The Association of SAA With Apolipoprotein B Affects Cardiovascular Risk

Clinical burden of CVD: CVD is the leading cause of death in developed nations and the VA
population is no exception. Despite decades of research, technical, and pharmacological
advances, CVD remains a major public health problem. This is partly due to our impaired
ability to identify subjects at greatest risk for CVD events and thus the best candidates for
pharmacological risk reducing therapies, and partly due to incomplete use or efficacy of
currently available therapies. Epidemiological studies have identified major risk factors for
CVD including elevated LDL cholesterol, low HDL cholesterol, hypertension, smoking and
diabetes. However, despite targeting individuals with these risk factors with aggressive
pharmacological interventions, CVD remains a major public health problem. Furthermore, even
in individuals with risk factors who are treated with pharmacological or lifestyle
interventions the CVD event rates are higher than in those who never had the risk factors.
Recent epidemiologic data evaluating the American Heart Association-identified cardiovascular
health metrics reported that the prevalence of having CVD risk factors at ideal levels is <
2%10; implying that >98% of the population are candidates for risk reduction. Clearly, health
systems cannot cope with pharmacological interventions for such enormous target populations.
Thus, additional risk stratifying markers are needed to identify those at highest risk for
events and thus at greatest likelihood of benefit. Several biomarkers, including the acute
phase reactants C reactive protein (CRP) and serum amyloid A (SAA) have been studied for
their role in predicting CVD events. Both CRP and SAA are chronically elevated in individuals
with obesity, metabolic syndrome (MetS), diabetes, rheumatoid arthritis, lupus and other
chronic inflammatory conditions associated with increased CVD rates, raising the question of
whether these biomarkers merely reflect underlying risk or play a causative role in CVD.
Although emerging evidence has cast doubt on the role of CRP as a causative factor the
investigators and others recently demonstrated that SAA is directly atherogenic in animal
models. Thus, in addition to its role as a biomarker for CVD, SAA may play a causal role in
CVD.

SAA: SAA is a family of acute phase proteins synthesized primarily in the liver. In healthy
individuals SAA concentrations are < 5 mg/L but during an acute phase response SAA can
increase up to 1000 mg/L for a few days, then it rapidly returns to baseline levels. However,
chronic inflammatory states such as obesity, MetS, diabetes, rheumatoid arthritis etc, are
associated with persistently and significantly elevated SAA concentrations of 30-100 mg/L.
Acute elevations in SAA are proposed to play a major role in response to injury and
inflammation, participating in cholesterol delivery to injured tissues, recruitment of
inflammatory cells, and induction of tissue repair cytokines. However, the chronic elevations
of SAA now prevalent in modern society likely reflect a maladaptive response and numerous
studies are now examining potential roles of SAA in disease pathology. Using murine models in
which acute phase SAA is over-expressed, the investigators and others demonstrated direct
increases in atherosclerosis development.

SAA and apolipoprotein B (apoB) containing lipoproteins: SAA is a lipid binding
apolipoprotein and lipid-free SAA has not been found in vivo. The dogma is that SAA is
exclusively an HDL associated lipoprotein; however, the investigators and others have
reported SAA on apoB-containing lipoproteins in both mice and humans. Several studies have
reported on a complex termed SAA-LDL associated with components of MetS, remnant like
particle cholesterol, smoking status, lifestyle interventions, and statin treatment. These
studies suggest that SAA-LDL is a risk factor for CVD. In new preliminary studies the
investigators demonstrate that SAA has a differential lipoprotein association in diabetes,
and in post-prandial lipoprotein metabolism, and the investigators demonstrate that the
presence of SAA on apoB-lipoproteins augments their proteoglycan binding, a key step in
atherosclerosis development. Thus, emerging evidence suggests that the presence of SAA on
apoB-lipoproteins may be a novel CVD risk factor, play a causal role in atherosclerosis, and
thus be a therapeutic target.

Post-prandial apoB-lipoprotein metabolism: The various lipoproteins are defined based on size
and density criteria, as well as by their protein constituents. However, even within each
lipoprotein class there is considerable heterogeneity, as the particles undergo continuous
remodeling. Briefly, lipids consumed in the diet associate with apoB-48 to form chylomicrons,
which are transported in intestinal lymphatics before entering the bloodstream. Various
enzymes act on newly formed chylomicrons shifting lipids and proteins between chylomicrons
and HDLs before the chylomicron remnants are taken up by the liver. The liver re-packages the
lipids into VLDL particles containing apoB-100. The hydrolysis of VLDL results in smaller
apoB-100 particles called VLDL remnants or intermediate density lipoproteins (IDLs).
Collectively, these particles are termed triglyceride rich lipoproteins (TGRLs).

Ongoing remodeling of TGRLs by various lipases leads to the formation of LDL. LDL can be
taken up by peripheral tissues, including the vasculature, or by the liver. The
sub-endothelial retention of apoB-containing particles initiates atherosclerosis.

Post-prandial lipoproteins and CVD: Elevated levels of LDL cholesterol and low levels of HDL
cholesterol are documented risk factors for CVD and contribute causally to atherogenesis.
However, individuals with obesity, MetS and diabetes do not typically have elevated LDL
cholesterol; their dyslipidemia is characterized by elevated triglycerides and low HDL
cholesterol. The role of triglycerides as a CVD risk factor remains controversial; however,
post-prandial triglycerides may be a more significant risk factor than fasting triglycerides.
As humans spend most of their lives in the post-prandial state, there is ongoing interest in
the role of post-prandial lipoprotein metabolism in CVD risk. However, most studies have
relied on fasting lipoprotein samples; triglycerides are the lipoprotein component most
affected by food consumption. The mechanisms accounting for the excess prevalence of CVD in
MetS and diabetic subjects beyond that predicted by the traditional CVD risk factors remain
unclear; however, insulin resistant states are characterized by increased intestinal apoB48
production, increased TGRL production and delayed lipoprotein clearance, which may contribute
to CVD prevalence. Retention of apoB- containing lipoproteins in the vascular wall by the
ionic interaction between apoB and proteoglycans, leads to the initiation of atherosclerosis.
Lipolysis of VLDL more than doubles its ability to cross the endothelium and deposit lipids
in the subendothelial space. TGRLs have proportionately more triglyceride than cholesterol:
however, their size means that they can deposit 5-20 times more cholesterol per particle in
the subendothelial space compared to an LDL particle. Increased TGRL production and delayed
particle clearance increases the likelihood of particle retention and cholesterol deposition
in the subendothelial space. The investigators have novel preliminary data demonstrating that
the presence of SAA on apoB-containing lipoproteins increases their proteoglycan binding. The
investigators propose that the increased presence of SAA on apoB-containing post-prandial
lipoproteins in insulin resistant states increases the atherogenicity of these particles and
could be a mechanism accounting for the increased CVD prevalence in insulin resistant states
such as MetS and diabetes.

HDL metabolism: Like VLDL and LDL, HDL comprises a range of particles; however, HDL does not
contain apoB, instead containing apoA-I. HDL is often separated into two major classes by
size and density: the large HDL2 and the smaller HDL3. As discussed above, HDL undergoes
continuous lipid interchange with various apoB-containing lipoprotein particles. A change in
lipoprotein structure or composition by various enzymes is termed remodeling. HDL is
typically thought to be an atheroprotective lipoprotein due to its ability to transport
cholesterol away from the periphery back to the liver. In addition, HDL has a number of other
beneficial properties including anti-inflammatory and anti-oxidative functions. In insulin
resistant states HDL levels tend to be low, and some studies suggest its beneficial
properties are reduced. Remodeling of lipoproteins affects their functionality and half-life;
for example, the remodeling of HDL by CETP (which transfers triglycerides from TGRL to HDL
and cholesterol ester from HDL to TGRL) predisposes HDL to enhanced catabolism and is thought
to contribute to the lower levels of HDL seen in insulin resistant states. Although the
paradigm is that SAA is a HDL associated lipoprotein, in preliminary studies the
investigators have found SAA on apoB particles in insulin resistant persons in the
post-prandial period. However, it is not clear how SAA associates with either HDL or
apoB-lipoprotein particles.

SAA lipoprotein association: In the setting of an acute phase response SAA levels can
increase up to 1000-fold; however, even at these highly elevated levels SAA remains
exclusively found on HDL particles. Thus, there is no evidence of a "maximum capacity" of HDL
for SAA. How SAA associates with either HDL or apoB-lipoprotein particles is not fully
understood. SAA is thought to be produced by the liver in a lipid-free form and bind
lipoproteins extracellularly, or in plasma SAA has been shown to induce HDL biogenesis via
ATP binding cassette 1 (ABCA1), which may be a major mechanism by which SAA associates with
HDL. Murine studies using knockout mice demonstrated that in the absence of HDL, SAA was
found on apoB-particles. However, the investigators and others have reported SAA on apoB
particles despite the presence of HDL. In new preliminary studies the investigators found
that the remodeling of HDL led to the liberation of both lipid-poor apoA-I and lipid-poor
SAA, and that lipid-poor SAA associates with apoB particles. Thus, the remodeling of HDL,
particularly in the post-prandial period, may lead to SAA shifting from HDL to apoB
particles; alternately, SAA could associate with apoB particles during their hepatic
secretion. Both HDL remodeling and hepatic apoB- particle secretion are increased in insulin
resistant conditions.

Role of lipoprotein-proteoglycan interactions in atherogenesis: There are several hypotheses
as to what triggers the initiation of atherosclerosis, with the "Response to Retention"
hypothesis well supported by biomedical evidence. As outlined in this theory, early fatty
streak lesions are initiated by deposition of atherogenic lipoproteins (LDLs and TGRLs) in
the subendothelial matrix by their retention by extra cellular matrix proteoglycans. Studies
show that lipoproteins migrate in and out of the subendothelial space, but once bound to
proteoglycans these lipoproteins are retained in this region, become more susceptible to
oxidation and other modifications, and are taken up by macrophages leading to the formation
of foam cells. TGRLs may be even more atherogenic than LDLs as they don't need modification
to be taken up by macrophages, and deliver 5-20 times more cholesterol than LDL on per
particle basis. The investigators have demonstrated the presence of SAA on apoB-containing
lipoprotein particles in mice, and recently confirmed this in humans. In preliminary studies
the investigators demonstrate that the presence of SAA on apoB-lipoproteins enhances their
proteoglycan binding. The investigators propose that the presence of SAA on apoB-containing
lipoproteins enhances their retention increasing atherogenesis.

Inclusion Criteria:

Up to 80 U.S. veterans age 50-75 will be recruited in the following three groups:

- Obese (BMI 27-45 kg/m2), metabolically healthy, (25-30 subjects)

- Obese (BMI 27-45 kg/m2), metabolic syndrome, (25-30 subjects)

- Obese (BMI 27-45 kg/m2), diabetic, (25-30 subjects)

Exclusion Criteria:

The use of:

- Statins (we will not exclude subjects on lipid lowering medications if they are
willing to discontinue them for 1-2 weeks prior to participation)

- Fibrates

- Niacin

- Anti-inflammatory drugs including Thiazolidinediones, non-steroidal
anti-inflammatories (NSAID), aspirin, steroids

- Estrogen replacement

Conditions such as:

- Acute illness

- Chronic inflammatory illness (such as psoriasis, rheumatoid arthritis, lupus, etc.)

- Infections

- Impaired renal function (eGFR < 60 ml/min)

- Hypo- or hyperthyroidism (subjects biochemically euthyroid on levothyroxine therapy
are permitted)

- Gastrointestinal dysfunction

Lifestyles including:

- Use of tobacco products

- Consumption of > 3 drinks /day