of HDL cholesterol and
C-reactive protein were at increased risk of incident
CVD.59,60 As a result of myeloperoxidase binding to HDL
at sites of inflammation, HDL is not only rendered less
able to promote cholesterol efflux in this setting, but is
converted to a proinflammatory particle.61 In addition,
inflammation induced by endotoxemia has been shown
to impair reverse cholesterol transport without alter-
ing HDL-cholesterol levels.62 HDL derived from rabbits
during the acute-phase response induced by croton oil
injection,52 from humans 3days after elective surgery,63
or from mice during an acute influenzaA infection64 has
been also shown to lose its anti-inflammatory properties,
independent of HDL-cholesterol levels.64
Mice deficient in apolipoproteinE (apo E) are geneti-
cally susceptible to atherosclerosis, and feeding these
mice an atherogenic diet induced the acute-phase
response, whereas mice with normal apo E levels were
unaffected.64 Even on a standard chow diet, the apo-
E-knockout mice had a chronic acute-phase response,
and their HDL was dysfunctional in its ability to inhibit
LDL oxidation and prevent MCP-1 production in cul-
tures of human artery wall cells.65 Normal HDL and its
components are able to remove or inhibit the activity
of lipids in LDL that are required for LDL oxidation.57
By contrast, HDL from patients with CAD is often dys-
functional in its ability to inhibit LDL oxidation and
Figure 3 | Anti-inflammatory properties of HDL. The anti-inflammatory properties of
HDL are exer ted through various pathways. In endothelial cells, HDL downregulates
the cell adhesion molecules VCAM-1 and ICAM-1, and prevents the expression of
IL-8 and MCP-1 under inflammatory conditions. It also upregulates endothelial NO
induction and availability, protects against the proinflammatory effects of
lipopolysaccharide, and supports the differentiation and performance of dendritic
cells. In several settings, HDL also prevents the activity of NADPH oxidase and
regulates production of TNF by macrophages. Abbreviations: EC, endothelial cell;
ICAM-1, intercellular adhesion molecule1; IL, interleukin; MCP-1, monocyte
chemoattractant protein1; NO, nitric oxide; PON 1, paraoxonase/arylesterase1;
TNF, tumor necrosis factor; VCAM-1, vascular cell adhesion molecule1.
Regulates macrophage TNF Prevents expression
of EC IL-8, MCP-1
Inhibits elevation of
NADPH oxidase activity Anti-inammatory HDL Interacts with
antioxidant PON 1
Preserves dendritic cell
differentiation and function Upregulates EC
NO production
Downregulates
EC VCAM-1, ICAM-1
Protects against endotoxins
REVIEWS
© 2011 Macmillan Publishers Limited. All rights reserved
NATURE REVIEWS
|
CARDIOLOGY VOLUME 8
|
APRIL 2011
|
227
MCP-1 production by human artery wall cells, and in
preventing the oxidation of phospholipids by fatty-acid
hydroperoxides.53,66 Interestingly, LCAT activity attenu-
ated the inverse relation ship between C-reactive protein
and HDL cholesterol in one study, indicating that LCAT
could mitigate the anti-inflammatory and antioxidative
properties of HDL.67
Patel etal. infused 15 mg/kg/h Intralipid® 20% fat
emulsion (Fresenius Kabi, Runcorn, UK) or saline into
eight fasting healthy human males and then isolated
their HDL to determine its ability to inhibit VCAM-1
and ICAM-1 expression in human coronary artery endo-
thelial cells stimulated with TNF.68 The investigators
found that the infusion of Intralipid® resulted in signifi-
cantly increased triglyceride-rich HDL (P <0.05) that had
impaired anti-inflammatory activity compared with HDL
isolated fromcontrols.
Proteins associated with HDL
In addition to apo A-I, a number of other proteins are
associated with HDL that confer additional properties
to this lipoprotein. With the use of mass-spectrometry-
based proteomics, Vaisar and colleagues identified 48
different proteins in HDL isolated from healthy indivi-
duals and patients with CHD.69 Proteins involved in the
acute-phase response, including complement factors,
were found to be associated with HDL, as well as proteins
involved in lipid metabolism.69,70 Although hemopexin, a
protein involved in iron transport, was identified in this
study,69 haptoglobin, which binds hemoglobin released
from red blood cells, was not found to be associated with
HDL. Other groups have, however, reported that hapto-
globin is associated with HDL and binds to the LCAT
binding site on apo A-I.71 Watanabe etal. reported that
hemoglobin was associated with apo A-I in HDL from
mice susceptible to atherosclerosis when they were fed an
atherogenic diet, but not when they were fed a standard
chow diet.72 The hemoglobin associated with HDLin
mice receiving the atherogenic diet was predominantly
in an oxidized state and led to the reduction of nitric
oxide levels and to contracted arterial vessels exvivo.
HDL from healthy individuals had low levels of hemo-
globin associated with its apo A-I, whereas HDL from
patients with CHD had significantly increased levels
of hemo globin (P <0.02). The HDL from both groups
was tested in a cell-based assay for its ability to inhibit
the production of MCP-1 by human aortic endo thelial
cells. HDL from plasma of healthy individuals was
anti-inflammatory regardless of LDL-cholesterol con-
centration. By contrast, HDL from plasma of patients
with CHD was proinflammatory regardless of the LDL-
cholesterol concentration in the plasma.72 In a subse-
quent study, Watanabe etal. reported that increased
hemoglobin levels associated with HDL in mice receiv-
ing an atherogenic diet and in humans with CHD corre-
lated with increased haptoglobin and hemopexin levels.63
The content of lipid hydroperoxides in HDL and HDL-
inflammatory index values (a measure of the inflamma-
tory activity of HDL) significantly corre lated with the
content of hemoglobin, haptoglobin, and hemopexin in
HDL (P <0.02, P <0.01, P <0.02, respectively). In the mice
analyzed in this study, HDL-associated haptoglobin was a
critical determinant of the ability of hemoglobin to bind
to HDL. Furthermore, the associ ation of hemoglobin,
haptoglobin, and hemopexin with HDL positively corre-
lated with inflammatory properties of HDL and systemic
inflammation in patients with CHD in this study. Of note,
HDL from hemopexin-knockout mice under athero-
genic conditions does not accumulate hemoglobin and
is anti-inflammatory, which indicates that hemopexin is
required for the association of hemoglobin with HDL
and that hemoglobin–hemopexin complexes regulate
the inflammatory properties of HDL.63 Interestingly,
treatment of apo-E-knockout mice with an apo A-I
mimetic peptide resulted in dissoci ation of hemoglobin–
hemopexin complexes from HDL and improvement of
HDL inflammatory properties. Watanabe and colleagues
have, therefore, suggested that HDL can become pro-
inflammatory via this pathway involving association of
hemoglobin–hemopexin complexes with apo A-I, and
dissociation of these complexes from apo A-I-containing
particles of HDL might be a novel target for the treatment
of CHD.63
Mice and patients with diabetes mellitus and the hapto-
globin2-2 genotype have been shown to have increased
hemoglobin associated with HDL.73–75 HDL from these
patients also contained increased levels of lipid hydro-
peroxides and was dysfunctional in its ability to promote
cholesterol efflux. Administration of vitaminE to indivi-
duals with diabetes and the haptoglobin2-2 genotype
signifi cantly improved HDL function (P <0.001), but
had no effect in individuals with the haptoglobin1-1
genotype. Interestingly, vitaminC blocked the ability of
vitaminE to improve HDL-mediated cholesterol efflux
from macro phages in mice with diabetes and with the
hapto globin2-2 genotype.75 No significant differences
were observed in the cholesterol efflux elicited by the
serum of Hp 1-1mice with diabetes supplemented
with vitaminC, α-tocopherol, or their combination,
as compared with placebo, the mechanism of which is
beingstudied.
Another protein that has been implicated in the anti-
inflammatory properties of HDL is the anti oxidant
enzyme PON 1.53 The importance of this enzyme in
atherosclerosis was demonstrated by Mackness and
Mackness, who showed that oxidation of LDL is prevented
by PON 1.76 The enzyme is largely synthesized in the liver
and transferred to HDL by an SRB1-mediated mecha-
nism.77 Bhattacharyya etal. reported the associ ation of
PON1 polymorphisms with PON 1 activity and systemic
oxidative stress.78 Serum PON 1 activityin patients with
the QQ192 genotype was higher than inthose with the
QR192 genotype, who in turn had higher PON 1 activity
than patients with the RR192 genotype. The incidence of
cardiac events was lower in patients in the highest PON 1
activity quartile than in those in the lowest quartile.
Moreover, a significant correlation was found between
PON 1 activity and plasma levels of oxidized fatty acids
(P <0.01).78 Imaizumi etal. reported that administra-
tion of the apo A-I mimetic peptide 4F (which increases
FOCUS ON LIPIDS AND CVD RISK
© 2011 Macmillan Publishers Limited. All rights reserved
228
|
APRIL 2011
|
VOLUME 8 www.nature.com/nrcardio
PON 1 activity79) led to decreased plasma levels of oxi-
dized fatty acids in mice, and that the levels of these
oxidized fattyacids measured in plasma significantly
correlated with the values of lipid oxidation obtained in
a cell-free assay of HDL function that is described below
(P <0.02).80 Morgantini etal. reported that the livers of
mice with diabetes contained high levels of oxidized fatty
acids, which were significantly reduced by treatment with
the 4F peptide (P <0.01).81 Concomitant with a decrease
in the levels of hepatic oxidized fatty acids, a significant
reduction in the development of atherosclerotic lesions
was seen in these mice (P <0.01).81
An import ant determinant of HDL-c holesterol
levels is the phospholipid transfer protein (PLTP). This
protein transfers phospholipids from triglyceride-rich
lipoproteins to HDL, which promotes HDL remodel-
ing. Yan etal. reported that mice deficient in PLTP had
low levels of HDL cholesterol, but that their HDL was
very anti-inflammatory and these mice were protected
from atherosclerosis.82 By contrast, the HDL from mice
overexpressing PLTP and apo A-I was dysfunctional in
its ability to accept cholesterol, and these mice had a 2.2-
fold increased atherosclerotic lesion area compared with
control mice.83
Assessment of HDL function
The HDL-inflammatory index
The ability of HDL to inhibit LDL oxidation and MCP-1
production can be quantified with the use of the HDL-
inflammatory index.6 The index provides a measure
of monocyte chemotactic activity when LDL is added
to endothelial cells in the presence or absence of an
indivi dual’s HDL. Values for healthy individuals were
0.38 ± 0.14 and 0.35 ± 0.11 in two substudies.6 By con-
trast, patients with CHD had values of 1.38 ± 0.91and
1.28 ± 0.29 before treatment. The average index for HDL
from healthy indivi duals was <1.0, therefore, whereas
patients with CHD before treatment had an average
index of >1.0. In a study by Ansell and colleagues, when
26 patients with CHD or another condition associ-
ated with an equivalent risk of a cardiac event were
treated with 40 mg simvastatin daily for 6weeks, the
HDL-inflammatory index significantly decreased from
1.38 ± 0.91to 1.08 ± 0.71 (P = 0.002); however, this value
was still >1.0, indicating that HDL from these patients
increased the production of MCP-1 compared with when
only LDL was added to the assay.6 The same assay was
used in a clinical trial in which 42 patients were enrolled
to test an oral apo A-I mimetic peptide.84 All of these
patients were required to have been receiving a stable dose
of statin (>4weeks) as a condition of enrollment. The
baseline index for these patients was 1.18 ± 0.03, which
was reduced to 0.88 ± 0.02 after treatment. This result is
remarkably similar to thevalue that Anselletal. reported
for their patients at the end of treatment with statins,
namely 1.38 ± 0.91 at baseline, decreasing to 1.08 ± 0.71
after treatment.6 In another study, healthy individuals
had an HDL-inflammatory index of 0.43 ± 0.05,85 which
is similar to that reported for the controls in the study by
Ansell and colleagues.6 By contrast, the index for patients
with end-stage renal disease was >2.0 owing to increased
inflammation associ ated with renal failure.85 In mice, the
values obtained from this cell-based assay correlated with
genetic suscept i bility to atherosclerosis much better than
did levels of HDL cholesterol.86 High indexes in humans
were associ ated with HDL that was less effective in pro-
moting cholesterol efflux from macrophages.86 Rabbits
fed a cholesterol-rich diet had indexes that signifi-
cantly corre lated with aortic lesions and levels of serum
amyloidA protein (P = 0.002 and P = 0.0079, respec-
tively), but not with HDL-cholesterol levels.87 Taken as a
whole, these studies indicate that the HDL-inflammatory
index is a better marker of dysfunctional lipid regula-
tion and could be more useful in assessing the risk of
cardiovascular events than HDL-cholesterol levels.
Cell-free assays of HDL function
The oxidative degradation of lipids (peroxidation)
leads to free-radical reactions and cell damage. A cell-
free assay measuring lipid hydroperoxides in plasma
was developed to determine the inflammatory proper-
ties of HDL.66 Lipid oxidation products in HDL inhibit
antioxidant and anti-inflammatory enzymes associated
with HDL, including PON 1.66 With the use of this assay,
poor outcomes were studied in a cohort of 189 patients
on hemodialysis.88 Patients who had proinflammatory
HDL, as measured by the cell-free assay, with an HDL-
inflammatory index >1, had a survival rate of 12% in
the 900days of the study, whereas patients with anti-
inflammatory HDL with an index <1 had a 63% survival
in that period.88 Ina small study of immigrants from
South Asia livingin the USA, carotid intima–media
thickness substantially correlated with dysfunctional
HDL after adjusting for age, family history of CVD, and
hypertension.89 Women with systemic lupus erythema-
tosus and dysfunctional HDL as determined by this cell-
free assay had an odds ratio of 16.1 of having a carotid
plaque as detected byultrasonography.90
The mRNA levels of the platelet-derived growth-
factor receptor (PDGFR)-β in monocytes were found
to be markedly increased in patients with dysfunctional
HDL.91 The HDL from these patients was shown to be
dysfunctional using the cell-free assay. Addition of HDL
from these patients to the monocytic leukemia THP-1
cell line invitro led to considerably upregulated mRNA
levels of PDGFR-β, MCP-1, PDGF-A, tenascin-C, and
integrin α-X, and also stimulated THP-1 cell migra-
tion. Interestingly, a peptide consisting of amino acids
113–122 from apo J and the apo A-I mimetic peptide 4F
ameliorated these effects.91 This observation is explained
by the high affinity that both of these peptides have
for oxidized lipids; by sequestering proinflammatory
lipids, the peptides reduced their stimulatory effects on
monocytes. In a study of 132 patients with rheumatoid
arthritis, dysfunctional HDL as assessed by the cell-free
assay correlated with age, disease activity, the presence
of erosive disease, non-white race, and smoking.92 These
findings demonstrate the usefulness of the cell-free assay
to study the function and the effects of HDL in the setting
of inflammation.
REVIEWS
© 2011 Macmillan Publishers Limited. All rights reserved
NATURE REVIEWS
|
CARDIOLOGY VOLUME 8
|
APRIL 2011
|
229
HDL dysfunction in other diseases
Studies of the function and properties of HDL in dis-
eases other than CVD have highlighted similarities in
the mechanisms of HDL dysfunction in the setting of
inflammation across a broad range of conditions. Normal
HDL preserves dendritic cell differentiation and function
during mycobacterial infection. The HDL-inflammatory
index was elevated in patients with leprosy to levels
similar to those seen in patients with CHD.93 HDL from
patients with leprosy had an impaired ability to promote
cholesterol efflux from macrophages. Additionally, their
HDL was unable to preserve dendritic cell differentiation
andfunction.93
Holven etal. reported that serum from patients with
hyperhomocysteinemia had low PON 1 activity, and the
ability of their plasma to induce cholesterol efflux from
lipid-loaded macrophages was impaired compared with
plasma from healthy individuals.94 The HDL from these
patients also showed impaired anti-inflammatory prop-
erties in cultures of human umbilical-vein endothelial
cells stimulated with TNF.94
Women with the antiphospholipid syndrome were
reported to have increased carotid intima–media thick-
ness, which was associated with dysfunctional HDL.95
HDL from these women reduced the production of
nitric oxide in cultures of human aortic endothelial cells,
whereas HDL from normal individuals increased the pro-
duction of nitric oxide. The HDL from women with this
syndrome was also less effective in decreasing VCAM-1
expression and superoxide production after treatment of
endothelial cells with TNF, and less effective in preventing
monocyte adhesion to endothelial cells than HDL from
healthy women.95
Insights from drug development
Reconstituted HDL
Reconstituted HDL, consisting of plasma-derived apo A-I
complexed with lipids, has been used to mimic some of
the functions of HDL and enables the study of HDL in
various settings. Reconstituted HDL has been shown
to reduce the development of atherosclerotic lesions in
animal models by increasing HDL levels, and could have
therapeutic potential in humans.96 In order to develop
useful preparations of reconstituted HDL, information on
the major subpopulations of HDL involved in a particular
pathological condition is required. Rye etal. have empha-
sized that HDL is structurally and functionally diverse,
and consists of numerous highly dynamic subpopula-
tions that vary in size, density, surface charge, composi-
tion, and anti-inflammatory properties.55 The possibility
that HDL subpopulations are functionally distinct raises
the important question of which subpopulations should
be therapeutic targets for increasing HDL levels. For
example, the relative importance of HDL containing
apo A-I versus HDL containing apo A-I and apo A-II,
large versussmall HDL, and preβ-1-migrating HDL
speciesversus α-migrating HDL, is currentlynotclear.
In a randomized cross-over study, 13 patients with
type2 diabetes received infusions of reconstituted
HDL (80 mg/kg).97 The reconstituted HDL consisted of
apo A-I isolated from human plasma and phosphatidyl-
choline isolated from soybean, which were combined in
the presence of sodium cholate in a molar ratio of 1:150
to form disc-shaped, noncovalently associated particles
resembling nascent HDL. HDL was isolated from patients
4 h and 72 h following infusion and tested for its ability
to influence the expression of ICAM-1 and VCAM-1 in
cultures of human coronary artery endothelial cells. The
ability of neutrophils to adhere to fibrinogen was deter-
mined, as well as the ability of plasma from these patients
to stimulate cholesterol efflux from THP-1 macro-
phages. The infusion of reconstituted HDL signifi cantly
reduced cell adhesion (P = 0.02) and increased cholesterol
efflux(P <0.05).97
As some studies have shown that HDL might act directly
on leukocytes to reduce inflammation, Peshavariya
etal. examined the effects of reconstituted discoidal
HDL (apo A-I complexed with 1-palmitoyl-2-linoleoyl
phosphatidyl choline in a molar ratio of 1:100) on NADPH-
oxidase activity in leukocytes.98 These investigators
showed that the activity of this enzyme, which generates
superoxide, was inhibited by reconstituted HDL, probably
by disruption of the assembly of NADPH-oxidase subunits
at lipid rafts. In another study, a rabbit model was used to
study arterial inflammation in which a nonoclusive peri-
arterial carotid collar was inserted, resulting in leukocyte
infiltration in the segment of the artery to which the collar
was applied.99 When the rabbits were infused with recon-
stituted HDL from normal human indivi duals, a reduc-
tion in neutrophil infiltration and VCAM-1 and ICAM-1
expression was observed. However, when the rabbits were
infused with reconstituted HDL made from nonenzymati-
cally glycated apo A-I, or taken from patients with dia-
betes, a decrease in the anti-inflammatory response was
seen, indicating that nonenzymatic glycation of apo A-I
reduces its anti-inflammatory properties.99
McGrath etal. showed in a cell-based assay that the
anti-inflammatory properties of reconstituted HDL were
in part due to an upregulation of 3β-hydroxysteroid-Δ24
reductase by HDL, which was dependent on SRB1.100 This
enzyme has known antioxidant and anti- inflammatory
properties, suggesting that low levels result in increased
levels of sterol intermediates, which could be pro-
inflammatory.100 In a sub seque nt study, Pate l etal. demon-
strated that infusion of reconstituted HDL into rabbits
fed with cholesterol significantly increased aortic mRNA
levels of 3β-hydroxysteroid-Δ24 reductase (P <0.0001)
and decreased aortic endothelial VCAM-1 and ICAM-1
expression.101 This observation is consistent with the
role of 3β-hydroxysteriod-Δ24 reductase in suppress-
ing nuclear factor kappaB, the transcription factor that
regulates endothelial VCAM-1 expression, and, to some
extent, ICAM-1 expression.101
Peptide mimetics of HDL
The use of peptide mimetics as research tools to perturb
and understand the basic mechanisms of HDL meta-
bolism and atherosclerosis has been firmly established.
However, determining the role of these peptides as thera-
peutic agents is likely to require many years of study in
FOCUS ON LIPIDS AND CVD RISK
© 2011 Macmillan Publishers Limited. All rights reserved
230
|
APRIL 2011
|
VOLUME 8 www.nature.com/nrcardio
a variety of patient populations. The use of apo A-I to
improve the function of HDL is not economically feasible
owing to the large size of this apolipoprotein, and many
research groups have, therefore, been working to develop
small peptide mimetics. Tabet etal. demonstrated in cul-
tures of human coronary artery endothelial cells and in
a rabbit carotid-collar model that a chemically synthe-
sized peptide of 37 amino acids called 5A was as effec-
tive as reconstituted HDL in reducing the expression of
VCAM-1 and ICAM-1, and, similar to HDL, exerted its
effects through ABCA1.102
One of the most commonly used peptides in studies of
HDL is the 18-amino-acid apo A-I mimetic peptide 4F.
Bloedon and colleagues reported on the safety and ability
of 4F to improve the protective capacity of HDL against
invitro LDL oxidation in patients at high cardio vascular
risk.84 The HDL-inflammatory index signifi cantly
improved in patients receiving the peptide compared
with patients receiving placebo (P <0.05), andwas well-
tolerated. Smythies etal. reported that 4F and apo A-I
stimulated cholesterol efflux from macrophages invitro,
leading to cholesterol depletion and disruption of lipid
rafts, which induced an anti-inflammatory response.103 Dai
etal. demonstrated that administration of 4F (10 mg/kg)
to lipopolysaccharide-treated rats resulted in retention
of lipopolysaccharide in the HDL fraction, which was
associated with neutralization of endotoxins and reduced
mortality.104 The 4F peptide also improved the HDL-
inflammatory index in a mouse model of systemic lupus
erythematosus and, when administered with pravastatin,
reduced the number of macrophages in athero sclerotic
lesions, increased the smooth muscle cell content of
these lesions, and reduced circulating levels of soluble
VCAM-1.105 Wool etal. reported that the 4F peptide
reduced nascent atherosclerosis and induced natural
antibody production against oxidation-specific epitopes
in apo-E-null mice.106
Another apo A-I mimetic peptide, consisting of 26
amino acids, ATI-5261, promoted cholesterol efflux
from macrophages invitro via ABCA1 at least as well
as apo A-I.107 When this peptide was administered by
injection, it significantly reduced atherosclerosis in LDL-
receptor-null mice by 30% (P = 0.011) and in apo-E-null
mice by 45% (P = 0.00016).
Ruchala etal. synthesized a number of other peptides
with sequences resembling that of a cholesterol-binding
domain found in pore-forming bacterial exotoxins.108
One of these peptides, oxpholipin-11D, bound chol-
esterol and oxidized phospholipids, and improved the
HDL-inflammatory index in a mouse model of athero-
sclerosis.108 This peptide had similar properties to the
well-characterized 4F peptide, and could be a potential
novel therapy. As the examples discussed here indicate,
results from studies of HDL mimetic peptides in animals
and early clinical trials are encouraging, but we will
probably have to wait some time before the outcomes of
definitive studies are known.
Conclusions
HDL-cholesterol levels are inversely related to the occur-
rence of clinical events, although many patients who
suffer a cardiovascular event have normal, or even high,
levels of HDL cholesterol. HDL-cholesterol levels provide
information about the size of the HDL pool, but do not
predict the composition or function of the HDL parti-
cles. The major functions of HDL are to promote reverse
cholesterol transport and modulate inflammation. These
functions are influenced by the composition of HDL,
which is altered during the acute-phase response. The
oxidative environment at a site of inflammation modi-
fies HDL proteins, which reduces the ability of HDL to
promote cholesterol efflux, and changes HDL from an
anti-inflammatory particle to a proinflammatory parti-
cle. Results from preclinical studies indicate that small
peptide mimetics of apo A-I can improve HDL function
and reduce inflammation without necessarily altering
HDL-cholesterol levels. Robust assays to test the func-
tion of HDL are needed to supplement the measurement
of HDL-cholesterol levels in the clinic.
Review criteria
The articles selected for this Review were obtained from
searches of PubMed using the terms “ATP-binding cassette
transporters ABCA1 and ABCG1”, “apolipoproteins”,
“apolipoprotein mimetic peptides”, “cholesterol efflux”,
“high-density lipoprotein”, “HDL”, “HDL-associated
proteins”, “HDL cholesterol”, “HDL inflammatory
properties”, “reconstituted HDL”, and “reverse cholesterol
transport”. Selected papers were full-text original articles
and reviews published between 1970 and 2010. Abstracts
were not included. Reference lists of the identified papers
were searched for additional material.
1. Gordon, T., Castelli, W.P. , H j o r t l a n d , M . C.,
Kannel, W.B. & Dawber, T.R. High density
lipoprotein as a protective factor against
coronary heart disease. The Framingham Study.
Am. J. Med. 62, 707–7 14 (19 77).
2. Lewington, S. etal. Blood cholesterol and
vascular mortality by age, sex, and blood
pressure: a meta-analysis of individual data from
61 prospective studies with 55,000 vascular
deaths. Lancet 370, 1829–1839 (2007).
3. Di Angelantonio, E. etal. Maj or lipids ,
apolipoproteins, and risk of vascular disease.
JAMA 302, 19 93–20 00 (2009 ).
4. Briel, M. etal. Association between change in
high density lipoprotein cholesterol and
cardiovascular disease morbidity and mortality:
systematic review and meta-regression analysis.
BMJ 338, b92 (2009 ).
5. van der Steeg, W.A. etal. High -density lipoprotein
cholesterol, high-density lipoprotein particle size,
and apolipoproteinA-I: significance for
cardiovascular risk. The IDEAL and EPIC-Norfolk
studies. J. Am. Coll. Cardiol. 51, 63 4–642 (2008).
6. Ansell, B.J. etal. Inf lamma tory /
antiinflammatory properties of high-density
lipoprotein distinguish patients from control
subjects better than high-density lipoprotein
cholesterol levels and are favorably affected by
simvastatin treatment. Circulation 108,
2751–2756 (2003).
7. Lund-Katz, S. & Phillips, M.C. High density
lipoprotein structure-function and role in reverse
cholesterol transport. Subcell. Biochem. 51,
183–227 (2010).
8. Vedha ch alam , C. etal. Influence of
apolipoprotein (Apo) A-I structure on nascent
high density lipoprotein (HDL) particle size
distribution. J. Biol. Chem. 285, 31965–31973
(2010).
9. Eisenberg, S. High density lipoprotein
metabolism. J.Lipid Res. 25, 10 17–1058 (1984).
10. Fagerholm, U. Prediction of human
pharmacokinetics—biliary and intestinal
clearance and enterohepatic circulation.
J. Pharm. Pharmacol. 60, 535–5 42 (20 08).
11. Praseetha, S. & Thampan, R.V. Regulatory
factors in steroid hormone biosynthesis. Crit.
Rev. Eukar yot. Gene Expr. 19, 25 3–265 (2009).
REVIEWS
© 2011 Macmillan Publishers Limited. All rights reserved
NATURE REVIEWS
|
CARDIOLOGY VOLUME 8
|
APRIL 2011
|
231
12. Simons, K. & Gerl, M.J. Revitalizing membrane
rafts: new tools and insights. Nat. Rev. Mol. Cell
Biol. 11, 688–69 9 (201 0).
13. Steck, T.L. & Lange, Y. Cell cholesterol
homeostasis: mediation by active cholesterol.
Tren ds Ce ll Bi ol. 20, 680 –687 ( 2010).
14. Levitan, I., Volkov, S. & Subbaiah, P.V. Oxidized
LDL: diversity, patterns of recognition, and
pathophysiology. Antioxid. Redox Signal. 13,
39–75 (2010).
15. Baldán, A., Bojanic, D.D. & Edward s, P.A.
The ABCs of sterol transport. J. Lipid Res. 50,
S80–S85 (2009).
16. Yvan-Charvet, L., Wang, N. & Tall, A.R. Role of
HDL, ABCA1, and ABCG1 transporters in
cholesterol efflux and immune responses.
Arterioscler. Thromb. Vasc. Biol. 30, 139–143
(2010).
17. Out, R. etal. Coexistence of foam cells and
hypocholesterolemia in mice lacking the ABC
transporters A1 and G1. Circ. Res. 102,
113–120 (2008).
18. de la Llera-Moya, M. etal. The ability to promote
efflux via ABCA1 determines the capacity of
serum specimens with similar high-density
lipoprotein cholesterol to remove cholesterol
from macrophages. Arterioscler. Thromb. Vasc.
Biol. 30, 796–80 1 (201 0).
19. Chau, P., Nakamura, Y., Fielding, C.J. &
Fielding,P. E. Mechanism of prebeta-HDL
formation and activation. Biochemistry 45,
3981–3987 (2006).
20. Wu, Z. etal. Double superhelix model of high
density lipoprotein. J. Biol. Chem. 284,
36605–36619 (2009).
21. Jones, M.K., Catte, A., Li, L. & Segrest, J.P.
Dynamics of activation of lecithin:cholesterol
acyltransferase by apolipoprotein A-I.
Biochemistry 48, 11196 –1121 0 (2009) .
22. Gogonea, V. etal. Congruency between
biophysical data from multiple platforms and
molecular dynamics simulation of the double-
super helix model of nascent high-density
lipoprotein. Biochemistry 49, 7323–7 343 (201 0).
23. Yvan-Charvet, L. etal. ABCA1 and ABCG1 protect
against oxidative stress-induced macrophage
apoptosis during efferocytosis. Circ. Res. 106,
1861–1869 (2010).
24. Wester terp, M. etal. Increased atherosclerosis
in mice with vascular ATP-binding cassette
transporter G1 deficiency—brief report.
Arteroscler. Thromb. Vasc. Biol. 30, 2103–2105
(2010).
25. Masson, D. etal. Increased HDL cholesterol and
apoA-I in humans and mice treated with a novel
SR-BI inhibitor. Arterioscler. Thromb. Vasc. Biol.
29, 205 4–2060 ( 2009) .
26. Wang, X . etal. Macroph age AB CA1 an d ABCG 1,
but not SR-BI, promote macrophage reverse
cholesterol transport invivo. J. Clin. Invest. 117,
2216–2224 (2007).
27. Rader, D.J., Alexander, E.T., W e i b e l , G . L.,
Billheimer, J. & Rothblat, G.H. The role of
reverse cholesterol transport in animals and
humans and relationship to atherosclerosis.
J. Lipid Res. 50, S189 –S194 (2009).
28. Sethi, A.A. etal. Hig h pre-β1 HDL concentrations
and low lecithin:cholesterol acyltransferase
activities are strong positive risk markers for
ischemic heart disease and independent of HDL-
cholesterol. Clin. Chem. 56, 1128 –1137 (2010).
29. Zheng, L. etal. Apolipoprotein A-I is a selective
target for myeloperoxidase-catalyzed oxidation
and functional impairment in subjects with
cardiovascular disease. J. Clin. Invest. 114,
529–541 (2004).
30. Zheng, L. etal. Localization of nitration and
chlorination sites of apolipoprotein A-I catalyzed
by myeloperoxidase in human atheroma and
associated oxidative impairment in ABCA1-
dependent cholesterol efflux from macrophages.
J. Biol. Chem. 280, 38–47 (2005 ).
31. Bergt, C. etal. The myeloperoxidase product
hypochlorous acid oxidizes HDL in the human
artery wall and impairs ABCA1-dependent
cholesterol transport. Proc. Natl Acad. Sci. USA
101, 130 32–1303 7 (200 4).
32. Shao, B. etal. My elope roxid ase impai rs ABCA1 -
dependent cholesterol efflux through methionine
oxidation and site-specific tyrosine chlorination
of apolipoprotein A-I. J. Biol. Chem. 281,
9001–9004 (2006).
33. Wu, Z. etal. The refined structure of nascent
HDL reveals a key functional domain for particle
maturation and dysfunction. Nat. Struct. Mol.
Biol. 14, 861–86 8 (200 7).
34. Shao, B., Cavigiolio, G., Brot, N., Oda, M.N.
& Heinecke, J.W. Methionine oxidation impairs
reverse cholesterol transport by apolipoprotein
A-I. Proc. Natl Acad. Sci. USA 105, 12224–12229
(2008).
35. Shao, B., Tang, C., Heinecke, J. & Oram, J.F.
Oxidation of apolipoprotein A-I by
myeloperoxidase impairs the initial interactions
with ABCA1 required for signaling and cholesterol
export. J. Lipid Res. 51, 1849–1 858 (201 0).
36. Proudfoot, J.M. etal. HDL is the major
lipoprotein carrier of plasma F2-isoprostanes.
J. Lipid Res. 50, 716– 722 (200 9).
37. Shao, B. & Heinecke, J.W. HDL, lipid
peroxidation, and atherosclerosis. J. Lipid Res.
50, 599 –601 (20 09).
38. Cavigiolio, G., Geier, E.G., Shao, B., Heinecke,J.W.
& Oda, M.N. Exchange of apolipoprotein A-I
between lipid-associated and lipid-free states:
a potential target for oxidative generation of
dysfunctional high density lipoproteins. J. Biol.
Chem. 285, 18847– 18857 (201 0).
39. Fogelman, A.M. etal. Malondialdehyde
alteration of low density lipoproteins leads to
cholesteryl ester accumulation in human
monocyte-macrophages. Proc. Natl Acad. Sci.
USA 77, 2214–2 218 (1 980).
40. Shao, B. etal. Modifying apolipoprotein A-I by
malondialdehyde, but not by an array of other
reactive carbonyls, blocks cholesterol efflux by
the ABCA1 pathway. J. Biol. Chem. 285,
18473–18484 (2010).
41. Steinberg, D. The cholesterol controversy is over.
Why did it take so long? Circ ulati on 80,
1070–1078 (1989).
42. Cohen, J.C. etal. Multiple rare alleles contribute
to low plasma levels of HDL cholesterol. Science
305, 869 –872 (20 04).
43. Gomaraschi, M. etal. Normal vascular function
despite low levels of high-density lipoprotein
cholesterol in carriers of the apolipoprotein
A-IMilano mut ant. Circulation 116, 2165–2172
(2007).
44. Ibanez, B. etal. Ra pid ch ange i n plaque s ize,
composition, and molecular footprint after
recombinant apolipoprotein A-IMilano (ETC-216)
administration: magnetic resonance imaging
study in an experimental model of
atherosclerosis. J. Am. Coll. Cardiol. 51,
1104–1109 (2008).
45. Alexander, E.T. etal. Macrophage reverse
cholesterol transport in mice expressing apoA-I
Milano. Arterioscler. Thromb. Vasc. Biol. 29,
1496–1501 (2009).
46. Inazu, A. etal. In creas ed hig h-density lipoprotein
levels caused by a common cholester yl-ester
transfer protein gene mutation. N. Engl. J. Med.
323, 123 4–1238 ( 1990) .
47. Edmondson, A.C. etal. Loss-of-function variants
in endothelial lipase are a cause of elevated
HDL cholesterol in humans. J. Clin. Invest. 119,
1042–1050 (2009).
48. Khera, A.V. , W o l f e , M . L., Cannon, C.P., Qin, J.
& Rader, D.J. On-statin cholesteryl ester transfer
protein mass and risk of recurrent coronar y
events (from the pravastatin or atorvastatin
evaluation and infection therapy-thrombolysis in
myocardial infarction 22 [PROVE IT-TIMI 22]
study). Am. J. Cardiol. 106, 451– 456 (201 0).
49. Castrillo, A. etal. Crosstalk between LXR and
toll-like receptor signaling mediates bacterial
and viral antagonism of cholesterol metabolism.
Mol. Cell 12, 8 05–81 6 (2003) .
50. Braun, A. etal. 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
apolipoproteinE-deficient mice. Cir c. Res. 90,
270–276 (2002).
51. Van E ck, M. etal. Increased oxidative stress
in scavenger receptor BI knockout mice with
dysfunctional HDL. Arterioscler. Thromb. Vasc.
Biol. 27, 2413–2 419 (200 7).
52. Van L enten, B. J. etal. Ant i-inflammatory HDL
becomes pro-inflammatory during the acute
phase response. Loss of protective effect of
HDL against LDL oxidation in aortic wall cell
cocultures. J. Clin. Invest. 96, 2758–2767
(1995).
53. Navab, M. etal. Normal high density lipoprotein
inhibits three steps in the formation of mildly
oxidized low density lipoprotein: steps 2 and 3.
J. Lipid Res. 41, 1495 –1508 (2000).
54. Kontush, A. & Chapman, M.J. Antiatherogenic
function of HDL particle subpopulations: focus
on antioxidative activities. Curr. Opin. Lipidol. 21,
312–318 (2010).
55. Rye, K.A., Bursill, C.A., Lambert, G., Tabet, F.
& Barter, P.J. The metabolism and anti-atherogenic
properties of HDL. J. Lipid Res. 50, S195–S200
(2009).
56. Navab, M. etal. Monocyte transmigration
induced by modification of low density
lipoprotein in cocultures of human aortic wall
cells is due to induction of monocyte
chemotactic protein1 synthesis and is
abolished by high density lipoprotein. J. Clin.
Invest. 88, 2039 –2046 (1991).
57. Navab, M. etal. Normal high density lipoprotein
inhibits three steps in the formation of mildly
oxidized low density lipoprotein: step 1. J. Lipid
Res. 41, 1481–1 494 (200 0).
58. Corsetti, J.P. , Z a r e b a , W . , M o s s , A . J.,
Rainwater,D.L. & Sparks, C.E. Elevated HDL is
a risk factor for recurrent coronary events in a
subgroup of non-diabetic postinfarction patients
with hypercholesterolemia and inflammation.
Atherosclerosis 187, 191–19 7 (2006) .
59. Corsetti, J.P. , G a n s e v o o r t , R . T. , S p a r k s , C . E.
& Dullaart, R.P. Inflammation reduces HDL
protection against primary cardiac risk. Eur. J.
Clin. Invest. 40, 483– 489 (201 0).
60. Corsetti, J.P. etal. Cholesteryl ester transfer
protein polymorphism (TaqIB) associates with
risk in postinfarction patients with high
C-reactive protein and high-density lipoprotein
cholesterol levels. Arterioscler. Thromb. Vasc.
Biol. 30, 1657–1 664 (201 0).
61. Undurti, A. etal. Modification of high density
lipoprotein by myeloperoxidase generates a
pro-inflammatory particle. J. Biol. Chem. 284,
30825–30835 (2009).
62. McGillicuddy, F.C. etal. Inflammation impairs
reverse cholesterol transport invivo. Circulation
119, 113 5–1145 ( 2009) .
63. Watanabe , J. etal. Hemoglobin and its
scavenger protein haptoglobin associate with
FOCUS ON LIPIDS AND CVD RISK
© 2011 Macmillan Publishers Limited. All rights reserved
232
|
APRIL 2011
|
VOLUME 8 www.nature.com/nrcardio
apoA-1-containing particles and influence the
inflammatory properties and function of high
density lipoprotein. J. Biol. Chem. 284,
18292–18301 (2009).
64. Van L enten, B. J. etal. Hig h-density lipoprotein
loses its anti-inflammatory properties during
acute influenzaA infection. Circulation 103,
2283–2288 (2001).
65. Navab, M. etal. Mildly oxidized LDL induces an
increased apolipoprotein J/paraoxonase ratio.
J. Clin. Invest. 99, 200 5–201 9 (1997) .
66. Navab, M. etal. A cell-free assay for detecting HDL
that is dysfunctional in preventing the formation
of or inactivating oxidized phospholipids. J. Lipid
Res. 42, 1308–1 317 (2001) .
67. Dullaart, R.P. etal. Plasma lecithin:cholesterol
acyltransferase activity modifies the inverse
relationship of C-reactive protein with HDL
cholesterol in nondiabetic men. Biochim.
Biohphys. Acta 1801, 84 –88 (201 0).
68. Patel, S. etal. Acute hypertriglyceridaemia in
humans increases the triglyceride content and
decreases the anti-inflammatory capacity of high
density lipoproteins. Atherosclerosis 204,
424–428 (2009).
69. Vais ar, T. etal. Shotgun proteomics implicates
protease inhibition and complement activation
in the antiinflammatory properties of HDL. J. Clin.
Invest. 117, 746– 756 (200 7).
70. Hoofnagle, A.N. & Heinecke, J.W.
Lipoproteomics: using mass spectrometr y-
based proteomics to explore the assembly,
structure, and function of lipoproteins. J. Lipid
Res. 50, 1967–1 975 (200 9).
71. Spagnuolo, M.S., Cigliano, L., D’Andrea, L.D.,
Pedone, C. & Abrescia, P. Assignment of the
binding site for haptoglobin on apolipoproteinA-I.
J. Biol. Chem. 280, 1193– 1198 (20 05).
72. Watanabe, J. etal. Differential association of
hemoglobin with proinflammatory high density
lipoproteins in atherogenic/hyperlipidemic mice.
A novel biomarker of atherosclerosis. J. Biol.
Chem. 282, 23698– 23707 (2007).
73. Asleh, R. etal. Correction of HDL dysfunction in
individuals with diabetes and the haptoglobin2–2
genotype. Diabetes 57, 2794–28 00 (2008 ).
74. Levy, A.P. etal. Haptoglobin: basic and clinical
aspects. Antioxid. Redox Signal. 12, 293–304
(2010).
75. Asleh, R. & Levy, A.P. Divergent effects of
α-tocopherol and vitaminC on the generation of
dysfunctional HDL associated with diabetes and
Hp 2–2 genotype. Antioxid. Redox Signal. 12,
209–217 (2010).
76. Mackness, B. & Mackness, M. Anti-inflammatory
properties of paraoxonase-1 in atherosclerosis.
Adv. Exp. Med. Biol. 660, 143 –151 ( 2010).
77. James, R.W. etal. The scavenger receptor
classB, typeI is a primary determinant of
paraoxonase-1 association with high-density
lipoproteins. Arterioscler. Thromb. Vasc. Biol. 30,
2121–2127 (2010).
78. Bhattacharyya, T. etal. Relationship of
paraoxonase1 (PON1) gene polymorphisms and
functional activity with systemic oxidative stress
and cardiovascular risk. JA MA 299, 1265–1276
(2008).
79. Navab, M. etal. Oral D-4F causes formation of
pre-β hig h-density lipoprotein and improves high-
density lipoprotein-mediated cholesterol efflux
and reverse cholesterol transport from
macrophages in apolipoproteinE-null mice.
Circulation 109, 3215–322 0 (200 4).
80. Imaizumi, S. etal. L-4F differentially alters
plasma levels of oxidized fatty acids resulting in
more anti-inflammatory HDL in mice. Drug
Metab. Lett. 4, 1 39–14 8 (2010) .
81. Morgantini, C. etal. ApoA-I mimetic peptides
prevent atherosclerosis development and
reduce plaque inflammation in a mouse model
of diabetes. Diabetes 59, 322 3–322 8 (2010) .
82. Yan , D. , Na vab , M. , Br uc e, C ., Fog elm an , A. M.
& Jiang, X.C. PLTP deficiency improves the anti-
inflammatory properties of HDL and reduces the
ability of LDL to induce monocyte chemotactic
activity. J. Lipid Res. 45, 1852 –1858 (2 004).
83. Moerland, M. etal. Atherogenic, enlarged, and
dysfunctional HDL in human PLTP/apoA-I double
transgenic mice. J. Lipid Res. 48, 2622–2631
(2007).
84. Bloedon, L.T. etal. Saf ety, phar macok ineti cs,
and pharmacodynamics of oral apoA-I mimetic
peptide D-4F in high-risk cardiovascular patients.
J. Lipid Res. 49, 1344 –1352 (2008).
85. Vazi ri, N.D., Mora di, H ., Pah l, M. V. ,
Fogelman,A.M. & Navab, M. Invitro stimulation
of HDL anti-inflammatory activity and inhibition of
LDL pro-inflammatory activity in the plasma of
patients with end-stage renal disease by an apoA-I
mimetic peptide. Kidney Int. 76, 43 7–444 (2009).
86. Navab, M. etal. The double jeopardy of HDL.
Ann. Med. 37, 1 73–17 8 (2005) .
87. Van L enten, B. J. etal. Lipoprotein inflammatory
properties and serum amyloidA levels but not
cholesterol levels predict lesion area in
cholesterol-fed rabbits. J. Lipid Res. 48,
2344–2353 (2007).
88. Kalantar-Zadeh, K., Kopple, J.D., Kam ran pour,
N., Fogelman, A.M. & Navab, M. HDL-
inflammatory index correlates with poor
outcome in hemodialysis patients. Kidney Int.
72, 114 9–1156 ( 2007) .
89. Dodani, S. etal. Can dysfunctional HDL explain
high coronary artery disease risk in South
Asians? Int. J. Cardiol. 129, 125– 132 (200 8).
90. McMahon, M. etal. Dysfunctional
proinflammatory high-density lipoproteins confer
increased risk of atherosclerosis in women with
systemic lupus erythematosus. Arthritis Rheum.
60, 242 8–2437 ( 2009) .
91. Skaggs, B.J., Hahn, B.H., Sahakian, L.,
Grossman, J. & McMahon, M. Dysfunctional, pro-
inflammatory HDL directly upregulates monocyte
PDGFRβ, che motax is and TNFα p roductio n. Clin.
Immunol. 137, 147–15 6 (2010) .
92. Charles-Schoeman, C. etal. Abnormal function
of high-density lipoprotein is associated with
poor disease control and an altered protein
cargo in rheumatoid ar thritis. Arthritis Rheum.
60, 287 0–2879 ( 2009) .
93. Cruz, D. etal. Host-derived oxidized phospholipids
and HDL regulate innate immunity in human
leprosy. J. Clin. Invest. 118, 2917–2 928 (2008) .
94. Holven, K.B. etal. The antiatherogenic function
of HDL is impaired in hyperhomocysteinemic
subjects. J. Nutr. 138, 2070 –2075 (2008).
95. Charakida, M. etal. Vas cular abnor malit ies,
paraoxonase activity, and dysfunctional HDL in
primary antiphospholipid syndrome. JAMA 302,
1210–1217 (2009).
96. Atkinson, D. & Small, D.M. Recombinant
lipoproteins: implications for structure and
assembly of native lipoproteins. Annu. Rev.
Biophys. Biophys. Chem. 15, 403–4 56 (1986 ).
97. Patel, S. etal. Recon stitu ted hi gh-density
lipoprotein increases plasma high-density
lipoprotein anti-inflammatory properties and
cholesterol efflux capacity in patients with type2
diabetes. J. Am. Coll. Cardiol. 53, 96 2–971 (2009).
98. Peshava riya, H. etal. Re const itute d high -density
lipoprotein suppresses leukocyte NADPH
oxidase activation by disrupting lipid rafts. Free
Radic. Res. 43, 7 72–78 2 (2009) .
99. Nobécourt, E. etal. Nonenzymatic glycation
impairs the antiinflammatory properties of
apolipoproteinA-I. Arterioscler. Thromb. Vasc.
Biol. 30, 766–77 2 (2010) .
100. McGrath, K.C. etal. Role of
3β-hydroxysteroid-Δ24 reductase in mediating
antiinflammatory effects of high-density
lipoproteins in endothelial cells. Arterioscler.
Thromb. Vasc. Biol. 29, 8 77–88 2 (2009) .
101. Patel, S. etal. Ant i-inflammatory effects of
apolipoprotein A-I in the rabbit. Atherosclerosis
212, 392 –397 (20 10).
102. Ta be t, F. etal. The 5A apolipoprotein A-I mimetic
peptide displays anti-inflammatory and
antioxidant properties invivo and invitro.
Arterioscler. Thromb. Vasc. Biol. 30, 246–252
(2010).
103. Smythies, L.E. etal. Apolipoprotein A-I mimetic
4F alters the function of human monocyte-
derived macrophages. Am. J. Physiol. Cell Physiol.
298, C15 38–C154 8 (201 0).
104. Dai, L. etal. The apolipoprotein A-I mimetic
peptide 4F prevents defects in vascular function
in endotoxemic rats. J. Lipid Res. 51,
2695–2705 (2010).
105. Woo, J.M.etal. Treatment with apolipoprotein
A-1 mimetic peptide reduces lupus-like
manifestations in a murine lupus model of
accelerated atherosclerosis. Ar th ri ti s Re s. T he r.
12, R93 (2010).
106. Wool, G. D. etal. 4F peptide reduces nascent
atherosclerosis and induces natural antibody
production in apolipoproteinE-null mice. FASEB
J. 25, 290–30 0 (201 1).
107. Bielicki, J.K. etal. A new HDL mimetic peptide
that stimulates cellular cholesterol efflux with
high efficiency greatly reduces atherosclerosis
in mice. J. Lipid Res. 51, 1496–150 3 (2010) .
108. Ruchala, P. etal. Oxpho lipin 11D: an an ti-
inflammatory peptide that binds cholesterol and
oxidized phospholipids. PLoS ONE 5, e10181
(2010).
Acknowledgments
This work was supported in par t by USPHS Grant
HL30568, and the Laubisch Fund and M.K. Gray
Fund at UCLA.
Author contributions
M. Navab and A.M. Fogelman contributed to
discussion of content for the article, researched data
to include in the manuscript, wrote the ar ticle,
reviewed and edited the manuscript before
submission, and revised the manuscript in response
to the peer-reviewers’ comments. S.T. Reddy and
B.J. Van Lenten reviewed and edited the manuscript
before submission.
REVIEWS
© 2011 Macmillan Publishers Limited. All rights reserved
