The Dark Side of HDL

High-density lipoprotein (HDL) — the molecule that normally scours from cells of vessel walls the excess cholesterol deposited there by low-density lipoprotein (LDL) — is itself vulnerable to corruption and conversion into a destructive form.

The modified, oxidized HDL and its major structural protein, apolipoprotein A1 (apoA1), not only are rendered dysfunctional, losing their cholesterol-scavenging capability, but gain potent pro-inflammatory capacities that activate endothelial cells and potentially contribute to vulnerable plaque development. Dysfunctional apoA1 is abundant in atherosclerotic plaques, and elevated levels of the oxidized, malevolent protein in patients’ blood have been shown to be associated with increased coronary artery disease (CAD) risk.

Insights more than a decade in the making

This profoundly altered view of HDL and its role in heart disease results from more than a decade of investigation by Cleveland Clinic cardiovascular researcher Stanley Hazen, MD, PhD, Head of the Section of Preventive Cardiology and Rehabilitation and Vice Chair of Translational Research, Lerner Research Institute.

His research team’s latest study, published in February’s Nature Medicine, reveals at the structural level how HDL is corrupted. Using innovative techniques, the team determined the HDL molecule’s complex atomic structure. Next they identified a targeted site (Trp72) on apoA1 where oxidation occurs, which disrupts apoA1/HDL cholesterol acceptor function. Then they genetically engineered an antibody that specifically recognizes the corrupted form of HDL and apoA1. The findings may soon produce a new diagnostic test to quantify patients’ dysfunctional HDL — and may eventually have therapeutic implications.

“We’re slowly decoding the structure and modifications that happen to apoA1 and HDL in the artery wall and understanding how that leads to changes in function,” Dr. Hazen says. “We’re starting to understand what’s going on at a very detailed, structural level, and the next step is figuring out how to block that.”

Unscrambling the HDL paradox?

HDL’s susceptibility to conversion from beneficial to harmful may help explain the HDL paradox that has baffled researchers.

Epidemiologic studies repeatedly have shown an inverse association between circulating HDL cholesterol or apoA1 levels and CAD: the higher the level of HDL cholesterol or apoA1, the lower the prevalence of CAD and risk of cardiovascular problems. But several interventional studies testing HDL cholesterol-raising drugs failed to show a reduction in cardiac risk, and studies testing direct infusions of different HDL formulations have yielded mixed results.

To Dr. Hazen, that strongly suggests the HDL particle’s functionality (or dysfunctionality), not its bulk amount, is more clinically important — a hypothesis his latest findings support.

From inflammation pathways to oxidative damage

Dr. Hazen, whose training is in biochemistry as well as internal medicine, didn’t set out to study HDL. His research had been focused on deciphering the inflammation pathways and processes in atherosclerosis.

His particular interest has been myeloperoxidase (MPO), an enzyme secreted by activated neutrophils and monocytes at sites of inflammation, including within atherosclerotic lesions. MPO churns out free radicals and diffusible oxidants that are toxic to microbes. But MPO also promotes oxidative damage of host tissues. Since MPO accumulates in the subendothelial space of the artery wall, that damage fosters development of CAD. Elevated MPO levels are associated with increased risk for CAD and coronary events.

Dr. Hazen’s team began trying to identify the major targets that MPO oxidizes and modifies in artery walls. HDL — and specifically apoA1 — emerged as the bull’s-eye. “If you quantify the degree of MPO’s oxidation of apoA1 vs. other proteins in its surroundings within the artery wall, it’s more than 500-fold selectively targeted,” Dr. Hazen says.

HDL as the bomb squad

Why the selective targeting? It happens because MPO binds directly to HDL in the artery wall. One of HDL’s jobs appears to be snaring potentially harmful enzymes and carrying them away for elimination. So it needs an accessible docking mechanism. “HDL has evolved to bind to and sequester heme proteins like MPO that can make reactive species,” Dr. Hazen says. “You need a way to get rid of the land mines. I call HDL the bomb squad.”

In this case, though, the hazardous cargo disables its transporter while also turning it toxic. Dr. Hazen’s research shows that MPO’s oxidation of apoA1 and HDL severely impairs the lipoprotein’s cholesterol-accepting ability and converts it into a pro-inflammatory particle.

Surprising findings on the lipoprotein’s structure

Confirming MPO’s selective affinity for HDL/apoA1 and understanding the process and disastrous consequences of oxidation required deeper knowledge of HDL’s structure. Dr. Hazen’s team needed to determine the lipoprotein’s shape and the architecture of its binding sites.

The traditional method for visualizing protein structures, X-ray diffraction, would require crystallizing HDL, which no one has been able to do. So for more than a decade, Dr. Hazen worked on alternative visualization means. His team was the first to use a technology called contrast variation neutron scattering to systematically map HDL. The results were a surprise.

Computational models of HDL had suggested it was a bilayered disc, like a coin, with a ring of protein around its rim. Neutron scattering revealed a much more complex structure. The protein and lipid components of spherical HDL, the most abundant form of HDL in the blood, were directly visualized, revealing a complex shape where the apoA1 surrounded a lipid core (Figure).

“HDL used to be thought of as just an oil slick with a protein like a serpent inside it,” Dr. Hazen says. “But it’s nowhere near that uncontrolled. It has a very defined structure, and the protein is not highly rigid; it’s highly dynamic. But it follows set rules of structure.”

Another analytical method called hydrogen/deuterium exchange enabled the team to map key contact sites on the apoA1 molecule of HDL with HDL-associated proteins. Proteomic studies of apoA1 from lesions identified the site at Trp72 where MPO oxidatively modifies apoA1.

Finding a superantibody to target dysfunctional HDL

Armed with that structural and binding site knowledge, Dr. Hazen began developing a monoclonal antibody to recognize MPO-oxidized apoA1. This would allow the researchers to determine diagnostically how much dysfunctional HDL was present in vivo. It also might function as a therapeutic aid, facilitating the corrupted lipoprotein’s removal.

Dr. Hazen’s team screened more than 30,000 candidate antibodies before identifying one with the right binding characteristics. But the antibody’s affinity was too low to be diagnostically reliable. The researchers had to carefully genetically modify the antibody to amplify its affinity without altering its binding characteristics. “In the end, we increased the antibody’s affinity more than 1,600-fold,” Dr. Hazen says.

Using this superantibody, the researchers confirmed in 2013 that atherosclerotic plaque-laden human aortas are teeming with dysfunctional apoA1, while there is far less in healthy vessel walls. And unlike in circulating blood, where apoA1 rides within HDL, the dysfunctional apoA1 in atherosclerotic plaque is lipid-free, unassociated with HDL.

What’s next? A diagnostic test and future research

Cleveland HeartLab, a spinoff company from Cleveland Clinic, is developing a diagnostic test for arterial inflammation and cardiac risk based on the antibody biomarker for dysfunctional apoA1. The company hopes to have the assay commercially available by the end of 2014, Dr. Hazen says.

Though several pharmaceutical companies are testing potential MPO-inhibiting drugs — which presumably would block inflammation — none currently is approved for use. So patients who test positive for dysfunctional apoA1 probably would be advised to take preventive steps to reduce cardiovascular risk, Dr. Hazen says, such as lowering LDL levels, increasing exercise, and controlling weight, blood pressure and diabetes.

Going forward, Dr. Hazen’s team will attempt to identify additional HDL sites where oxidation-induced dysfunctionality can occur.

“We used to have good and bad cholesterol,” he says. “I think we’re now going to have dysfunctional HDL, and that may be a composite of several things. The real question is what you do about it. That will be the next five, 10 years of work.”

Image was originally published in the Journal of Biological Chemistry (2011;286[14]:12495-12508). ©The American Society for Biochemistry and Molecular Biology.