American Association for Clinical Chemistry
Better health through laboratory medicine
December 2007 Clinical Laboratory News: HDL-C

December 2007: Volume 33, Number 12

The Changing Testing Paradigm
Part 2

By Alan T. Remaley, MD, PhD, and G. Russell Warnick, MS, MBA




Assessing a patient’s risk of coronary heart disease (CHD) routinely involves measuring one of the major classes of lipoprotein particles, high-density lipoprotein (HDL). The concentration of cholesterol carried on HDL particles, HDL-C, has been the main metric for this assessment, originating from early studies that first showed the link between cholesterol and atherosclerosis. But some patients whose HDL-C levels and other traditional risk factors fall within the recommended levels set by the National Cholesterol Education Program’s (NCEP) Adult Treatment Panel III (ATP III) still suffer adverse CHD events.

Research has now conclusively shown that the major lipoprotein classes are not homogeneous entities, but rather heterogeneous and polydisperse populations of particles with varying composition, physical characteristics, and pathophysiologic significance for the development of atherosclerosis. Researchers have discovered that not all HDL-C particles confer the anti-atherogenic properties ascribed to this so-called “good” cholesterol molecule.

In the first part of this review (CLN, November 2007, page 10), we described the latest findings on the atherogenic properties of HDL and the status of HDL lab measurements. Here we discuss the various lab methods for measuring the subclasses of HDL and the current controversy in the use of these measurements in predicting CHD.

Analytical Methods for Measuring HDL Subclasses

Table 1 lists the various methods for determining HDL subclasses. The assays each resolve and detect HDL subclasses using different physical properties, so the results of different assays are not necessarily interconvertible.

Table 1
Methods for Analysis of HDL Subclasses
Number of HDL Subclasses
Research Tests analytical ultracentrifugation   density reflectance
density gradient ultracentrifugation   density cholesterol
2D-gel electrophoresis   charge and size immunoblot of apolipoproteins
Reference Tests gradient gel electrophoresis Berkeley HeartLab
(Burlingame, Calif.)
4-30% stepped gradient gel of d<1.21 protein by
Coomassie dye
5 + E-rich
nuclear magnetic resonance LipoScience
(Raleigh, N.C.)
proton shift 0.8ppm terminal methyl group
protons of phospho-lipids, triglycerides, free and esterified cholesterol
vertical auto profile Atherotech
(Birmingham, Ala.)
vertical rotor ultracentrifugation cholesterol
Products gel electrophoresis
(Redondo Beach, Calif.)
linear gel Sudan black pre-stain
high-pressure liquid chromatography
Skylight BioTech
(Okazaki, Japan)
gel permeation cholesterol


Scientists first used analytical ultracentrifugation to resolve HDL subclasses. This method was followed by density gradient ultracentrifugation, which can resolve up to five different subclasses of HDL. While tedious to perform, it is still considered the gold standard for separating HDL lipo-protein subclasses.

A higher resolution method, two-dimensional agarose gel electrophoresis, can resolve up to 14 different subclasses of HDL. In the first dimension, HDL subclasses separate largely based on charge into pre-beta, alpha, and pre-alpha migrating forms. The second dimension separates the particles by size under nondenaturing conditions (1). This method is technically demanding, very tedious, and therefore impractical for clinical laboratories. However, a variety of subclass separation methods, including the two-dimensional separations, have consistently revealed that the large apoA-I containing lipid-rich, less-dense forms of HDL appear to be atheroprotective. Furthermore, in some studies the smaller, denser forms of HDL positively correlate with CHD risk (1, 2).

The remaining tests in Table 1 are more practical alternatives for HDL subclass identification and are currently offered by reference labs that specialize in advanced lipid and lipoprotein testing. Berkeley HeartLab (Burlingame, Calif.) offers a stepped 4–30%, nondenaturing, gradient polyacrylamide gel separation for HDL subclass separation (2). After the HDL subclasses are separated on the gel and stained, the gel is analyzed by densitometric scanning to quantify the amount of each subclass (Figure 1, below). Initially validated against analytical ultracentrifugation, this method electrophoretically resolves HDL and an apoE -rich form of HDL that can migrate with LDL into five subclasses similar to that seen in ultracentrifugation. Consistent with other methods, gradient gel electrophoresis reveals that the larger apoA-I containing particles are more atheroprotective (2).   

Figure 1
Analysis of HDL Subclasses By Gradient Gel Electrophoresis

Panel A: Stained polyacyrlamide gel showing separation of HDL subclasses. The arrow indicates origin of gel. Size standards are present at ends of the gel and in the middle lane.

Panel B: Densitometric scan of a representative gel. Top right panel is a gel from an individual with a low level of a large HDL subclass (HDL2b), whereas the bottom right panel is from a subject with a high level of HDL2b.


A simpler method for the electrophoretic separation of HDL uses a linear tube gel that separates HDL into three subclasses. Quantimetrix Corporation (Redondo Beach, Calif.) is developing this assay, which is similar to an LDL-subfraction method that is currently offered by the company. This type of assay would fall within the technical capability of most clinical labs.

Another test for resolving HDL subclasses uses nuclear magnetic resonance (NMR). LipoScience’s (Raleigh, N.C.) NMR analysis of lipoproteins depends on the proton vibrational signal generated by the terminal methyl groups on the lipids in HDL. The frequency and shape of the NMR signal is affected by the lipoprotein particle size (3). After a mathematical deconvolution of the signal, the method provides the amount of three subclasses of HDL. LipoScience calibrated the assay using different HDL subclasses isolated by density gradient ultracentrifugation. In addition to HDL subclass information, the method also yields information on HDL particle number, which, like LDL particle number by NMR versus LDL-C, appears to be superior to HDL-C for predicting CHD risk (4).

Another method for separating HDL into two subclasses by density involves a vertical rotor. A continuous-flow cholesterol analyzer, called a vertical auto profiler (VAP), monitors fractions from the centrifuge tube for cholesterol (5). Compared to standard ultracentrifugation procedures, VAP analysis is easier and faster to perform. Atherotech (Birmingham, Ala.) offers this test, along with other lipid and lipoprotein testing.

The final method for separating HDL subclasses is based on a method developed by Okazaki and colleagues (6). The method separates HDL particles into four size forms by high-pressure liquid chromatography (HPLC). Skylight BioTech, Inc. (Akita, Japan) offers this method under the name LipoSEARCH.

HDL Immunoassays

Researchers have also worked to develop immunoassays against specific subfractions of HDL. An ELISA produced by Daiichi Sankyo (Tokyo, Japan) and distributed in the U.S. by Polymedco (Courtland Manor, N.Y.) measures pre-beta HDL, the lipid-poor discoidal-shape HDL that is especially good in mediating cholesterol efflux from cells by the ABCA1 transporter (See Figure from part 1 of this article, below). Interestingly, although pre-beta HDL has anti-atherogenic properties when tested in vitro, its concentration in plasma positively correlates with CHD risk (7). It may be that the accumulation of pre-beta HDL in vivo indicates that there is an aberration in the HDL maturation process, and when this occurs there is a decrease in the net flux of cholesterol by the reverse cholesterol transport pathway. In fact, genetic defects in HDL metabolism—such as mutations in the ABCA1 transporter in Tangier disease—lead to an overall decrease in HDL but to a relative increase in pre-beta HDL and an increased risk for CHD (8).


Table 1
The Reverse Cholesterol Transport Pathway and Other Sites of HDL Action on the Pathogenesis of Atherosclerosis


This diagram depicts the various known mechanisms whereby HDL protects against the development of atherosclerosis. The so-called “reverse cholesterol transport pathway” is believed to be the main protective effect of HDL (1). This pathway results in the transfer of excess cholesterol from peripheral cells—such as macrophages in atherosclerotic plaques, —to HDL, which eventually transfers its cholesterol to the liver for excretion. HDL is made in the liver and intestine where apoA-I, the main protein of HDL, is produced. Nascent HDL is formed when lipid-poor apoA-I, such as the small discoidal-shaped particle with pre-beta mobility on agarose gels, is secreted and interacts with the ABCA1 transporter on the surface of hepatocytes and enterocytes (4). This interaction results in the transfer of phospholipids and a limited amount of cholesterol to apoA-I to form a more lipid-rich, intermediate size form of HDL. Such HDL particles can then remove additional cholesterol from peripheral cells by ABCA1 and other transporters, such as the ABCG1 transporter, which primarily transfers cholesterol to only the larger, lipid-rich forms of HDL. Unesterified cholesterol on the particle surface is then moved to the hydrophobic core of HDL after its conversion to cholesteryl esters by lecithin:cholesterol acyltransferase (LCAT), creating a spherical-shaped HDL particle. HDL then returns cholesteryl esters to the liver via the SR-BI receptor, where apoA-I is recycled and returned to the circulation. Cholesterol is off-loaded and either converted into a bile salt by hepatocytes or is directly excreted into the bile. A significant fraction of cholesterol on HDL is also delivered to the liver by way of the LDL-receptor, after first being transferred from HDL to LDL by CETP.


HDL Functional Assays

Based on the knowledge of HDL’s in vivo function, researchers have also attempted to develop functional assays for measuring it, such as the ability of HDL to remove cholesterol from cells, to inhibit the expression of the adhesion protein VCAM-I on endothelial cells, and to inhibit the production of pro-inflammatory cytokines from macrophages. At this time, these assays are mostly reserved for research studies, but they may lead to new insights into developing more practical tests for assessing the anti-atherogenic functions in the clinical laboratory.

One approach that shows early promise is a test for the anti-oxidative capacity of HDL (9). The first assays of this function were complicated, involving the use of cells, but cell-free assays have now been developed that could potentially be automated. The anti-oxidative capacity of HDL appears to be particularly useful in distinguishing fully-functional good HDL from dysfunctional or even bad-good HDL. In small studies, the assay has even been shown to be better than HDL-C in predicting CHD risk (9).

Outlook for HDL Testing

Clearly, there is still much uncertainty about how HDL works to protect against CHD and how its composition impacts its function. In terms of testing in clinical labs, compositional assays are almost always easier to implement than functional assays. But the current analytical procedures used to isolate and measure HDL may alter its composition, particularly for those constituents that are weakly associated.

There are many ongoing studies examining the clinical utility of HDL subclass assays in different populations, and more information on their clinical utility should be available shortly. Importantly, at a December 2006 meeting of the FDA’s Clinical Chemistry and Clinical Toxicology Devices Panel, members of the advisory board recommended that HDL and LDL lipoprotein subfraction tests may be useful to help determine the need to treat patients at intermediate risk for CHD as determined by more conventional tests. They further advised that more data are needed to support more widespread use of these assays.

In summary, although HDL-C has been useful for predicting CHD risk and for managing patients with dyslipidemia, its predictive ability is still rather limited as demonstrated by the Torcetrapib (Pfizer, New York, N.Y.) clinical trial discussed in part one of this review. Briefly, the drug raised HDL-C approximately 50% by increasing the amount of large-sized HDL; however, the trial was stopped because participants experienced a significant increase in cardiovascular deaths (10).

Future advances in HDL testing will most likely go beyond just measuring cholesterol content. Evidence is accumulating that HDL-C is not synonymous with HDL and that this single parameter of HDL does not embody all its diverse compositional, structural, and functional characteristics. Better indices of HDL function, structure, and quality are clearly needed now for research and drug development, and perhaps ultimately in the establishment of HDL-based treatment goals for patients.

For most clinical labs, the only viable alternative to measuring HDL-C is apoA-I, which is probably underused, given its superior predictive ability over HDL-C in many clinical studies. Consequently, there is a growing interest and support in the possible inclusion of apolipoprotein tests in national testing guidelines for CHD risk as an alternative to HDL-C and LDL-C tests.

While HDL subfraction tests currently available from specialty reference labs can be useful for deciding how to treat patients at intermediate risk for CVD, such tests will likely have limited use until more convenient formats become available that are suitable for routine clinical lab use. With new insights gained from research on HDL metabolism and better characterization of the proteome and lipidome of HDL, clinical laboratories will likely have tests that better distinguish the good and “bad-good” forms of HDL. Such tests represent the next major advance in lipid and lipoprotein testing for CHD risk assessment.


  1. Asztalos BF, Collins D, Cupples LA, et al. Value of high-density lipoprotein (HDL) subpopulations in predicting recurrent cardiovascular events in the Veterans Affairs HDL Intervention Trial. ATVB. 2005; 25:2185–91.
  2. Warnick GR, McNamara JR, Boggess CN, et al. Polyacrylamide gradient gel electrophoresis of lipoprotein subclasses. In: Smith DA, guest editor. Clinics Lab Med. 2006; 26:803–46.
  3. Jeyarajah EJ, Cromwell WC, Otvos JD. Lipoprotein particle analysis by nuclear magnetic resonance spectroscopy. In: Smith DA, guest editor. Clinics Lab Med. 2006; 26:847–70.
  4. Otvos JD, Collins D, Freedman DS, Shalaurova I, Schaefer EJ, McNamara JR, Bloomfield HE, Robins SJ. Low-density lipoprotein and high-density lipoprotein particle subclasses predict coronary events and are favorably changed by gemfibrozil therapy in the Veterans Affairs High-Density Lipoprotein Intervention Trial. Circulation 2006; 113(12):1556–63. 
  5. Kulkarni KR. Cholesterol profile measurement by vertical auto profile method. In: Smith DA, guest editor. Clinics Lab Med 2006; 26:787–802. 
  6. Okazaki M, Usui S, Fukui A, et al. Component analysis of HPLC profiles of unique lipoprotein subclass cholesterols for detection of coronary artery disease. Clin Chem 2006; 52:2049–53
  7. Hirayama S, Miida T, Miyazaki O, Aizawa Y. Pre beta1-HDL concentration is a predictor of carotid atherosclerosis in type 2 diabetic patients. Diabetes Care 2007;30(5):1289–91.
  8. Zannis VI, Chroni A, Krieger M. Role of apoA-I, ABCA1, LCAT and SR-BI in the biogenesis of HDL. J Mol Med 2006;84:276–94. 
  9. McMahon M, Grossman J, FitzGerald J, Dahlin-Lee E, Wallace DJ, Thong BY, Badsha H, Kalunian K, Charles C, Navab M, Fogelman AM, Hahn BH. Pro-inflammatory high-density lipoprotein as a biomarker for atherosclerosis in patients with systemic lupus erythematosus and rheumatoid arthritis. Arthritis Rheum 2006;54:2541–9.
  10. Schaefer EJ, Asztalos BF. Where are we with high-density lipoprotein raising and inhibition of cholesteryl ester transfer for heart disease risk reduction? Curr Opin Cardiol. 2007;22:373–8.  

Alan T. Remaley, MD, PhD, is a member of the senior staff of the Department of Laboratory Medicine at the National Institutes of Health (Bethesda, Md.) and is the Section Chief of the Lipoprotein Metabolism Laboratory at the National, Heart, Lung and Blood Institute. He has served on the FDA’s Chemistry and Clinical Toxicology Panel since 2002. 




G. Russell Warnick, MS, MBA, is Chief Scientific Officer and Vice President for Laboratory Operations at Berkeley Heartlab, Inc. (Burlingame, Calif.). Previously he directed the Core Lipoprotein Laboratory of the Northwest Lipid Research Center, associated with the University of Washington (Seattle) and founded Pacific BioMetrics, Inc. (Seattle, Wash.), a specialty service laboratory supporting pharmaceutical and diagnostic development. He serves as editor of The Fats of Life, a newsletter published by the Lipoproteins and Vascular Disease Division of AACC.